Back to previous post: Reading for Rockets

Go to Making Light's front page.

Forward to next post: USA Sweet USA

Subscribe (via RSS) to this post's comment thread. (What does this mean? Here's a quick introduction.)

May 26, 2015

E Pluribus Hugo: Out of Many, A Hugo
Posted by Abi Sutherland at 02:11 PM * 826 comments

In this thread we will hammer out the formal language of the proposal, any FAQs we wish to include, and strategize for the presentation at the business meeting itself. At this point, we’ll consider the system itself locked in, so we are really only looking at the language.

1. RME instead of 6th place
2. (1,1), (1,2), or (1,2,2) for ties in points
3. Option 2a (if there is a tie for nominations, eliminate the one with fewer points; if there is a tie for both nominations and points, eliminate them both)

There is one more issue that is still up for debate: Should we explicitly empower the Hugo admins to use further tie breakers in the future if they decide it’s necessary? I’ve written the proposal and FAQ explanations assuming that we do, however, a case can be made for not worrying about giving them the power explicitly. We should settle that question here. I think that the way I’ve written the “empowerment” makes it okay to include it, but for myself, I don’t feel a strong need to. I’m definitely not opposed to it, however.

If I’ve forgotten any signers, please let me know!!

Kilo


Short Title: E Pluribus Hugo (Out of the Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations), section 3.9 (Notification and Acceptance), and section 3.11 (Tallying of Votes) as follows:

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the following multi-round process described in 3.8.8.
3.8.2: The Worldcon Committee shall determine the eligibility of nominees and assignment to the proper category of works nominated in more than one category.
3.8.3: Any nominations for “No Award” shall be disregarded.
3.8.4: If a nominee appears on a nomination ballot more than once in any one category, only one nomination shall be counted in that category.
3.8.5: No nominee shall appear on the final Award ballot if it received fewer nominations than five percent (5%) of the number of ballots listing one or more nominations in that category, except that the first three eligible nominees, including any ties, shall always be listed.
3.8.6: The Committee shall move a nomination from another category to the work’s default category only if the member has made fewer than five (5) nominations in the default category.
3.8.7: If a work receives a nomination in its default category, and if the Committee relocates the work under its authority under subsection 3.2.9 or 3.2.10, the Committee shall count the nomination even if the member already has made five (5) nominations in the more-appropriate category.
3.8.8: The final Award ballots shall list in each category the eligible finalists as determined by successive rounds of a two-phase elimination process. In this process, each member gets a single nomination “point” for each category, and that point will be divided equally among their nominated works in that category.
3.8.8.1: Selection Phase: In each round, all the works having the least number of points will be selected for the Elimination Phase (3.8.8.2). If there is only one work with the least number of points, then all of the works with the second-least number of points will be also be selected for the Elimination Phase in addition to the lowest-point work.
3.8.8.2: Elimination Phase: Of the works identified in the Selection Phase, the one(s) that appear on the fewest number of nomination ballots will be removed from all nomination ballots for subsequent rounds as if they had never appeared on any ballots.
3.8.8.4: Ties: If two or more works are tied for appearing on the fewest number of nomination ballots, the tied work with the lowest point total will be eliminated. If there is a tie for appearing on the fewest number of nomination ballots as well as for lowest point total, then all members of that tie will be eliminated. Should they deem it necessary to do so in the future, the Hugo administrators are empowered to take further measures to break this type of tie, provided those measures are announced at the beginning of the nomination period for the Hugo Award.
3.8.8.5: If elimination would reduce the number of finalists to fewer than the minimum number, then instead none of the works from that round shall be eliminated, and all remaining works shall appear on the final ballot, extending it if necessary.
3.8.8.6: Subsequent rounds begin by reallocating points as follows: All works eliminated in previous rounds are removed from all nomination ballots and treated as if they had never appeared on any nomination ballot. Members’ points are then reallocated equally among their remaining nominated works, if any.

Section 3.9: Notification and Acceptance.
3.9.1 Worldcon Committees shall use reasonable efforts to notify the nominees finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each nominee finalist shall be asked at that time to either accept or decline the nomination. If the nominee any finalist(s) declines the nomination, that nominee finalist(s) shall not appear on the final ballot. In this event, once all finalists have had the opportunity to decline, the nomination system shall be re-run as described in Section 3.8 with the declined nomination(s) removed from the nomination ballots on which they appeared. The eligible finalists from this re-run shall be merged with the remaining potential finalists from the original run. If this merge would result in more than the maximum number of finalists, then the ballot shall be extended to include the finalists from both the original and the re-run of the nomination system. This procedure shall also be used in the event that a finalist is deemed ineligible.

Section 3.11: Tallying of Votes.
3.11.4: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. During the same period a record of all rounds of the selection process for each category shall also be published.

Submitted by: Keith “Kilo” Watt, Tammy Coxen, Yoana Yotova, Joshua Kronengold, Chris Battey, CJ Cabourne, Steven Halter, P.J. Evans, David Goldfarb, Seth Gordon, Ginger Tansey, Steve Wright, Catherine Faber, Andy Holloway, Duncan J. Macdonald, Claudia Beach, Derry Earnshaw, Jason Skiles, Soon Lee, David Harmon, Lydy Nickerson

Commentary: “E Pluribus Hugo” - the name of this proposal is also its goal: recognizing the many opinions within fandom as to what works might be worthy of the Hugo. It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members: They still list the works, unranked, they feel are Hugo-worthy (up to the maximum permitted) in any categories they choose, just as they have in the past. In the past, we have counted the number of nominations each work received, and the top five works were put on the final ballot. However, because SF fandom typically nominates a large number of different works, it was easy for an organized slate to make it so that no other works made the final ballot.

Using this system, fandom isn’t penalized for nominating a wide variety of works. If you nominate something that ends up not having a chance to make the final ballot, then your remaining choices automatically get more of your support instead of just being wasted. In other words, you can safely nominate anything you feel is Hugo-worthy. If enough people agree with you, it will make the final ballot. If they don’t, that’s okay - when that work is eliminated, your other choices will have a greater chance of making the final ballot. In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which works should be voted on to be the final Hugo winner.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no minority of members - of any sort - can disproportionately dominate an entire category. This system allows the broadest range of works that are popular with fandom-at-large to be considered for the Hugo Award.


FAQ’s:
1. Can you explain the system in plain language?
The important thing to remember is that nothing changes in how you nominate. If you think a work is Hugo-worthy, then nominate it. That’s all. There’s no need to rank your choices at the nomination stage, and there’s no reason not to nominate something you think even might be Hugo-worthy. All we are doing at this stage is throwing names into a hat. The final Hugo voting system, which actually chooses the winner, is unchanged. We could, in theory, simply put everyone’s nominations on the final ballot, but that would make for a very long ballot indeed. We therefore need to narrow the nomination list down. This system narrows down the list by eliminating the least popular works until only five (under current rules) finalists remain. Here are the basic steps to the elimination process:
a. You have one nomination “point” for each category that will be divided equally among the works you choose to nominate in that category. So, if you nominate two works in a category, each will get half a point; if you nominate three works, each will get one-third of a point, and so on.
b. All the points given to each work from all nomination ballots are added together. The two works that got the least number of points are eligible for elimination. One of these works is the least popular and will be eliminated. (We call this the Selection Phase.)
c. To determine which of these two works is least popular, we compare the total number of nominations they each received (that is, the number of nomination ballots on which each work appears). The work that received the fewest total number of nominations is the least popular and now completely vanishes from the nomination process as though it never existed. (We call this the Elimination Phase.)
d. We start over for the next round and repeat the process, however, if one of your works was eliminated, then you now have fewer works on your nomination ballot. This means that each work gets more total points, since you aren’t dividing your point among as many works. For example, if one of your five nominated works was eliminated, your remaining works now get one-fourth of a point each instead of one-fifth of a point. If four of your nominated works are eliminated, your remaining work now gets your full point.

2. E Pluribus Hugo?
The designers felt this name accurately reflects what the Hugo nomination system should be trying to achieve. Fandom is the Many. Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in SF. This is a defining characteristic of fandom, and this system embraces it. Out of all of these many tastes in works, this system allows a single set of nominations to slowly emerge as the most popular candidates for the Hugo Award: “Out of the Many, a Hugo.”

3. But surely that’s not what the system is formally called?
Okay, sure. Formally, in the context of election theory, this system would be called “single divisible vote with least popular elimination” or SDV-LPE for short. E Pluribus Hugo is both cooler and geekier, as everything is better with a little Latin.

4. How does this system eliminate slate or bloc voting?
It doesn’t, exactly, nor should a work be automatically eliminated just because it appears on a slate. On the other hand, any slate which nominates a full set of five works will find that each of its nominations only count 1/5 as much. With “non-slate” nominating, some of your works will be slowly eliminated, so your remaining works get more and more of your support. Since slate works tend to live or die together, they tend to eliminate each other until, in general, only one slate work remains. With a large enough support behind the slate (five times as much), the slate may still sweep a category; however, if that many voters support the slate, they arguably deserve to win, and no fair and unbiased system of nomination will prevent that. The answer in that case is, simply, to increase the general pool of voters. Regardless, with SDV-LPE, slates will never receive a disproportionate share of the final ballot, as occurred in the 2015 Hugos.

5. What if there are multiple slates (slate wars, “parties”, etc.)?
As with a single slate, the more works a slate nominates, the less their nominations count for each work. The end result is that even multiple slates are unable to sweep the nominations.

6. What happens if a broadly popular work is nominated by a group of unrelated people?
If it is broadly popular, the system will still select that work for the final ballot.

7. What happens if a broadly popular work also appears on a slate?
Even if it is on a slate, if the work garners support from individuals then the system will select that work for the final ballot. Slates neither help nor hurt any given work.

8. Why re-run the system if someone declines a nomination? Why not just take the work that was eliminated in the next to last round?
We can do that, and in fact have an amendment for that ready if that’s what the people decide they want to do. The “re-run, merge, and extend” (RME) system described in the proposed language has only one major difference between it and taking the “6th place” work. Under RME, if a slate work makes the final ballot and that nomination is declined, the declined nomination will tend to be replaced by another work from the slate. Under “6th place”, if a slate work makes the final ballot and that nomination is declined, the declined nomination will tend to be replaced by a non-slate work. Either way works, but since our goal is for all sections of fandom to be fairly represented on the final ballot, RME seemed more in line with that philosophy.

9. Isn’t it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws. The objective is to choose a system whose flaws are not in an area of concern to the electorate.

10. What are SDV-LPE’s flaws?
In rare cases, it is possible for eliminating both members of a tie to change the final ballot slightly from what it would be if the tie were broken so that only one member was eliminated. The change is usually in the least popular of the finalists, and requires the two works in question be very close in popularity. This situation was extremely rare using realistic data (in fact it never occurred in any of our situations with realistic data), so the statistical probability of how ties are handled affecting the results is extremely small. There are a number of additional tie breakers that could be used if the Hugo administrators deemed it necessary; however, simulations that have been run by the designers of this system show that it really shouldn’t be required. The proposal language leaves open that option if later Hugo administrators want to add those layers of complexity to the process. SDV-LPE itself works the same regardless.

11. What are SDV-LPE’s benefits?
Simply put, it reduces the power of bloc voting without eliminating the chance that works that do appear on slates will make it to the final ballot. Conversely, it makes it very difficult for slates to prevent non-slate works from appearing on the ballot.

12. Couldn’t slates just recommend a single work for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate is large enough that is certainly a viable possibility - it’s also completely fair. It does not force any other works off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn’t a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a work deserves a Hugo nomination, then it should, in fact, be represented on the final ballot.

13. What happens with a large field with no stand-out favorites when a slate votes?
Even in this case, we were unable to find a simulation in which no non-slate works appeared on the final ballot. Slate works did receive a larger proportion of nomination slots than they did otherwise, however, again, this could be considered a fair and valid result. If there was no general favorite, then nominators really had no collective preference.

14. How do the results of this system compare to the results under the current nomination system?
Statistical tests showed that this system and the current system overlap in about 4.5 out 5 candidates. In the absence of slates, it generally gives identical results to the current nomination system.

15. I think we should just increase the number of nomination slots on the final ballot to (for example 6), and decrease the number of slots a voter can vote for to a smaller number (for example, 4). Wouldn’t that be simpler and easier?
Unfortunately, this simply means that the largest slate will receive four of the nominations and the next largest will receive the remaining two. It doesn’t solve the problem of forcing works off the ballot that had a chance to win the final election. In general, we want fandom to nominate as many works as they feel are Hugo-worthy, since under SDV-LPE there is no strategic reason not to do so. For this reason, most of the designers of this system would prefer that members not be limited to nominating only four works. Keep in mind, however, that SDV-LPE will work with this (or most any other) change as well, so one does not preclude the other.

16. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
This could work. The problem is that now you have a small group of people who serve as literal gatekeepers to the Hugo Awards. In spite of the word on the Internet, this has never been the case in the past. Establishing it now means that those groups who believed it existed in the past will now be correct. Ultimately, human judgement is fallible. The fairness of a committee’s decisions will forever be subject to opinion. The end result is that the prestige of the Hugo Awards will forever be tarnished.

17. I think we should use [insert other mathematical voting system].
We considered essentially every applicable type of voting system currently in the literature, guided by two experts in the field. It should be kept in mind, however, that the goals and requirements for choosing a set of representatives in a political situation are different from those for choosing a set of Hugo finalists. Some of these systems do, in fact, have positive properties that speak for them. None of them were as simple or as intuitive as SDV-LPE, yet SDV-LPE meets all of the stated goals for a Hugo nomination system.

18. Won’t SDV-LPE be complicated to code and implement?
Actually, no. One of our non-experts coded a full simulator for the system in a matter of days. A full web-based app would not be much more difficult to handle.

19. Wasn’t this system just designed by Social Justice Warriors to block the Good Stuff?
It is true that much of the discussion for this system occurred on Teresa and Patrick Nielsen Hayden’s “Making Light” discussion board, and it is also true that groups such as the Sad Puppies and the Rabid Puppies consider TNH and PNH to be The Enemy, and therefore completely biased and not to be trusted. Other than serving as occasional moderators, TNH and PNH had no real input in the discussions of the system, however. Those of us who worked on the system were very clear that our goal was not to keep the Sad/Rabid Puppies off of the Hugo ballot, and that any system which specifically targets any type of work is inherently wrong and unfair. One of the members of the group is a retired US Naval officer, a combat veteran, a certified Navy marksman, a Christian, and considers Robert Heinlein to be the greatest science fiction author who has ever lived. In short, he is exactly the Puppies’ demographic. But any slate, of any sort, be it a Sad Puppy or a Happy Kitten of Social Justice, breaks the Hugo Award because a small percentage of voters can effectively prevent any other work from appearing on the final ballot. This is a major flaw in the Hugo nomination system, and it is a flaw that must be fixed if the integrity of the award is to be maintained. Politics should play no role whatsoever in whether a work is Hugo-worthy or not.

Comments on E Pluribus Hugo: Out of Many, A Hugo:
#1 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 02:27 PM:

Welcome to the new thread!

As the initial post mentions, at this point we're going to consider the system itself locked in. The intent here is get the proposal language in as clear and unambiguous form as possible, make sure we have FAQs to handle the most common questions people will have, and then plan the presentation at the business meeting itself. As someone pointed out in the other thread, we only have less than 20 minutes to make our pitch, so we should focus on making things as tight as possible. There is a lot of support (both here and elsewhere) for the proposal, so I think we've got a good thing going here.

I'd like us to finish the admin discretion question, since that seems to be the only open issue remaining. I am recording changes to the proposal language in a Word file. Once we feel we're going to go, I'll ask Jameson to post the final document to his Google docs site for one last edit-and-check, and then we'll release the proposal to the wild for comments.

Sound like a plan?

Thanks,
Kilo

#2 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 02:28 PM:

Just to carry over what I wrote in the last thread...

Admin empowerment is an open issue, as far as I'm concerned. The only way admin empowerment could ever come into play is if in the past someone had serious problems with a tie break and wanted something different for the next year's Hugo. This would enable them to do that without an amendment. I don't think there's any way to give the admins discretion after the results have been determined (which is why I stipulated any changes had to be announced at the start of the Hugo nomination process).

Personally, I don't think it's needed. The cases where it even makes a difference are going to be so uncommon that once every 20 years just isn't going to be an issue. And when some case does comes up, the odds of it being so different that it's contentious are so small that I still don't think it's worth worrying about.

I'm open to what the group wants, but if left to me, I'd leave that part out and just do another constitutional amendment in 20 years if someone really wants to. I just don't see it as that big an issue.

Kilo

#3 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 02:32 PM:

Various typos and uncontroversial fixes from the end of the last thread:

Tim@997:
I think there's other problems in the wording (like the use of "works" in 3.8.8, which should be "nominee"), but we can tackle them in the next thread...

Aan@998:
Typo fix for FAQ #10:
in fact >it

#4 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 02:33 PM:

That was weird... I added one more in the preview, but it didn't show up in the actual post.

Kevin@1000:
Whatever you propose here, I hope it will use "nominee" to mean "work/person nominated in the initial voting" and "finalist" to mean "work/person appearing on the final ballot."

#5 ::: Jameson Quinn ::: (view all by) ::: May 26, 2015, 02:58 PM:

I aesthetically like the idea of leaving some power in admins' hands. But it's really not worth fighting about one way or another. I'm happy to stand behind Keith, as he's the guy herding us cats. So if Keith says admin discretion isn't worth it, I'll drop it. And I think I was making at least half of the pro-discretion noise, so I think if I drop it, it's probably dead.

On the FAQ:

12: remove "any"; that's overreach.

14: I'd like a way to say that 4.5 is a conservative estimate. Perhaps "an average of at least"? This is a bit technical and confusing, but the simple option of just "at least" makes it into nonsense...

Also: I'd really like to search-and-replace "nominations" with "approvals". "Nominations" is confusing because at first some people will assume that it refers to the outcome, not the input, of this process. I know that it's impossible to read this carefully and keep that misconception, but it is best if we can avoid it arising in the first place.

I'd suggest we drop the "five (5)" verbiage. This is ascii, nobody is going to misread a letter.

New FAQ question:

Why are there two phases (selection and elimination)? Couldn't you simplify it by having one step per elimination?

The selection phase looks for the works with the fewest points, and the elimination phase eliminates the one with the fewest approvals. Looking at two different measures of quality adds robustness to the system. With honest, non-slate voting, as in most historical Hugo elections, the works with the fewest points will tend to be the same ones as those with the fewest approvals. But while might be possible for strategic blocs to manipulate one of these two metrics, as the "puppy" factions did in 2015, it would be very hard for them to manipulate both in the same direction. A strategy that focuses on approvals for the elimination phase, such as bloc voting, will be stymied by the selection phase; a strategy that focuses on points for the selection phase, such as bullet voting, will turn out to be powerless or counterproductive in the elimination phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.

#6 ::: Aan ::: (view all by) ::: May 26, 2015, 03:03 PM:

Keith @ #4:

Yeah, you need to encode your < as &lt; otherwise it's interpreted as the start of an HTML element, and then stripped until the next > (which generally isn't present). Tripped me up on first preview as well...

#7 ::: Jameson Quinn ::: (view all by) ::: May 26, 2015, 03:07 PM:

Final update on the gofundme:

Fully funded, and beyond!

I'm truly in awe of the generosity this community has shown, both to me personally and to the cause of voting reform. Not only has the main campaign received $1440, beyond the goal of $1400; but I've also been offered a Sasquan attending membership, so in effect it's actually $250 over the goal. All of that extra money, and any further that is raised from here on, and any amount I can save by being frugal in Spokane, will go to support Electology.org (the Center for Election Science) in our work on voting reform in general.

For all those who have given: thank you so much. I'm now definitely going, with no worries about money, and I hope to see many of you at Sasquan.

If you're motivated to help even further, here are a few things you can do:

1. Share this campaign with anyone else who you think might be interested. Electology.org is a growing organization on a shoestring budget, and every bit helps.

2. If you can help me find shared lodging at Sasquan, that would be great. I'm a 40something family man, and I try to be a friendly, helpful, and respectful guest. I think you can email me via this campaign. [extra note: my google email is my first and last names, with a dot in between].

3. We'd love it if you'd sign up for our newsletter at electology.org, or browse around our website to learn more about voting reform issues.

4. If you're a Sasquan organizer, I'd be happy to have a short session where I and other panelists (such as Keith, of course, if he's willing) discuss this proposal, including the voting theory, statistics, and simulations that went into making it.

Again, thanks to everyone who donated.

#8 ::: Jameson Quinn ::: (view all by) ::: May 26, 2015, 03:08 PM:

Oh, and I want to sign on to the proposal, of course. My membership paperwork hasn't gone through yet, but I expect it will be in my name, as written here.

#9 ::: abi ::: (view all by) ::: May 26, 2015, 03:08 PM:

Forgot to mention that I fixed the typo Aan's reported when I posted, because it was easy and required no judgment on my part.

The fact that I have no spoons for actual judgmental changes is evinced by the fact that I did not report the fix.

#10 ::: abi ::: (view all by) ::: May 26, 2015, 03:12 PM:

You can add me to the list too. Abigail Sutherland.

#11 ::: Kevin Standlee ::: (view all by) ::: May 26, 2015, 03:14 PM:

Jameson Quinn @ 7:

If you're a Sasquan organizer, I'd be happy to have a short session where I and other panelists (such as Keith, of course, if he's willing) discuss this proposal, including the voting theory, statistics, and simulations that went into making it.
I suggest that you reach out to program@sasquan.org about this rather than expecting them to contact you. While I'm a Sasquan area head (Business Meeting), I have nothing to do with the scheduling of the rest of the convention.

Given that the first time this proposal will come before the meeting will be shortly after 10 AM Thursday (the second day of the convention), I suggest that if you're offered the opportunity to do a panel on this, you ask for sometime on Wednesday, the first day of the convention, and that you ask that if Sasquan schedules a panel about "Introduction to the Business Meeting" (a relatively common-but-lightly-attended panel on Day 1), they not schedule it against your discussion.

Alternatively, if you are confident that the Preliminary Business Meeting won't kill your proposal out of hand, you might ask for them to schedule such a panel sometime after 1 PM Thursday, so people who first encounter the proposal at the PBM can go to an informal "hearing" on the subject.

#12 ::: Zack ::: (view all by) ::: May 26, 2015, 03:16 PM:

I'm generally in favor of the proposal (and don't have time to help wordsmith language, unfortunately) but I think that there should be a separate proposal to limit the number of times a single person can be shortlisted, (for instance) "Within one year's Hugo ballot, no person may receive a nomination for more than one work per category, nor be nominated in more than three categories total. If the nomination procedure leads to someone receiving more nominations than this limit, the Hugo administrators shall ask the nominee which nomination(s) they would prefer to receive. If the nominee does not respond in a timely fashion the administrators may drop excess nominations at their discretion."

Is anyone working on something like that?

#13 ::: Jameson Quinn ::: (view all by) ::: May 26, 2015, 03:24 PM:

@11: Thanks for the advice. That's basically what I was hoping for from raising it here: not that somebody would be like "sí, cómo no, of course, here's your panel sir", but just that I might get a bit of advice about how and where to raise the possibility.

Also: Keith, do you want to be in on this? I'll be at Sasquan primarily for the purpose of doing this kind of thing, so I'm totally willing to work around your schedule.

@12: sounds like a good idea, but clearly separate from this one. Note that you'd have to clarify whether the Campbell award counts as a "category".

#14 ::: spacefaringkitten ::: (view all by) ::: May 26, 2015, 03:52 PM:

If this has already been discussed in previous threads, please just ignore me, but I feel that maybe there should be one Q&A in the FAQ along the lines of "Does this mean that I have to think strategically how many works I'll nominate?"

I've brought up this proposal in a couple of places, and the first reaction of many people seems to be something like that. Stating clearly that that isn't the case would maybe help.

#15 ::: Lee ::: (view all by) ::: May 26, 2015, 03:58 PM:

I will sign onto the proposal as well. Lee Billings.

#16 ::: JDC ::: (view all by) ::: May 26, 2015, 04:15 PM:

Suggested edits (forgive me if they are covered elsewhere in the rules or threads which I have not read):

In subsection 3.8.1, delete "following multi-round" as unnecessary. Add "subsection" before "3.8.8".

In subsection 3.8.8.1, suggest changing first sentence to read "In each round and for each category, all the works having the least number of points (or works in the case of a tie for least number of points) will be selected for the Elimination Phase (subsection 3.8.8.2).".

Subsection 3.8.8.3 is missing (that is, a numbering error).

In subsection 3.8.8.5, does "minimum number" need defining or expanding? Also suggest deleting "instead" as unnecessary. And is "extending" the accepted Worldcon usage? I would spell it out more, e.g. "regardless of the number of works on the final ballot".

Finally, a general observation that the concept of "category" does not appear as often as I think it should in the 3.8.8.n subsections. That is, selection happens within categories within the ballot, not merely in the ballot. This seems obvious but in my drafting I prefer to drive the obvious home. As my torts professor was fond of saying, "Legal writing is to writing as military music is to music."

#17 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 04:16 PM:

Jameson@13:

I'm not actually sure a panel is a good idea. I can see that an extended Q&A might be useful if it's necessary, but I'm -really- hoping that we write this proposal so that it's not. I really think a seminar on election theory would do more harm than good. I do recognize that's one of the goals of your organization, but I think we need to keep in mind that general voting systems isn't really the goal here. A technical seminar of any sort might just turn people off from the proposal. I might can see it on Sunday after the proposal has been voted on maybe.

My wife and I won't be arriving in Spokane until late on Wednesday night, and we will be leaving around 1:30 on Sunday (3:00 flight). In the between time I'll be talking up the proposal when I'm at the various parties and such when then opportunity arises, of course.

Once we have the formal proposal finished, we should discuss how we want to present it. I don't know if we have PowerPoint available or just talk, etc. But we can cross that bridge when we come to it.

Kilo

#18 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 04:21 PM:

Zack@12:

To my knowledge, no one is working on the type of thing that you're proposing. Kevin has asked that we keep things as separated as possible, so I don't think we should include it in this proposal. There's nothing stopping anyone from proposing it, however -- it won't affect SDV-LPE either way.

Kilo

#19 ::: Christopher Battey ::: (view all by) ::: May 26, 2015, 04:28 PM:

[SIGNING]

Apparently I'm registered with Sasquan as "Christopher Battey" (rather than "Chris Battey"). I usually go by Chris, but if the submitter list needs to match up against Sasquan's member list, I should be listed as Christopher.

#20 ::: Tim Illingworth ::: (view all by) ::: May 26, 2015, 04:32 PM:

Zack@12: if you want to write to me offlist (tim@smofs.org) I'll draft something for you.

#21 ::: Jameson Quinn ::: (view all by) ::: May 26, 2015, 04:54 PM:

@17: I think that people who think "math, ick" wouldn't come to that kind of panel; and the people who did come, would be the kind ready to listen to the theory. So I think it's a good idea. Of course, I would think that...

I think the best way to convince people that this is a good idea is to be as open as possible about how we're doing it. For instance, we're working things out here, in public; that's a good thing. I think a panel, in which we honestly help people understand the system, is more of the same.

But if you strongly oppose the idea of a panel, then me doing it without you might look quarrelsome. So I'd really like it if we could agree on this, one way or another.

#22 ::: Doire ::: (view all by) ::: May 26, 2015, 04:56 PM:

Well, it's printed out, double-spaced, and I'm wielding my editing pens. Please don't hate me.

I'll try to get the thoughts of someone who hasn't been involved, but should be able to get the ideas easily enough, in case any assumptions or other holes have snuck in. See you tomorrow.

#23 ::: Oskari Rantala ::: (view all by) ::: May 26, 2015, 04:59 PM:

[SIGNING], in case you take onboard people who support this but have not really done anything else than follow the discussion with interest.

I'm a supporting member and my name's up there.

#24 ::: rcade ::: (view all by) ::: May 26, 2015, 05:15 PM:

This looks good, and I'm evaluating whether to sign on to it.

Is there a full dataset available of any past year's Hugo nominating ballots with personally identifying information removed? I'd like to run simulations with data like that and add bloc voters at several different participation levels, to see what the final ballot would be.

Lacking that, I have trouble figuring out how to simulate a realistic nominating process.

#25 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 05:20 PM:

rcade@24:

Yes, we have the actual data for 1984 which we have been modifying for various situations (slates of various types, bullet voting, etc.). I believe it was PJ who made that available to us, so look back in the previous thread to find his email.

Additionally, we had another member generate a random sampling of ballots that was reasonably proportional to the 2013 data, though in that case we didn't have the actual ballots available.

Kilo

#26 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 05:22 PM:

Jameson@21:

I agree it should probably be both of us or none. Let's table it for now and look at it again when we're discussing the actual presenting at the meeting.

Kilo

#27 ::: P J Evans ::: (view all by) ::: May 26, 2015, 05:44 PM:

25
repeating it here (and not 'he', although I do answer to 'hey you'), with ROT-13:
cw.rinaf88@tznvy.pbz

#28 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 06:16 PM:

Whoops! Sorry about that, PJ! I come from a family of seven, so I was six before I realized "hey you" wasn't my actual name...

K

#29 ::: Jameson Quinn ::: (view all by) ::: May 26, 2015, 06:45 PM:

@14:

Does this mean that I have to think strategically how many works I'll nominate?

The short answer is, no.

The long answer is: there is a mathematical theorem that in any possible voting system, strategic voting is possible in at least some cases. But under this system, strategy would have to focus either on the selection phase or the elimination phase. But as a rule, a strategic advantage on just one of the two is not enough to win.

In theory, there could be exceptions to this rule; cases where strategy focusing on one of the phases could be enough to win, because without that strategy a given work was doing unusually poorly on the phase. However, in order for that to happen, several unlikely factors have to come together in an even-more-unlikely balance. Aside from being improbable, this is in practice impossible to predict. So without 20/20 hindsight, there's no way to plan a strategy that works.

#30 ::: JJ ::: (view all by) ::: May 26, 2015, 07:08 PM:

Jameson Quinn #29:

#14: Does this mean that I have to think strategically how many works I'll nominate?

The short answer is, no.
The long answer is: there is a mathematical theorem that in any possible voting system, strategic voting is possible in at least some cases. But under this system, strategy would have to focus either on the selection phase or the elimination phase. But as a rule, a strategic advantage on just one of the two is not enough to win.
In theory, there could be exceptions to this rule; cases where strategy focusing on one of the phases could be enough to win, because without that strategy a given work was doing unusually poorly on the phase. However, in order for that to happen, several unlikely factors have to come together in an even-more-unlikely balance. Aside from being improbable, this is in practice impossible to predict. So without 20/20 hindsight, there's no way to plan a strategy that works.

Bless your heart, Mr Quinn, because you've done a lot of wonderful work on this, but that explanation will be as clear as mud to the average Worldcon voter. I'm not able to articulate a plain English translation as the moment, but I will put some thought into how it would be worded.

#31 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 07:19 PM:

sfk@14:

Here was my first draft attempt...

20. So what kinds of strategy should I use when nominating? Is there a particular number of works I should nominate?
Under SDV-LPE there’s no reason to use any strategy at all. The system was crafted specifically to make most strategies irrelevant. For example, suppose you wanted to give your full support behind one work that you really feel strongly about. You can do that, particularly if there isn’t anything else you feel is Hugo-worthy. On the other hand, you generally won’t hurt your candidate by nominating something else that you feel is worthy. If your one work gets eliminated, then your full support will go towards the remaining work. In general, the best strategy is simple: nominate as many works as you feel are worthy.

#32 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 07:24 PM:

and since it's somewhat related, here was my edit of Jameson's post on phases...

K

21. Why are there two phases (Selection and Elimination)? Couldn’t you simplify it by doing it all in one step per elimination?
The selection phase looks for works with the fewest points, and the elimination phase eliminates the one with the fewest nominations. Looking at the two different measures of popularity adds robustness to the system. With non-slate nominations, as occurred with most of the Hugo Awards to date, the works with the fewest points will tend to be the same ones as those with the fewest number of nominations. But while it might be possible for strategic blocs to manipulate one of these two measurements (as occurred in 2015), it would be very difficult to manipulate both in the same direction. A strategy that focuses on nominations for the elimination phase (such as bloc voting) will be stymied by the selection phase; a strategy that focuses on points for the selection phase (such as nominating only one work) will turn out to be counterproductive in the elimination phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.

#33 ::: Jameson Quinn ::: (view all by) ::: May 26, 2015, 08:03 PM:

@30: Well, I was interrupted several times while writing it... so eventually I just hit "send" because I was tired of editing it. But yes, it definitely needs further work.

I guess something simpler like @31 is probably the best way to go. I was trying to recapitulate the point about robustness, but you're right, it did not particularly work. If anybody can make another attempt to say something similar, I'd be grateful.

#34 ::: junego ::: (view all by) ::: May 26, 2015, 08:25 PM:

I think this got lost as it was right at the end of the last thread. It isn't clear, to me, in the current language exactly what the first step in the process entails, so...

#1006 ::: junego ::: (view all by) ::: May 26, 2015, 01:51 PM:
@Kilo
[PROPOSAL]
3.8.8: The final Award ballots shall list in each category the eligible finalists as determined by successive rounds of a two-phase elimination process. In this process, each member gets a single nomination “point” for each category, and that point will be divided equally among their nominated works in that category.

I think we need more description of the process here. I'd suggest something like below to modify the final sentence above:

"In this process,...category; all the points from all ballots will be summed for each nominee; and the total number of ballots each nominee appears on will be tallied. These two numbers will be used in the process described below."

#35 ::: junego ::: (view all by) ::: May 26, 2015, 08:30 PM:

[3.8.1]

Small, small nit-pick: drop "following" from the last sentence. It's redundant and awkward.

#36 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 08:35 PM:

Jameson@33:

Your strength is in showing the relationships between the various aspects of the system. For example, I hadn't thought about the fact that slates need to succeed in two opposing directions in order to control the ballot. It totally makes sense, I just hadn't thought to make that comparison explicit. It's very well done.

But you've got the same issue I had when I was in grad school. I was studying general relativity with Charlie Misner, and was living every moment in tensor calculus and Riemannian geometry. My wife was also in grad school in astrophysics, so we could "sling the lingo" with each other without thinking about it. But when we tried to talk to grad students outside our field, we had a really hard time communicating, not because they were idiots, but because they just didn't have the background we had -- not even what we, by that time, considered to be the simple stuff. It wasn't until I started teaching undergraduates that I finally learned how to explain GR in truly nontechnical terms. Prior to that, I was just too close to it.

So, your insights are absolutely important. But you might keep in mind that most people -- even though they are smart -- don't even have the background of the easy stuff in your field. I'm a reasonably smart guy with a good background in statistics, but I had to go get a mini-education before I could follow even the basics of the deeper theory. But I don't think most people want the deeper theory. We need it to invent something sound, but it's more important that we have reasonable explanations for why things work rather than mathematical ones. We have a new assistant professor at work that I'm mentoring. I tell him what my first mentor told me, "Your explanations must be correct, but they shouldn't be more detailed than is necessary for the topic at hand. If you use an analogy, you have to explain its limitations. But be ready to provide a detailed proof for when you come across a student who is ready for it."

Just thoughts,
Kilo

#37 ::: Keith "Kilo" Watt ::: (view all by) ::: May 26, 2015, 08:42 PM:

All:

Incidentally, as suggestions are being made I'm updating the Word doc. Once we have a good stack of edits -- or if it seems the discussion is settling down -- I'll post the next draft. Some suggestions work as-is, some need some editing to fit in the whole, but my goal is for everyone's voice to be heard and contributed.

K

#38 ::: P J Evans ::: (view all by) ::: May 26, 2015, 08:55 PM:

36
I used to call it 'asking dumb questions, because then you get the best answers'.

I did QC for years, and learning what I needed to know required asking a lot of dumb questions. (And I learned more about piping than I ever expected to know.)

#39 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 01:48 AM:

All:

So, what is the group's feeling on the current 3.8.8.4 (which will be 3.8.8.3 in the next draft)? Should we give the admins discretion, or should we require another amendment at a later date to change tie breakers?

Kilo

#40 ::: junego ::: (view all by) ::: May 27, 2015, 02:24 AM:

Kilo (and everyone else)

[PROPOSAL]

I've just spent a couple of hours reformatting and wordsmithing the proposal language. As I was looking things over to suggest changes, I realized that it was easier for me to present all my suggestions as one complete entity rather than as a whole bunch of puzzle pieces.

I hope everyone doesn't think I'm doing this out of hubris, but this is one area where I have a certain amount of experience. I worked for 25+ years in manufacturing, testing and quality engineering for vendors/contractors to NASA, US DoD, aerospace industries/agencies in US, UK, France, Germany, etc, and US medical device industry. I have been on the receiving and giving end of good and bad rules/regs.

I'm applying the following general principles for this reformat:
1) Describe the basic process as clearly and simply as possible -first-!
2) Describe the ifs/ands/ors/buts equally clearly and simply -after- 1).
3) More but smaller steps are usually better than fewer but more convoluted steps.
4) Use/define terms consistently.
5)If revising an extant document, use existing conventions as much as possible.

It'll probably take me another hour or so to cut, paste and reformat what I've got. I didn't see a lot of new suggestions, but if someone had an important change that I missed, I apologize.

Even if we don't go with my particular suggestions, I recommend that our final proposal follow the 'principles', at least in general.

Kilo, I'm not dissing what you've done either. It's a whole lot easier to edit someone else than create from scratch (and with dozens of competing suggestions thrown at you, too.) :^]

#41 ::: Soon Lee ::: (view all by) ::: May 27, 2015, 02:25 AM:

[3.8.8.4]

If we are to explicitly allow admins discretion, I would prefer not to restrict them, so I would delete the following: "provided those measures are announced at the beginning of the nomination period for the Hugo Award."

We already trust the admins with a certain amount of discretion e.g. assigning works to the appropriate category in edge cases, or in deciphering intent where a nominator has been ambiguous. So on balance (given we expect this discretion to be used rarely), I say yes to giving admins discretion.

#42 ::: felice ::: (view all by) ::: May 27, 2015, 02:37 AM:

Keith "Kilo" Watt @39: "So, what is the group's feeling on the current 3.8.8.4 (which will be 3.8.8.3 in the next draft)? Should we give the admins discretion, or should we require another amendment at a later date to change tie breakers?"

If we want to give the admins discretion, we shouldn't specify what to do with ties at all. If we are specifying that ties should be eliminated, giving the admins discretion is a bad idea; they'd have no idea if they'd need to exercise it until after seeing the nomination data, and messing around with how they count it at that stage isn't going to look good. And suggesting that they might need discretion is implying that there's a flaw with the system, which won't help the proposal.

#43 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 02:44 AM:

Junego@40:

No worries at all! I'll be looking forward to seeing what you have. This is a group effort -- I've not got an ego to bruise here (if you'll pardon the paraphrase :) ).

Kilo

#44 ::: Steven desJardins ::: (view all by) ::: May 27, 2015, 02:51 AM:

I disagree with both felice and Soon Lee: giving the administrators a bit of leeway to tweak the procedure is no bad thing, but letting them decide how to break ties after they see what ties need to be broken is a downright terrible idea. That would just foster further paranoia among the (admittedly delusional) people who are already skeptical of the process.

I support giving them flexibility to add tiebreakers, with an item in the FAQ noting that simulated data has shown that including further tiebreakers is unlikely to make any difference to the final result. If we get the sense that the Business Meeting doesn't like the provision, we can drop it then.

#45 ::: spacefaringkitten ::: (view all by) ::: May 27, 2015, 02:56 AM:

I like Keith "Kilo" Watt's draft @31 a great deal.

My guess is that many people will only read the plain language version and skim the FAQ (because time is precious), so keeping things simple may be advisable.

I really appreciate the effort you people have put into this. Hope it'll make it in the business meeting!

#46 ::: Soon Lee ::: (view all by) ::: May 27, 2015, 03:27 AM:

Steven desJardins #44:

I don't have strong feelings about this, but your comment is convincing. We should to keep it as simple & unambiguous as possible.

#47 ::: junego ::: (view all by) ::: May 27, 2015, 03:44 AM:

[PROPOSAL REFORMAT]

Sorry, this is a long post and my threatened reformat of the proposal.

Original proposal is in italics.
My reformats are in [bold brackets].
I have put a few explanatory comments in {plain text and sqiggle brackets}

I've only wholesale redone section 3.8.8.


***********************************************************************

Short Title: E Pluribus Hugo (Out of the Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations)...

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the following multi-round process described in 3.8.8.

[ remove the word "following" from the last sentence. It's redundent.]

{I snipped all the sections where there are no proposed changes}

3.8.8: The final Award ballots shall list in each category the eligible finalists as determined by successive rounds of a two-phase elimination process. In this process, each member gets a single nomination “point” for each category, and that point will be divided equally among their nominated works in that category.

[3.8.8: The three-phase finalist selection process: Initial steps: First, the total number of nominations shall be tallied for each nominee in a category. Next, a single point shall be assigned to each ballot for a category and divided equally among all nominations in that category. Last, all these points from all ballots shall be totaled for each nominee in a category. These two numbers, nominations and points, shall be used to determine the finalists in a category from successive elimination rounds as described below.]

{The first sentence was unnecessary, restating 3.8.1. The process description was incomplete. I'm not completely happy with the way I describe the steps as "first", "next", "last" but brain was too tired to find better. }

3.8.8.1: Selection Phase: In each round, all the works having the least number of points will be selected for the Elimination Phase (3.8.8.2). If there is only one work with the least number of points, then all of the works with the second-least number of points will be also be selected for the Elimination Phase in addition to the lowest-point work.

[3.8.8.1: Selection Phase: Each round shall start by selecting the two (or more, as described in 3.8.8.4) nominees with the lowest points for the Elimination Phase (3.8.8.2).]

3.8.8.2: Elimination Phase: Of the works identified in the Selection Phase, the one(s) that appear on the fewest number of nomination ballots will be removed from all nomination ballots for subsequent rounds as if they had never appeared on any ballots.

[3.8.8.2: Elimination Phase: Nominees chosen in the Selection Phase (3.8.8.1) shall be compared and the nominee that has the least nominations shall be eliminated and removed from all ballots for the Recalculation Phase (3.8.8.4) of all subsequent rounds. (See 3.8.8.4 for ties)]

3.8.8.6: Subsequent rounds begin by reallocating points as follows: All works eliminated in previous rounds are removed from all nomination ballots and treated as if they had never appeared on any nomination ballot. Members’ points are then reallocated equally among their remaining nominated works, if any.

[3.8.8.3: Recalculation Phase: After the Elimination Phase (3.8.8.2), those ballots that contained the eliminated nominee(s) shall have their one point redistributed among their remaining nominees (if any) in that category and all points shall be recalculated for all nominees. The next round shall start at 3.8.8.1. The rounds shall continue until the defined minimum number of finalists remain (except as modified by 3.8.8.5)]

{I reodered the original 3.8.8.6 in accord with principle of explaining the basic process first and added a new "phase" because it actually IS the next step after each elimination and it makes it clearer to understand and explain, imo, to have this be a separate named thing.}

3.8.8.4: Ties: If two or more works are tied for appearing on the fewest number of nomination ballots, the tied work with the lowest point total will be eliminated. If there is a tie for appearing on the fewest number of nomination ballots as well as for lowest point total, then all members of that tie will be eliminated. Should they deem it necessary to do so in the future, the Hugo administrators are empowered to take further measures to break this type of tie, provided those measures are announced at the beginning of the nomination period for the Hugo Award.

[3.8.8.4: Ties:
3.8.8.4.1: If, at the Selection Phase, two or more nominees are tied for the lowest points, all such nominees shall be moved to the Elimination Phase.
3.8.8.4.2: If, at the Selection Phase, one nominee has the lowest points and two or more nominees are tied for the second lowest points, all such nominees shall be moved to the Elimination Phase.
3.8.8.4.3: If, at the Elimination Phase, two or more nominees are tied for the fewest number of nominations, the nominee with the lowest points at that round shall be eliminated and the Recalculation Phase shall be implemented before starting a new round.
3.8.8.4.4: If, at the Elimination Phase, two or more nominees are tied for both nominations and points, then all such nominees tied at that round shall be eliminated and the Recalculation Phase shall be implemented before starting a new round.]

[I'VE LEFT OFF THE ADMIN DISCRETION LANGUAGE, IT CAN BE ADDED IF NECESSARY AFTER CONSENSUS IS REACHED]

{I broke out the different types and rules for ties into separate sub-headings to, again, make them easier to understand and explain.}

3.8.8.5: If elimination would reduce the number of finalists to fewer than the defined minimum number, then instead none of the works from that round shall be eliminated, and all remaining works shall appear on the final ballot, extending it if necessary.

[3.8.8.5: If any elimination round would reduce the number of finalists to fewer than the defined minimum number, then none of the nominees in that round shall be eliminated and all remaining nominees shall appear on the final ballot, extending the maximum number of finalists, if necessary.]


Section 3.9: Notification and Acceptance.

{I have some thoughts and questions about this section, but they can wait until tomorrow.}

#48 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 07:34 AM:

20. So what kinds of strategy should I use when nominating? Is there a particular number of works I should nominate?

Under SDV-LPE there’s no reason to use any strategy at all. The system was crafted specifically to make most strategies irrelevant. For example, suppose you wanted to give your full support behind one work that you really feel strongly about. You can do that, particularly if there isn’t anything else you feel is Hugo-worthy. On the other hand, you generally won’t hurt your candidate by nominating something else that you feel is worthy. If your one work gets eliminated, then your full support will go towards the remaining work. In general, the best strategy is simple: nominate as many works as you feel are worthy. The more works the average voter nominates, the better the final result can reflect the will of the majority.

Voting theorists know that it's impossible to make a system that's 100% strategy-proof. But SDV-LPE comes pretty close. A strategy that dishonestly helps your favorite works in the selection phase will tend to hurt them in the elimination phase, and vice versa. These factors usually balance out exactly (over 99% of the time, in our simulations); and even when they don't, it's impossible to predict which way the balance will swing unless you have perfect knowledge of how everyone will vote. So before attempting a coordinated voting strategy, you'd have to ask yourself the Dirty Harry question: are you feeling lucky, punk?

#49 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 07:39 AM:

Sorry, the above @48 should have had [FAQ].

....

@47 [junego's edits]: I like this revision in general, but I still think we should avoid talking about "nominations". If we're giving precise meanings to words, they should be words that are conceptually related, but where people don't have an intuitive definition that's not the one we want. So I'd guess "approvals" is best; or something else, but certainly not "nominations".

#50 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 08:00 AM:

[FAQ]

@48: When I say it's "impossible" to create a strategy-proof voting system, I think it needs to be strengthened; this isn't "we've found it impossible", but "we are certain that it will forever be impossible". So how about "flat-out impossible" or "literally impossible"?

#51 ::: junego ::: (view all by) ::: May 27, 2015, 08:00 AM:

Jameson @ 49

@47 [junego's edits]: I like this revision in general, but I still think we should avoid talking about "nominations". If we're giving precise meanings to words, they should be words that are conceptually related, but where people don't have an intuitive definition that's not the one we want. So I'd guess "approvals" is best; or something else, but certainly not "nominations".

I was using "tallies" or "ballots" for a while, but settled on "nominations" because the constitution currently uses the term to mean the count of votes/approvals/whatevers that are equal to the number of times a nominee is listed on member ballots. If the word is already understood to mean this, it's just easier to use the same word than define a new one. Less confusion, less resistance.

#52 ::: J Thomas ::: (view all by) ::: May 27, 2015, 08:11 AM:

@14 spacefaringkitten

"Does this mean that I have to think strategically how many works I'll nominate?"

Short answer: No, don't bother. You do just fine by nominating whatever you think deserves to be a finalist. It doesn't help to think strategically.

Long answer:

When you nominate a particular candidate, you increase the chance it will be a finalist. You never hurt its chances by nominating it.

In any system, nominating one candidate is likely to slightly reduce the odds that another candidate will win, because there are many candidates competing for a few finalist places. With this way of voting, when nominee Alpha gets more votes than candidate Beta, then when you nominate both, they both get slightly less support but candidate Beta loses more than candidate Alpha does.

This matters when candidate Beta is on the edge of fifth place. When candidate Beta will lose anyway then you don't hurt it by nominating Alpha too, and when candidate Beta will win anyway you don't hurt it by nominating Alpha too.

When you care so much more about Beta that you would feel bad if Alpha won and Beta didn't, vote for only Beta. It means you don't like Alpha that much.

And if you are sure that Alpha is ahead of Beta, and Beta is on the edge of fifth place, then vote only for Beta. Are you sure you are right that Alpha is ahead of Beta and that Beta is on the edge of fifth place? If not, just nominate the ones you think should be finalists. It isn't worth any strategy unless you know who's in fifth place.

#53 ::: junego ::: (view all by) ::: May 27, 2015, 08:23 AM:

felice @42

If we want to give the admins discretion, we shouldn't specify what to do with ties at all. If we are specifying that ties should be eliminated, giving the admins discretion is a bad idea; they'd have no idea if they'd need to exercise it until after seeing the nomination data, and messing around with how they count it at that stage isn't going to look good. And suggesting that they might need discretion is implying that there's a flaw with the system, which won't help the proposal.

Excellent points, felice, I agree totally.

I initially thought it would be a good idea to give them flexibility, but when I tried to walk through the scenarios where they might actually use it, they almost all came back to "they can't change the system once they start tallying the nominations"! So the discretion is essentially useless. Putting it in is adding a needless point for argument with absolutely no reasonable gain for the proposal or the Hugos.

If you think the admins need or would use this flexibility, please describe the circumstances under which it would work. Maybe we're wrong and there is some great good to adding this proviso.

#54 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 08:43 AM:

[discretion]

I said in @5 that I'm willing to drop this idea, and it looks as if consensus is forming that we don't need it. So, I think wordsmithing is all we have left.

#55 ::: junego ::: (view all by) ::: May 27, 2015, 08:57 AM:

Keith "Kilo" Watt @17

Once we have the formal proposal finished, we should discuss how we want to present it. I don't know if we have PowerPoint available or just talk, etc. But we can cross that bridge when we come to it.

Speaking of presentation materials, wasn't it JT who had started an animation that visually explained the tally process? I thought it was a really good first attempt and having the coded animation for a possible seminar (I think a seminar *could* be a good idea) and the meeting itself could be really helpful in selling the proposal. If actual animation wasn't available then a cartoon with similar printed images could be used to good effect.

HEY! JT, DO YOU STILL HAVE THAT ANIMATION?

#56 ::: junego ::: (view all by) ::: May 27, 2015, 09:17 AM:

Keith "Kilo" Watt @32
21. Why are there two phases (Selection and Elimination)? Couldn’t you simplify it by doing it all in one step per elimination?
The selection phase looks for works with the fewest points, and the elimination phase eliminates the one with the fewest nominations. Looking at the two different measures of popularity adds robustness to the system. With non-slate nominations, as occurred with most of the Hugo Awards to date, the works with the fewest points will tend to be the same ones as those with the fewest number of nominations. But while it might be possible for strategic blocs to manipulate one of these two measurements (as occurred in 2015), it would be very difficult to manipulate both in the same direction. A strategy that focuses on nominations for the elimination phase (such as bloc voting) will be stymied by the selection phase; a strategy that focuses on points for the selection phase (such as nominating only one work) will turn out to be counterproductive in the elimination phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.

I think we should keep my new formulation of a 3 phase system not only because it makes it easier to explain, but I think that Recalculation is a really important step. That's where the points that are so vital as the second measure of popularity are actually redistributed to the remaining nominees. I'm not going to try to reword this section until and if we agree on reconfiguring this as 3-phase. (But I do like your description of the process in general)

#57 ::: junego ::: (view all by) ::: May 27, 2015, 09:23 AM:

Jameson Quinn @54
[discretion]

I said in @5 that I'm willing to drop this idea, and it looks as if consensus is forming that we don't need it. So, I think wordsmithing is all we have left.

This isn't directed at you, JQ. We're working on a consensus with the other half of the support for the idea. :^]

#58 ::: Duncan J Macdonald ::: (view all by) ::: May 27, 2015, 09:30 AM:

junego @ 51: I was using "tallies" or "ballots" for a while, but settled on "nominations" because the constitution currently uses the term to mean the count of votes/approvals/whatevers that are equal to the number of times a nominee is listed on member ballots. If the word is already understood to mean this, it's just easier to use the same word than define a new one. Less confusion, less resistance.

I concur. The membership is already familiar with "nominating" works, and we've said that there is no change in the nominating process (FAQ #1). There already exists a proposal to change the terminology throughout the constitution, we really shouldn't add to the confusion.

Secondly, I also agree with not adding an explicit "Admins have the following discretion" section.

#59 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 09:33 AM:

@58: We all agree we don't want to add to the confusion. The question is, how best to avoid that.

#60 ::: Steven desJardins ::: (view all by) ::: May 27, 2015, 10:02 AM:

The advantage I see in letting the administrators add tie-breakers is that it allows WSFS to tweak the counting process without spending 2-3 years amending the Constitution. I envision the administrators looking at the first year's results with real data, saying, "Hmm, we could use another tie-breaker, here's something that makes sense," and suggesting it to the following year's committee. I think it's a minor improvement over *not* allowing tweaks, worth putting in the proposal but not worth fighting over.

#61 ::: junego ::: (view all by) ::: May 27, 2015, 10:34 AM:

Jameson Quinn @ 7 plus later posts

If you're a Sasquan organizer, I'd be happy to have a short session where I and other panelists (such as Keith, of course, if he's willing) discuss this proposal, including the voting theory, statistics, and simulations that went into making it.

It might be a more interesting panel and garner a larger audience if it included people proposing all the different voting reforms, like "4/6". Everyone would have a chance to explain their ideas. People could ask questions and compare proposals. The organizers might agree this would be interesting. It also might make Kilo"s participation less decisive.

Kevin Standlee said:
"Given that the first time this proposal will come before the meeting will be shortly after 10 AM Thursday (the second day of the convention), I suggest that if you're offered the opportunity to do a panel on this, you ask for sometime on Wednesday, the first day of the convention, and that you ask that if Sasquan schedules a panel about "Introduction to the Business Meeting" (a relatively common-but-lightly-attended panel on Day 1), they not schedule it against your discussion."

"Alternatively, if you are confident that the Preliminary Business Meeting won't kill your proposal out of hand, you might ask for them to schedule such a panel sometime after 1 PM Thursday, so people who first encounter the proposal at the PBM can go to an informal "hearing" on the subject."

I'm not sure how to gain confidence of passing the PBM. Obviously, we need to find out who and how to do this.

If we do think we can jump the PBM hurdle, I think Kevin's suggestion of after 1pm on Thurs sounds good.

#62 ::: Tim Illingworth ::: (view all by) ::: May 27, 2015, 10:52 AM:

The PBM is a business meeting like any other. It can kill a proposal by Postpone Indefinitely, which takes a 2/3 majority (or by Object To Consideration, which takes 3/4 these days).

#63 ::: Cally Soukup ::: (view all by) ::: May 27, 2015, 11:12 AM:

junego @47

[REWRITE]

As seems to be my role in this process, I'm mostly about the copyediting. In 3.8.8.2 of your rewrite, you have "least nominations" where you should have "fewest", the number of ballots involved being a countable thing.

And I remember Kevin(?) saying we should avoid the word "nominations". Unfortunately, the only rephrase I can think of is "ballots on which a candidate appears", which is really clunky.

#64 ::: Duncan J Macdonald ::: (view all by) ::: May 27, 2015, 11:41 AM:

On the "nomination" question:

There is an amendment (A.3) pending before Sasquan passed on by Loncon3 to modify the language of Sections 3.7, 3.9, 3.10, and 3.11 to "...change
references to those works or people that appear on the final Hugo Award ballot to “finalist” and to change references to “candidate” to “nominee,” ..."

From that amendment, I feel that our using the term nomination and/or nominee for portions of Section 3.8 fits well with the flow of the process: from nominations to nominees to finalists to winner (or Noah Ward).

#65 ::: David Harmon ::: (view all by) ::: May 27, 2015, 11:56 AM:

junego #61:

It might be a more interesting panel and garner a larger audience if it included people proposing all the different voting reforms, like "4/6". Everyone would have a chance to explain their ideas. People could ask questions and compare proposals.

Oh, you've got to be kidding me. We're finally finishing up a proposal after 3,000+ messages in a moderated forum, 50 days clock time, and a fair bit of programming and simulation in the background. Do you really want to try to replay the entire process in an hour or two of meatspace interaction, the day before the proposal gets presented? If there is a panel, it badly needs to stay focused on the proposal as made, with discussion of the "paths not taken" limited to "why we didn't pick that".

#66 ::: Cadbury Moose ::: (view all by) ::: May 27, 2015, 12:09 PM:

David Harmon @ #65

It'd be interesting all right, as in "interesting.cn".

If there's going to be a panel then it needs to be more of a presentation (possibly with handouts) of the finished proposal and very limited "questions from the floor".

Not to mention a seriously good moderator in case of likely disruptive attempts from the canine cabal.

#67 ::: Seth Gordon ::: (view all by) ::: May 27, 2015, 12:13 PM:

@Jameson, is the following an accurate answer to the FAQ regarding strategic voting?

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. And suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all should submit ballots that only have Castle in the Best Novel category, leaving off Rocket: your ballots, and your undivided points, would allow Castle to displace Tentacles.

In practice: You won’t actually know the preferences of all the other Hugo voters to such a high level of precision before the election. Maybe Rocket isn’t so popular after all and needs your votes to get over the top. Maybe Tentacles has a broad enough base of support that your votes for Castle aren’t enough to displace it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance.

So honesty is your best strategy.

#68 ::: Aan ::: (view all by) ::: May 27, 2015, 12:18 PM:

Steven @ #60: The advantage I see in letting the administrators add tie-breakers is that it allows WSFS to tweak the counting process without spending 2-3 years amending the Constitution.

It taking 2-3 years to amend the constitution is a feature, not a weakness. If we want to give the admins powers to make decisions for the good of us all when something not entirely to our liking comes up, then there's many other sections where it'd do a whole lot more "good" than just here in this tiny and extremely unlikely-to-actually-matter tie-breaking section.
I'm quite in favour of them not having that power in other sections, so think we also shouldn't explicitly give it to them here.

#69 ::: P J Evans ::: (view all by) ::: May 27, 2015, 12:55 PM:

65
By my count, about 3.4MB of discussion, so far.

#70 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 12:58 PM:

@67 Seth: Here's my edit for accuracy:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. And suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all should submit ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points. If you push it up to 5th place, it will still be eliminated by Tentacles; but if you push it all the way up to 4th place, Tentacles will be eliminated by some other work (whatever would have been in 4th place with honest ballots).

In practice: You won’t actually know the preferences of all the other Hugo voters to such a high level of precision before the election. Maybe Rocket isn’t so popular after all and needs your votes to get over the top. Maybe Tentacles has a broad enough base of support that your votes for Castle aren’t enough to displace it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy doesn't change Castle's score by enough to put it in 4th place. There are many ways for your strategy to fail, or even backfire; and only one, highly-specific way for it to work.

So honesty is your best strategy.

....

I realize that's not as clear as what you wrote, but unfortunately what you wrote is not quite true.

#71 ::: Seth ::: (view all by) ::: May 27, 2015, 01:01 PM:

3.8.8.5 contains "minimum number". It needs to be specified (it appears to be 5, not 3).

3.9.1 contains "maximum number". It needs to be specified (it appears to be 5).

3.8.5 is currently "No nominee shall appear on the final Award ballot if it received fewer nominations than five percent (5%) of the number of ballots listing one or more nominations in that category, except that the first three eligible nominees, including any ties, shall always be listed."

3.8.5 would remove, *after* reduction to 5 is done, all candidates that don't appear on at least 5% of the ballots "except the first three". Which are the "first three"? This clause does not play well with the new procedure, and needs to be modified. I see choices including

3.8.5.A Prior to beginning the elimination procedure, all candidates that do not appear on at least 5% of the number of ballots listing one or more candidates in that category, except that if this would reduce the number of candidates below X, then the Y candidates appearing on the most ballots (or more if there are ties for Y'th place) will not be eliminated. Fill in the blanks for X and Y.

3.8.5.B After the elimination procedure, all candidates that do not appear on at least 5% of the number of ballots listing one or more candidates in that category will be eliminated, except that if the procedure would reduce the number of candidates below 3, then the 3 selected candidates appearing on the most ballots (or more, if there are ties for 3rd) will remain.

3.8.5.C (Just remove that clause)

#72 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 01:04 PM:

@65: The fact is that 4/6 is not just "another option", but an actual proposal that will be before the BM. I agree we shouldn't be discussing "what if we tried..." hypotheticals, but I think a panel discussing all the actual proposals that exist is not out of place. And yes, we can point our multi-megabyte discussion archive and github projects and simulation output and all that, and I think people will understand that we've gone into greater depth than the 4/6 people, but I don't think it's necessary to shut them out.

#73 ::: Seth ::: (view all by) ::: May 27, 2015, 01:04 PM:

Please add me (Seth Breidbart) as a co-signer.

#74 ::: Mike Kennedy ::: (view all by) ::: May 27, 2015, 01:08 PM:

Computational Accuracy

I'm a "visitor" to this thread, having seen notice of it on the SMOFS mailing list. I'd like to say I'll check in again, but the demands of life make that unlikely. I do have one issue -- which may well have already been considered -- that I'd like to bring up.

For purposes of this example, presume that computations are performed to 2 decimal digit accuracy. The actual accuracy of computation will doubtless be higher than that, but there will be some finite accuracy. My question is whether the following (general) case is taken into account by the proposed system.

Suppose that 3 nominators each nominate 3 works. One of the works, call it Alice, is in common between the 3 nominators. Thus Alice receives 0.33 points from each nominator, for a total of 0.99 points.

Suppose further that 2 nominators each nominate 2 works, with Billy being in common between them. Thus Billy receives 1.00 points (0.50 + 0.50).

So, does Alice or Billy advance? As I understand the system's intent, it would be Alice; but computationally it will be Billy.

Many other scenarios could also be posited where finite accuracy calculation defeats the system on pathological cases.

#75 ::: P J Evans ::: (view all by) ::: May 27, 2015, 01:08 PM:

72
It's a separate proposal. Putting them in the same panel will confuse people. (People don't need big explanations of 4/6: it's very simple.)

#76 ::: P J Evans ::: (view all by) ::: May 27, 2015, 01:10 PM:

74
As it works in practice, it's not going to be that much of a problem. (The simulations were run with actual nomination data.)

#77 ::: Cally Soukup ::: (view all by) ::: May 27, 2015, 01:17 PM:

Seth: the reason not to put minimums and maximums in this proposal is that they are addressed in another part of the Hugo Rules, and may well (this year, even) have the process started to change them. There's no need to hard-code them into this proposal, and good reason not to.

#78 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 01:19 PM:

In FAQ 12, I believe we use "Slate" in the first sentence ("size of the slate") to mean "the number of people nominating the slate"; which is not how we use the word anywhere else. So we should spell it out, or rephrase somehow.

Then, in the second sentence of the answer, when we say it doesn't force any other work off the ballot--huh? Often it will. It's just that, given that there are enough people voting for the one-item slate, it's fair that that work be a finalist, and for that to happen some other work may get eliminated from being a finalist.

#79 ::: Seth Gordon ::: (view all by) ::: May 27, 2015, 01:20 PM:

Jameson @70: Corrections gratefully accepted.

Mike @74: Computation of the point values does not need to use floating-point arithmetic. Modern programming languages allow you to do math on fractions without converting them to floating-point approximations; in the Python language, for example, code for doing this is part of the standard library. Or, since the fractions involved will all be multiples of 1/60, the code could simply do all of its computation using sixtieths-of-a-point.

#80 ::: David Harmon ::: (view all by) ::: May 27, 2015, 01:21 PM:

Seth Gordon #67: I like it. Though it's still worth saying explicitly, "no general strategy will help or hurt much".

#81 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 01:22 PM:

I'm not a strong authority on this, so we should ask people with loads of Business Meeting experience. But from reading about it online for years, I'm doubtful that the Admin discretion idea would play well politically at the Business Meeting. (Note that this is independent of whether it's a good idea in general! And therefore might not be decisive even if there's a clear consensus of those in the know. But it worries me enough I wan t to raise it.)

#82 ::: junego ::: (view all by) ::: May 27, 2015, 01:22 PM:

David Harmon @ 65
junego #61: said:
"It might be a more interesting panel and garner a larger audience if it included people proposing all the different voting reforms, like "4/6". Everyone would have a chance to explain their ideas. People could ask questions and compare proposals."

Oh, you've got to be kidding me. We're finally finishing up a proposal after 3,000+ messages in a moderated forum, 50 days clock time, and a fair bit of programming and simulation in the background. Do you really want to try to replay the entire process in an hour or two of meatspace interaction, the day before the proposal gets presented? If there is a panel, it badly needs to stay focused on the proposal as made, with discussion of the "paths not taken" limited to "why we didn't pick that".

We know the 4/6 proposal is already on the agenda and, from seeing peoples' statements around the interwebs, there is a certain level of support for that proposal. I haven't heard of any other serious proposals, but there might be others heading for the BM as we speak.

Our proposal will work with 4/6, although our boffins think it will have a slight negative impact on E pluribus' effectiveness. It might be in our interests to be at least friendly to them (and others). We all want the same thing, to insulate the Hugos from slates. If we can get those who already favor other proposals and some of the general membership to understand how robust our system is and how it reduces slate influence, we may be able to get more support for our proposal.

Those are the thoughts I had when I suggested this. It definitely has potential to become a free-for-all, so caution and planning would be required. If everyone thinks the opportunity to educate and persuade isn't worth the risk, I guess it won't go anywhere.

#83 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 01:34 PM:

3.8.8.1 Selection phase: This is weird. Discussion of why the requirement of eliminating more than one candidate each round is necessary doesn't belong in this thread (was there such discussion? Can somebody point me to it?), but assuming there's a good reason, I think there needs to be a FAQ on this, because it's an obvious weirdness, so people will wonder about it.

3.8.8.4 Ties: The complexity of this, plus the "admin discretion" proposal, worries me. It ends up sounding like we're not sure if we have an adequate tie-breaker in place.

#84 ::: junego ::: (view all by) ::: May 27, 2015, 01:40 PM:

#63 ::: Cally Soukup ::: (view all by) ::: May 27, 2015, 11:12 AM:
junego @47

[REWRITE]

As seems to be my role in this process, I'm mostly about the copyediting. In 3.8.8.2 of your rewrite, you have "least nominations" where you should have "fewest", the number of ballots involved being a countable thing.

Do *not* disparage copy editing, most of our books would be nearly unreadable without it! 😋

I will note the correction, but Kilo is the one keeping track, I hope. And thank you, this really is a necessary job.

And I remember Kevin(?) saying we should avoid the word "nominations". Unfortunately, the only rephrase I can think of is "ballots on which a candidate appears", which is really clunky.

I don't remember that at all. Jameson Quinn has objected to it, but "nominations" is the word used in the current constitution and, afaik, there is no proposal to change it. If you recall/find a reference from Kevin, would appreciate a heads up.

#85 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 01:46 PM:

@84: Given @64, I have no further problem with "nominations".

#86 ::: Seth ::: (view all by) ::: May 27, 2015, 01:48 PM:

#77 Defining "minimum" and "maximum" by reference to other articles is fine. As it is, the numbers are ambiguous. (Is the minimum 3 or 5?)

#83 We don't eliminate more than one candidate each round (unless there are ties). We select more than one for possible elimination, then choose among those for actual elimination.

#87 ::: Cally Soukup ::: (view all by) ::: May 27, 2015, 02:01 PM:

As for the word "nominee":

Kevin Standlee wrote HERE about a pending WSFS constitutional amendment that changes "nominee" to "finalist" and deprecates the use of "nominee" as a reference to "person or work appearing on the final ballot. He concludes:

Whatever you propose here, I hope it will use "nominee" to mean "work/person nominated in the initial voting" and "finalist" to mean "work/person appearing on the final ballot."

I was (and may well be again) somewhat confused about which was a "nominee" and which a "finalist", but I suspect I wasn't alone.

And now I've typed "nominee" too many times and it just looks like a jumble of letters....

#88 ::: Joshua Kronengold ::: (view all by) ::: May 27, 2015, 02:05 PM:

I don't think a panel will be amis.

I don't see any sign at the moment of this proposal being killed in the perliminary BM.

Re ties: There are lots of things left to admin discression, but it's possible some people would object to a "last ditch" tiebreak being so left explicitly (rather than implicitly). I don't think there's a good way to word this that would allow the ghosts or conjoined ties ideas.

I'm not happy with the OP tiebreak rules -- if we specify something, it's as specified. So if we want to leave some thing up to admin ruling/custom/discression, we want something more like:

3.8.8.4: Ties: If two or more works are tied for appearing on the fewest number of nomination ballots, the tied work with the lowest point total will be eliminated. If there is a tie for appearing on the fewest number of nomination ballots as well as for lowest point total, then that tie should be broken, if possible, by another manner -- if not possible, all tied works should be eliminated.

#89 ::: Ian GIllespie ::: (view all by) ::: May 27, 2015, 02:07 PM:

"nothing changes in how you nominate"

"the final Hugo voting system, which actually chooses the winner, is unchanged"

This seems like a bug, not feature, and needs to be better explained in the FAQ.

No new input is being solicited from voters? This is simply a mathematical counting mechanism?

Best I can tell, this is a two phase system in which Single Divisable Vote is used in the first phase of each round of voting, then the bottom two vote getters in a given round (or all those tied for last place) go "head-to-head" in a straight up vote count to see which gets eliminated. This process repeats itself, round after around, until all but five canadidates are eliminated (with some exceptions for ties and edge cases).

If my summation is correct, then the following statement fromt he FAQ is false:

"there’s no reason not to nominate something you think even might be Hugo-worthy"

If I have only one work that I want to see nominated above all else, I should vote for only that work to reduce the chances it will be selected for the elimination phase in any given round of voting.

Also, I don't see what this system does to discourage tacctical other than devalue the votes of all those who vote for more than one work. For example, if I vote for one popular work that is a contender to become a finalist and four works for which no else voted, then the value of my vote is reduced in the "selection" phase by the same amount as a slate voter who selects five works based on a tactical voting scheme. Conversely, while works appearing on a slate will be at a disadvantage in the "selection", they will benefit from the full impact of tactical voting in the elimination phase. This seems likely to result in those works being selected for elimination over and over again and truncing whichever works they are up against in the elimination phased due to the EXACT SAME ISSUE that caused the problem this year (an orgnaized, tactical focus on a few specific works vs. a disorganized open list).

Perhaps I am misunderstanding how this whole system works, but, if so, I'd say the FAQ needs clarifcation.

#90 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 02:09 PM:

Morning all!

Quite a bit of catching up to do this morning, so I'm going through all the posts. I've extracted junego's re-write into a separate file and am looking at merging the two versions now.

[3-PHASE]
One thing about junego's version that instantly jumps out at me is that maybe the third phase should be the "calculation phase" at the beginning and not the "recalculation phase" at the end. As it is, there are actually four phases mentioned in his re-write, though one of them is only done once. I think the sequence is more clearly shown as:

1. Calculation (or I would call this "distribution" maybe)
2. Selection
3. Elimination

You then loop through 1,2,3. This may seem more reasonable to me just because that's how the code is written, I don't know.


[PANEL]
I'm starting to feel strongly that this is not a good idea. If we can't explain SDV-LPE in 20 minutes, there's a problem. And I don't think there is a problem. It's really a very simple system. Yes, the theoretical underpinnings and proofs can be quite complex, but while those were necesary to make sure we had a good and robust system, they are -not- needed to explain the system or why it's a good idea. And as someone else pointed out, if you've ever been to a panel, the potential for it to get out of control is real. This would leave a bad taste for everyone regarding the system. I just see lots of potential for bad and virtually no potential good. I would much rather that we focus on a tight presentation that can be convincingly done in 20 minutes than to have an hour-long panel that might not accomplish anything.


More soon...

Kilo

#91 ::: Jeremy Leader ::: (view all by) ::: May 27, 2015, 02:12 PM:

Ian Gillespie @89: I think you're overlooking the redistribution of points. If you nominate two works, A and B, and then B gets eliminated in an early round, from that point on, A is treated as if you had never nominated B.

#92 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 02:13 PM:

[DISCRETION]

I'm not seeing a lot of support for admin discretion at this point. If there's anyone who still strongly feels it needs to be in, go ahead and pipe up, otherwise we'll move on without it.

K

#93 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 02:17 PM:

Seth@71: The sections you're referring to are in the constitution now. We aren't changing them with this proposal, but there are other proposals (sponsored by other people) who are looking to change these. Kevin Standlee (the Business Meeting chair) has asked us to keep the different proposals as separated as possible, so that's what's we've done. If one of those other proposals pass, then they will modify the sections you've quoted. We needed to write our proposal to be neutral with regard to those wording changes.

That said, there's probably no need to quote them in our proposal, just to avoid confusion.

Kilo

#94 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 02:20 PM:

Mike@74:

Yeah, when it comes time to code the actual web-based app, or whatever the Hugo admins decide to use, they can address the computational issues then. As someone else mentioned, it never came up as an issue in our sims, though I can see where you'd still want to plan for it in the code. Our sims were more to make sure what we thought would happen actually does happen when run with real data. Web and database programming isn't my thing, so I'll leave that to the experts. I plan to offer my help coding the algorithm itself, of course.

Kilo

#95 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 02:22 PM:

So, while we're nitpicking: I think that a tie for 5th place should not be broken, even by points. As I read the current text, that is not how it would work. That kind of tie is not actually that unlikely; I'd bet it would happen in some category in the first or second year of voting. I know, more special cases is annoying, but this one is probably worth it.

@89: You're right, "there’s no reason not to nominate something you think even might be Hugo-worthy" is false. We should say "there’s no good reason...", which is true, as explained in @70. So I guess that's an [FAQ CORRECTION].

As for your paragraph beginning with "also...": yes, you are not understanding this system. If you vote for four works which nobody else nominated, they will quickly be eliminated, so your vote will count full-strength for your viable choice. And you're right that slate candidates will tend to win in the elimination phase... but only until two slate works face each other, in which case (at least) one will have to be eliminated. (And the stuff that gets eliminated before that happens probably had no chance anyway.)

We've spent over 3000 comments working on this, and we really wouldn't have made a mistake that basic.

#96 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 02:24 PM:

David@78:

I thought about that, too. I had already re-written it as follows (the change is that they don't force -all- works off the ballot, which is what we really mean). As far as "large" I don't think we need to define it, becuase how large they need to be depends on how many total nominators there are and how those nominations are distributed.

12. Couldn't slates just recommend a single work for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate is large enough that is certainly a viable possibility – it’s also completely fair. It does not force all other works off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn't a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a work deserves a Hugo nomination, then it should, in fact, be represented on the final ballot.

#97 ::: junego ::: (view all by) ::: May 27, 2015, 02:31 PM:

David Dyer-Bennet @83
3.8.8.1 Selection phase: This is weird. Discussion of why the requirement of eliminating more than one candidate each round is necessary doesn't belong in this thread (was there such discussion? Can somebody point me to it?), but assuming there's a good reason, I think there needs to be a FAQ on this, because it's an obvious weirdness, so people will wonder about it.
3.8.8.4 Ties: The complexity of this, plus the "admin discretion" proposal, worries me. It ends up sounding like we're not sure if we have an adequate tie-breaker in place.

Not sure what you mean about 3.8.8.1 so I'll answer the questions I *think* you're asking. If I'm wrong, let me know.

If your question/concern is that we shouldn't complicate the explanation at this point by bringing up this issue, I agree and had already planned to change the language here to "(See 3.8.8.4 for ties)".

If your question/concern is why would we ever pick more than 2 nominees for the Elimination Phase, that's more complicated and involves how and why we break ties. There was, in fact, probably 500+ comments in the last thread discussing ties. Which segues into your question about 3.8.8.4: Ties.

Short answer is that we must have this level of detail in the proposal because this is the *law* we're trying to pass about how to decide who are the finalists. One example of why is that It has to be clear and detailed enough for someone to program a computer to do the calculations correctly.

In the new system, since we are eliminating from the bottom up and reapportioning points based on those eliminations, the issue of ties in either points, nominations or both have to be addressed.

An example: We select and compare the nominees with the lowest points, right? What do we do if there is one nominee with the fewest points and 2 nominees tied for 2nd fewest points? How do you decide which one(s) to eliminate, in what order? Thus the detailed section on ties. It may look very complicated at first, but it really isn't.

I agree there should be explanation in the FAQ what this is about.

#98 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 02:37 PM:

DavidD@83:

I don't believe 3.8.8.1 discusses eliminating anything at that stage. This phase just finds the two (or more if there are ties) works that are -eligible- for elimination. Because we need at least two things to have a comparison, we'll always need to have at least two works. In the elimination phase, we choose (usually) one of those two selected works to be eliminated. If there is a tie here, we have one tie break procedure, otherwise we just eliminate all the tied works as being equally the least popular.

Kilo

#99 ::: junego ::: (view all by) ::: May 27, 2015, 02:44 PM:

Cally Soukup @87

As for the word "nominee":
Kevin Standlee wrote HERE about a pending WSFS constitutional amendment that changes "nominee" to "finalist" and deprecates the use of "nominee" as a reference to "person or work appearing on the final ballot. He concludes:
Whatever you propose here, I hope it will use "nominee" to mean "work/person nominated in the initial voting" and "finalist" to mean "work/person appearing on the final ballot."
I was (and may well be again) somewhat confused about which was a "nominee" and which a "finalist", but I suspect I wasn't alone.

Thank you, I remember this now, I think I complied with this proposed amendment. Almost all the candidates/works/individuals we're dealing with in Section 3.8 are still nominees "nominated in the initial voting". In fact, Section 3.8 is where nominees get turned into finalists "appearing on the final ballot". I think we've used the 2 terms correctly. If you see a specific problem, please speak up.

And now I've typed "nominee" too many times and it just looks like a jumble of letters....

Yeah, my eyes were crossing while typing all that last night!

#100 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 02:50 PM:

Ian@89:

Keep in mind that there are actually two phases to awarding the Hugo. Currently, in the nomination phase, members submit works they think are Hugo-worthy and the top five nomination-getters are put on a final ballot. In the second phase, all Worldcon members (whether or not they nominated anything) rank their preferences using a process called "instant runoff voting" (you can find an explanation of it here).

What is broken is the nomination system, not the voting system, so this proposal affects only the nomination system. What you're referring to in the next part of your message is called "bullet voting" -- nominating a single candidate. In this case, it doesn't really help you.

Let's take your example where you nominated one work that was a finalist and four that no one else nominted. In this case, the odds are that your four non-finalists will be eliminated fairly early. The key to SDV-LPE (and this may be the part that you are missing) is that your points get re-calculated each round. Since you now only have one work your nomination ballot, then it gets your full point. Slates will tend to operate the same way: Their most popular candidate will eliminate their four least popular candidates. The end result is that they are very likely to get one work on the final ballot, but they are not able to get all five slots. If you nominate four other things that you think might be Hugo-worthy, you aren't hurting your favorite, and if enough people agree with your other four, some of them might make it, too. If enough people don't agree with your favorite, there's nothing you can do to get it on the final ballot, because you will never have more than one point.

Our goal isn't to prevent slates from being represented, our goal is to keep slates from keeping -others- from being represented.

Does that help?

Kilo

#101 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 02:54 PM:

Yes, I mis-read 3.8.8.1. And I think it's just me, I don't think the description is at fault. Sorry for that distraction!

(Thanks Kilo@98, junego@97, seth@86)

#102 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 02:57 PM:

Kilo@96: To me, a slate being "large" means it lists a lot of works. I think in FAQ 12 it's being used to mean that that slate has a lot of supporters (people who will nominate from it). That second is (at least to me) an unusual or even incorrect usage (the "slate" is the list of works, it is not the set of people supporting it), and it differs from our usage everywhere else, and I think it's confusing.

#103 ::: Seth ::: (view all by) ::: May 27, 2015, 03:04 PM:

Keith@93: The proposal refers to "minimum" and "maximum". Those terms are not specifically defined elsewhere; I believe that under current rules, both would be 5, but for purposes of 3.8.5 the minimum is 3.

I think the proposal needs to specify them, by reference to other sections if appropriate (so as to play well with other proposals that change those numbers).

#104 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 03:13 PM:

DavidD@102:

Ah, I think I see what you're saying. As a result of the Sad/Rabid Puppies this year, a "slate" is a complete set of 5 works for a category. However, I've also seen people referring to the Puppies themselves as a "slate" fairly commonly, but I agree this is technically inconsistent. It probably is more clear to say "if a slate has enough supporters that is certainly a viable possibility".

Does that re-write make more sense?

Kilo

#105 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 03:15 PM:

Seth@103:

Yeah, referencing section numbers is probably a good way to do it. I'll see what I can do with the next draft.

K

#106 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 03:18 PM:

Kilo@104: Yes, that's what I was having trouble with. And yes, that change does fix it for me.

(And if I'd bothered to do my own re-write at the beginning maybe it would have been clear what I meant from the beginning.)

#107 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 03:23 PM:

@90: [PANEL]: OK. I still disagree, but I respect your decision. And without you, even I don't think a panel is a good idea.

Which makes me think: T-shirts. I would love an "E Pluribus Hugo" shirt, and I suspect others would like it too. And if worn at the con, I think such shirts would be decent conversation-starters, to help us explain the proposal to those who don't know the details.

Would it be against the rules to make such shirts for personal use? How about for sale? How about for sale at the con itself? (I suspect that last one is probably verboten, or at least heavily regulated, but might as well ask.) And if it's not against the rules, who would be interested in getting one? And are there any artists who would be interested in designing one?

(Obviously, if selling them is legal, then we'd pay the artist a fair fee, and any profits left over would go to charity. If it were up to me, that would mean 50/50 to Worldcon and Electology.org, but obviously at this point I'm getting way way ahead of myself, because there are probably Rules About That, or at least norms.)

#108 ::: Joshua Kronengold ::: (view all by) ::: May 27, 2015, 03:35 PM:

I know of no rule (or even custom) against selling or giving away WSFS BM promo material (and still have some promo material from a joke BM proposal a few years ago). With the exception of the joke BM proposal (Yngvi is NOT a Louse, of course) I haven't seen someone do it, though.

#109 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 03:48 PM:

Jameson@107:
Which makes me think: T-shirts. I would love an "E Pluribus Hugo" shirt, and I suspect others would like it too.

That is an -awesome- idea. I usually end up spending most of my Worldcons at the various publisher parties anyway, so I had planned to do a lot of the "campaigning and education" that was mentioned for the panel there. Something like this would be a great conversation starter.

But if I can suggest a variation: There is a Worldcon tradition of making buttons on-site and passing them out. I have some that mean absolutely nothing to anyone who wasn't at that particular party, but are some of the most cherished momentos I have. What would you guys think about making a bunch of buttons for "E Pluribus Hugo" and passing them out to everyone?

Kilo

#110 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 03:52 PM:

OK, so with that august (and STUDLY) precedent on our side, are there any artists here who would be interested in designing a nice T-shirt? If we had a nice image, we might be able to use it for both T-shirts (for sale) and buttons (as freebies).

#111 ::: junego ::: (view all by) ::: May 27, 2015, 04:00 PM:

Keith "Kilo" Watt @90

[3-PHASE]
One thing about junego's version that instantly jumps out at me is that maybe the third phase should be the "calculation phase" at the beginning and not the "recalculation phase" at the end. As it is, there are actually four phases mentioned in his re-write, though one of them is only done once. I think the sequence is more clearly shown as:
1. Calculation (or I would call this "distribution" maybe)
2. Selection
3. Elimination
You then loop through 1,2,3. This may seem more reasonable to me just because that's how the code is written, I don't know.

It could be structured this way, but you'd then have to have 2 explanations/descriptions for Calculation/Distribution. The first 1-point allocation and division, then summation of all points. The second description would need to cover reallocating and resumming points after the elimination phase.

It just worked out easier for me to describe it more like: Initial conditions/explanations, Step 1, step 2, step 3, etc.

It logically works either way. Since the calculation/distribution done after an elimination phase changes the points each time, it also works to set it apart from the initial tabulation of points.

I have no strong preference for what it's called.

I still think there needs to be a clear, separate statement that the two numbers (nominations and points) are used to determine selection and elimination.

That "fourth phase" (tallying nominations) could also be broken out and named, but it isn't part of the repeated selective rounds. That would logically be the first phase, I guess.

[PANEL]
I'm starting to feel strongly that this is not a good idea. If we can't explain SDV-LPE in 20 minutes, there's a problem. And I don't think there is a problem. It's really a very simple system. Yes, the theoretical underpinnings and proofs can be quite complex, but while those were necesary to make sure we had a good and robust system, they are -not- needed to explain the system or why it's a good idea. And as someone else pointed out, if you've ever been to a panel, the potential for it to get out of control is real. This would leave a bad taste for everyone regarding the system. I just see lots of potential for bad and virtually no potential good. I would much rather that we focus on a tight presentation that can be convincingly done in 20 minutes than to have an hour-long panel that might not accomplish anything.

I've never been to a panel at a con, or more than a day tripper to some local cons a couple decades ago, so I am willing to bow to greater experience. Obviously we need ways to grow support for E pluribus, but that 's a topic for another post.

#112 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 04:22 PM:

@110: I'm definitely in for a t-shirt if someone can design one. My stunning lack of artistic abilities is something of a standing joke among my students...

I think using the same design for buttons to give out would be great as well -- random people are more likely to wear a button to show support (along with all the others they've collected) than wear a t-shirt, so having these for freebies would be great.

Kilo

#113 ::: DavidS ::: (view all by) ::: May 27, 2015, 04:26 PM:

Regarding the FAQ statement "there’s no reason not to nominate something you think even might be Hugo-worthy".

That's not literally true, as I understand the system. Suppose I like broadly popular work A and obscure work B. I might want to nominate only work B, so that it will get my full vote, rather than 1/2 a vote, protecting it in the early rounds. Or am I confused?

#114 ::: Seth ::: (view all by) ::: May 27, 2015, 04:35 PM:

#113 That's correct. If the popular work you like is going to get nominated anyway (there's always some risk it isn't as popular as you believe) then including it on your ballot lowers the point score of the other work to 1/2.

#115 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 04:40 PM:

DavidS@113: That's theoretically possible, but very unlikely, and very hard to determine is happening in advance. Enough people are involved that for one work to have any hope of surviving through to become an actual finalist, the difference between 1 and 1/2 vote (or even 1/5) on one ballot, initially, is very unlikely to be relevant.

So: Yes, it can be a better tactic. But nearly everybody who thinks about it that way is going to, in fact, be representing their preferences less well than by just listing the works they think are worth considering.

One of the problems in this level of voting math discussion is that, at the mathematical level, most of the good properties aren't actually "true"; in the mathematician's sense that they're always true. However, practically, if you don't have access to hindsight, behaving as if the good properties are true works better than trying to out-think the system. Because the corners where it behaves strangely don't have particularly clear up-front signatures, and they're very rare in real-world data.

That's kind of an unsatisfying answer, isn't it? Yeah, I feel that way too. Maybe there's another level of breakthrough in the theoretical understanding that will make all this simply explainable again; or maybe the logic of the situation simply doesn't conform to human prejudices (I know which way I'm betting on that one).

#116 ::: Soon Lee ::: (view all by) ::: May 27, 2015, 04:42 PM:

Ian@89:

Welcome. And welcome as well to other new commenters. It's good at this stage to get feedback from fresh eyes; helps point out what needs changing to improve clarity & ease of understanding.

#117 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 04:42 PM:

@113, 114: That's right, but... chances are that 1/2 point will not make a difference. Even if it helps your favorite have higher points, it would have to climb all the way to 4th place to avoid having its nominations checked anyway. That's tough to do, especially because chances are good that most of the other supporters of obscure B do not also like popular A, so there are only a limited number of strategic half-point-boosts available to B.

So I think (as I stated above) that we should change the FAQ text to "no good reason", but otherwise leave it unchanged.

#118 ::: DavidS ::: (view all by) ::: May 27, 2015, 04:43 PM:

Sorry, I had not read all the way down before posting. 67 is a perfect response to my question, written before I asked it.

#119 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 04:46 PM:

@116: Yes, welcome to all. In particular, if there are any puppies lurking here, any good-faith comments are welcome too.

#120 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 04:54 PM:

@115: Thanks for inspiring me to get back to work on my research. I have designed a voting system that actually is a bit of a breakthrough (Condorcet winners are always a unique strong Nash equilibrium). It's not applicable here, but it is a single-winner property that I would have told you was probably impossible to attain if you'd asked me 5 years ago. So I should really be writing that up.

So breakthroughs are not impossible. But yes, you'd be a sucker to bet on getting exactly the breakthrough you wanted.

#121 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 04:54 PM:

[PROPOSAL DRAFT]

Okay, all, next draft time! I have mergeed the excellent re-write that junego did with a couple of changes I had already made based on feedback. THere is some re-wording of junego's ideas, as well, but they aren't changed a whole lot. One thing that is different is that I have moved his "recalculation phase" to the beginning and called it the "calculation phase". I think this makes more sense as it's how the code actually is written.

If I've forgotten any signers, please alert me using the [SIGNING] tag. Also, if you think you may be attending the business meeting (and haven't let us know before), letting us know with the [ATTENDING] tag would be helpful for organizing a meetup.

Let me know your thoughts!

Kilo

=======================================

Short Title: E Pluribus Hugo (Out of the Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations), section 3.9 (Notification and Acceptance), and section 3.11 (Tallying of Votes) as follows:

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the process described in subsection 3.8.8.

3.8.8: The finalist selection process shall be conducted in rounds consisting of three phases:
3.8.8.1: Calculation Phase: For each category, the total number of nominations shall be tallied for each nominee. Next, a single “point” shall be assigned to each category on each nomination ballot, and that point shall be divided equally among all of the nominations on that ballot in that category that were not eliminated in previous rounds. Finally, all points from all nomination ballots shall be totaled for each nominee in that category. These two numbers, total nominations and total points, shall be used in the Selection and Elimination Phases described in 3.8.8.2 and 3.8.8.3
3.8.8.2: Selection Phase: Each round shall start by selecting the two (or more, as described in 3.8.8.5) nominees with the lowest point total for the Elimination Phase.
3.8.8.3: Elimination Phase: Nominees chosen in the Selection Phase shall be compared, and the nominee with the fewest total nominations shall be eliminated and removed from all ballots for the Calculation Phase of all subsequent rounds.
3.8.8.4: These rounds continue until the number of finalists specified in 3.8.1 remain. If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then instead none of the nominees from that round shall be eliminated, and all remaining nominees shall appear on the final ballot, extending it if necessary.
3.8.8.5: Ties shall be handled as described below:
3.8.8.5.1: If during the Selection Phase two or more nominations are tied for the lowest point total, all such nominees shall be selected for the Elimination Phase.
3.8.8.5.2: If during the Selection Phase one nominee has the lowest point total and two or more nominees are tied for the second-lowest point total, then all such nominees shall be selected for the Elimination Phase.
3.8.8.5.3: IF during the Elimination Phase two or more nominees are tied for the fewest number of nominations, the nominee with the lowest point total at that round shall be eliminated.
3.8.8.5.4: If during the Elimination Phase two or more nominees are tied for both fewest number of nominations and lowest point total, then all such nominees tied at that round shall be eliminated.

Section 3.9: Notification and Acceptance.
3.9.1 Worldcon Committees shall use reasonable efforts to notify the nominees finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each nominee finalist shall be asked at that time to either accept or decline the nomination. If the nominee any finalist(s) declines the nomination, that nominee finalist(s) shall not appear on the final ballot. In this event, once all finalists have had the opportunity to decline, the nomination system shall be re-run as described in Section 3.8 with the declined nomination(s) removed from the nomination ballots on which they appeared. The eligible finalists from this re-run shall be merged with the remaining potential finalists from the original run. If this merge would result in more than the maximum number of finalists, then the ballot shall be extended to include the finalists from both the original and the re-run of the nomination system. This procedure shall also be used in the event that a finalist is deemed ineligible.

Section 3.11: Tallying of Votes.
3.11.4: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. During the same period a record of all rounds of the selection process for each category shall also be published.

Submitted by: Keith “Kilo” Watt, Tammy Coxen, Yoana Yotova, Joshua Kronengold, Chris Battey, CJ Cabourne, Steven Halter, P.J. Evans, David Goldfarb, Seth Gordon, Ginger Tansey, Steve Wright, Catherine Faber, Andy Holloway, Duncan J. Macdonald, Claudia Beach, Derry Earnshaw, Jason Skiles, Soon Lee, David Harmon, Lydy Nickerson, Jameson Quinn, Abigail Sutherland, Lee Billings, Christopher Battey, Oskari Rantala, Seth Breidbart

Commentary: “E Pluribus Hugo” – the name of this proposal is also its goal: recognizing the many opinions within fandom as to what nominees might be worthy of the Hugo. It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members: They still list the nominees, unranked, they feel are Hugo-worthy (up to the maximum permitted) in any categories they choose, just as they have in the past. In the past, we have counted the number of nominations each nominee received, and the top five nominees were put on the final ballot. However, because SF fandom typically nominates a variety of different nominees, it was easy for an organized slate to make it so that no other nominees made the final ballot.

Using this system, fandom isn't penalized for nominating a wide variety of nominees. If you nominate something that ends up not having a chance to make the final ballot, then your remaining choices automatically get more of your support instead of just being wasted. In other words, you can safely nominate anything you feel is Hugo-worthy. If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot. In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which nominees should be voted on to be the final Hugo winner.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no minority of members – of any sort – can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Award.

#122 ::: David Gallaher ::: (view all by) ::: May 27, 2015, 05:03 PM:

@107 @109 Ribbons are traditional and quite common, and can be given out to supporters and to people you talk with.

#123 ::: ULTRAGOTHA ::: (view all by) ::: May 27, 2015, 05:07 PM:

"Which makes me think: T-shirts. I would love an "E Pluribus Hugo" shirt, and I suspect others would like it too. And if worn at the con, I think such shirts would be decent conversation-starters, to help us explain the proposal to those who don't know the details."

There is a Tshirt vendor who makes the Sasquan shirts who I assume will be dealing at the con.
http://www.offworlddesigns.com/sasquan-friends-t-shirt/


Also, badge ribbons are cheaper than buttons and would go down well.

#124 ::: Cally Soukup ::: (view all by) ::: May 27, 2015, 05:18 PM:

[PROPOSAL DRAFT]

In 3.9.1 you speak of "finalists" before they've accepted, and then later you call the same works/people "potential finalists" after they've accepted when new finalists are being merged into them following the declining of a nomination, and calling the new, pre-notification works/people "finalists". This is exactly backwards to my way of thinking.

In 3.8.8.5.1 would it be possible/a good idea to explicitly point out that if works tie for last place, we don't bring in second-to-last before comparing, or is it just me that had to read that section more than once?

#125 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 05:19 PM:

Yeah, buttons are by no means passé, but ribbons seem to be the current leading way of showing affiliation (and making jokes). Stickers to go on badges aren't gone either, but tend to be small for what we probably want to say. And t-shirts of course, but they're more expensive and most fans (despite rumors about fannish hygiene) only wear them one day between washings.

So, I think the ideal choice is to have both t-shirts and ribbons (horizontal self-stick ribbons, last I checked at a Worldcon).

#126 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 05:21 PM:

I think using the official Hugo logo in a shirt design for us is completely legit (no trademark issues). Anybody who knows anything think differently? The key issue is that we're using it to identify the thing that it actually, legally, does identify, the real Hugo awards given out by WSFS.

#127 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 05:24 PM:

Cally@124:

Thanks, people had decided they wanted "finalist" instead of"potential finalist". I thought I had removed all the "potential" instances, but they are slippery beasts!

Kilo

#128 ::: Cally Soukup ::: (view all by) ::: May 27, 2015, 05:25 PM:

[SWAG]

As well as, or even more than, buttons, in recent years badge ribbons have become extremely popular, with many people collecting dozens of them at the cons they attend. You can have (fairly simple) art on them, not just words, and they're quite inexpensive per unit, especially if you're buying by the hundreds. There's no reason you couldn't do both (here's a thought: Nancybuttons might cut you a volume discount on the buttons! She goes to a bunch of cons, though I don't know if she's going to Worldcon).

#129 ::: Cadbury Moose ::: (view all by) ::: May 27, 2015, 05:51 PM:

DDB @ #126

It needs to be cleared by the Mark Protection Committee, though.

(It also might need to be an official WSFS product, too.)

IANAL(NDIPOOTV)

#130 ::: Chris Suslowicz ::: (view all by) ::: May 27, 2015, 05:54 PM:

Assuming supporting members can be submitters, you may add me to the list if desired.

#131 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 06:57 PM:

Chris@130: Yep, supporting and attending members can both be signers. I've added you to the list!

K

#132 ::: Christopher Battey ::: (view all by) ::: May 27, 2015, 07:00 PM:

[SIGNING]

Sorry my name is causing confusion. I'm now listed in the submitter list as both "Chris Battey" and "Christopher Battey"; please remove the former (as the latter is the name under which I am registered).

[SWAG]

I'm looking into the badge-ribbon possibility. Any idea what a good quantity of ribbons would be for the purpose? Since the cost per unit drops quite a bit on larger orders, those of us planning to purchase and distribute ribbons should coordinate a single order, which we can then divide up among us at the convention.

Simplicity in the design is best here; the idea I'm working from at the moment is just "E PLURIBUS HUGO" as serif text (as close as possible to that used on US seals, etc), printed in silver on black ribbons.

This could coordinate well with a t-shirt design parodying the US seal - perhaps an eagle clutching a sheaf of ballots in one talon and a Hugo in the other, with "E PLURIBUS HUGO" ringing the text at the top and "2015" at the bottom? Using the Hugo rocket element as a small part of a larger design may make it less likely to run afoul of the Mark Protection Committee.

#133 ::: Cally Soukup ::: (view all by) ::: May 27, 2015, 07:07 PM:

[SWAG]

If the eagle is clutching the rocket (or sheaf of rockets?) more or less by its base, so the distinctive fins can't be seen, would that help get it past any problems with the Marks Committee? It's their job to be paranoid about the symbol, after all. I think we need to send up the Standlee Signal; he knows the Marks Committee much better than any of us, and thus would have a better-informed opinion.

#134 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 07:34 PM:

Chris@132: Got it entered, thanks for letting me know!

[SWAG]
500 ribbons probably isn't unreasonable at all. Popular ones can go like hot cakes. I'd be willing to contribute to the production costs, as well.

K

#135 ::: David Gallaher ::: (view all by) ::: May 27, 2015, 07:52 PM:

[SIGNING]
Please add my name as a submitter.

[ATTENDING]
I will be attending the Business Meeting.

#136 ::: Adam Tilghman ::: (view all by) ::: May 27, 2015, 08:01 PM:

[SIGNING]
Please add me to the submitters list. I plan to attend the business meeting.

[SWAG]
@132: I'll happily contribute to ribbon production! 500-1000 would probably suffice.

#137 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 08:22 PM:

[SWAG] I'd be happy to help do the ribbons. But I'm still hoping to hear from bona fide artists/designers.

As to the "eagle with hugo" idea: could be good. Non-US people: would that seem too US-centric? What if the eagle were drawn in Northwest Indian (Kwakwaka'wakw or whatever) style? Or maybe use the Sasquan raven, but in an eagle-like pose? Would that help make it clear it's Sasquan-related, or would it be too "thematically busy"?

Other ideas:
-A rocket ship made out of a lot of little rocket ships
-A rocket ship made out of the names of all past "best novels".
-something like this

#138 ::: Fragano Ledgister ::: (view all by) ::: May 27, 2015, 08:41 PM:

I support this proposal and am willing to sign on to it. No voting system is perfect for every purpose, but this one will serve to improve the basic function of Hugo voting. That is to give the fans who are voting members of Worldcon the chance to express their beliefs regarding the quality of the works of sf/f they have read or viewed in the previous year.

#139 ::: Cally Soukup ::: (view all by) ::: May 27, 2015, 08:54 PM:

[SWAG]

I'd personally recommend against a super-busy design for ribbons, or one that required really fine lines. I've seen some fairly blotchy ribbons (and also, to be fair, some really well-printed ones), and don't know what company printed which so as to recommend a really good company. Best to err on the side of simpler design, in my opinion.

#140 ::: Lori Coulson ::: (view all by) ::: May 27, 2015, 09:09 PM:

I'm willing to sign on as well.

#141 ::: David Harmon ::: (view all by) ::: May 27, 2015, 09:10 PM:

[SWAG]

Personally I'd rather have a T-shirt or even a button than a ribbon, but then, I'm sadly out-of-date in con fashions. :-) I'd certainly buy any of the three.

#142 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 09:22 PM:

David@141:
I hear you. I'm partial to buttons, too, but I'm also likely to be an old fogey when it comes to con fashions. :) I'll go with whatever the group wants.

K

#143 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 09:22 PM:

[SWAG]

@139: For the ribbons, it should probably just be the slogan, in the same font as used on the T-shirt.

For the T-shirt, a rocket made out of all the winning book names, with the slogan underneath, would be awesome. I think it would look great, it would clearly get the message across, and it would be something that had a meaning beyond just this proposal. And yes I know that the Hugo is more than "best novel", but I think it's fair to use that as representative.

#144 ::: Jeffry Herman ::: (view all by) ::: May 27, 2015, 09:26 PM:

[SIGNING]

I have followed the discussions with interest and I am very much in favor of this proposal. If it would be useful in any way for you to add me as a submitter, please feel free to do so. I am a supporting member of Sasquan.

#145 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 09:28 PM:

David@141:
I hear you. I'm partial to buttons, too, but I'm also likely to be an old fogey when it comes to con fashions. :) I'll go with whatever the group wants.

K

#146 ::: Jameson Quinn ::: (view all by) ::: May 27, 2015, 09:34 PM:

[SWAG AND MONEY]
So people pay cost for ribbons or buttons? Or are they usually given for free?

In any case, I'd like to talk about money. Part of the idea of the T-shirts was as a fundraiser. I originally suggested that it should go 50/50 to Worldcon and Electology.org. But... now that I look into it, "Worldcon" isn't really a thing; that is, if I'm not mistaken, each separate convention has its own budget, with no money rolling over year to year. So, is there some generally accepted Hugo-related nonprofit? And do people agree that using this as a fundraiser might be a good idea in general? Obviously, I'm biased; but I really do think that electology.org is a good cause, and I'm not in any sense trying to put this money into my own pocket.

#147 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 09:35 PM:

Some preliminary t-shirt design ideas (wrong format for ribbons, and I think too fancy, and some use of colors).

I'll call them #1, #3, and #5.

Originals done in Adobe Illustrator entirely from vector elements. Just some ideas to consider and work from -- perhaps actual designers / illustrators will be motivated to work on this after seeing my suggestions :-).

From my understanding of trademark law prior approval is not required. The law isn't about who may use the logo, it's about what the logo is used to identify, and we're using it to identify the thing it does legally identify, the actual Hugo Awards. Some members of the mark-protection committee have participated in the discussion, their saying there's not a problem would in practice settle it, eh?

#148 ::: Tim Illingworth ::: (view all by) ::: May 27, 2015, 09:35 PM:

[MPC] Moose @ 129: I'm pretty sure we like it, as it's being used to indicate the WSFS Hugo.

#149 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 09:41 PM:

Tim@148: Excellent timing! (identical timestamp to my #147 saying members of MPC are active in this discussion)

#150 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 09:45 PM:

One obvious design element for a t-shirt or whatever is, um, something canine. I chose to avoid that myself.

I personally don't agree with defining what we're doing here as solely anti-puppy, in particular. And also I think it's poor political communication on top of that. When expedience and actual conviction point the same way, I go that way!

I note nobody else has suggested it yet, so perhaps everybody is already on-board with that approach and this message is unnecessary.

#151 ::: Tammy Coxen ::: (view all by) ::: May 27, 2015, 09:51 PM:

The Mark Protection Committee is a standing committee that carries on from year to year, and they always need money to continue their work (which includes protecting the Hugo and related services marks) so that's a possible Worldcon-related recipient for receipts from T-shirts.

#152 ::: Tammy Coxen ::: (view all by) ::: May 27, 2015, 09:53 PM:

DD-B - I'm getting 404 errors on your links.

#153 ::: David Goldfarb ::: (view all by) ::: May 27, 2015, 10:00 PM:

I'm not sure if everyone here knows that Lee and her partner are professional t-shirt / button / jewelry makers. I don't know whether they're planning to go to Sasquan, but I think it's worth pointing her at this discussion. Excuse me a moment --

#154 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 10:14 PM:

Jameson@146:
So people pay cost for ribbons or buttons? Or are they usually given for free?

Buttons and ribbons are always handed out for free. It's part of the con experience. I think if even some of the group is willing to pitch in, we can get what we need for very little cost to individuals.

I'm not in favor of doing any kind of fundraiser. Our goal is to convince people to vote in favor of the proposal, so we really don't need money to achieve that goal. The people at the business meeting have already paid a huge sum just to be there. Our asking for money would look -- odd.

I'm not -totally- opposed to offering t-shirts at cost, but the odds of us being left with quite a few are pretty high. I would say to plan on leaving the t-shirts as something the supporters here and their friends can order if they want.

Kilo

#155 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 10:16 PM:

Tammy@152: fixed.

Don't know why it had that effect; fiddling with the auto-index setting changed the behavior of a link to a specific file, which makes no sense. But changing it back fixed it, so fixed.

#156 ::: Tammy Coxen ::: (view all by) ::: May 27, 2015, 10:22 PM:

I agreed with Keith re selling T's at cost in advance to supporters being a better idea. If we can get art, then we can use CafePress or Zazzle or one of those kinds of sites, so no one has to commit to a certain size order and be stuck with them.

#157 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 10:23 PM:

[ATTENDING]

Here's the current list of Business Meeting attendees, according to my notes. If I've missed anyone, please let me know!

Kilo
===================================

Worldon Business Meeting Attendees

Keith "Kilo" Watt
Tammy Coxen
Yoana Yotova
Chris Battey
Joshua Kronengold
Jameson Quinn
David Gallaher
Adam Tilghman

#158 ::: Tim Illingworth ::: (view all by) ::: May 27, 2015, 10:37 PM:

[Attending]

I hope to be there, but I've just accepted a new job, which may throw a spanner in the works.

I expect Kevin Standlee will also be there...

#159 ::: Lee ::: (view all by) ::: May 27, 2015, 10:43 PM:

David G., #153: I haven't been tracking the discussion closely. The short version from this end is:

1) Re T-shirts, we are not set up to do full-color separation printing. We can do up to 4 colors on a shirt, but fewer is preferable. We need line-art, and at this point halftones are problematic (software issues).

2) Re buttons, those could be done in full-color on white if necessary; however, we haven't produced buttons regularly for about 5 years now and a few things would have to be re-done for it to happen again. My partner assures me that we still have in the vicinity of 1,000 blanks.

3) Ribbons we've never done; it's much easier to order them from a specialist.

#160 ::: JJ ::: (view all by) ::: May 27, 2015, 10:57 PM:

Keith "Kilo" Watt, #157:

I'll be Attending, but would prefer to have my name left off any official lists, for reasons.

Have you an e-mail you can rot13, or are you one of the numerous Keith Watts on Facebook?

#161 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 10:57 PM:

[SWAG]

Idle thoughts here...

One thing about buttons is that we can make the design match the t-shirts. If the presenters are wearing the t-shirt the day of voting, and we have a number of people wearing buttons that match, that could make a strong statement. I'd imagine people will have lots of ribbons, so ours won't be quite as obvious in the crowd.

But again, I'm not opposed to ribbons at all, we'd just need a much simpler design for it. Nothing that says we can't do both, if the cost isn't prohibitive.

Kilo

#162 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 11:11 PM:

Rough ribbon design idea -- close to the most obvious possible design I believe.

I couldn't find the right suppliers in a quick look, so I don't think the aspect ratio is right for the normal ribbons. If somebody can point me at that info I can revise this, or other people can use it.

ribbon #1.

Again, original is in Adobe Illustrator from all vector elements.

#163 ::: JJ ::: (view all by) ::: May 27, 2015, 11:16 PM:

David Dyer-Bennet, #147:

I like #1 far, far better than the other two -- with the exception that the "Ee'd Plebnista" script needs to be just a little less curly, so that it's readable.

#164 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 11:17 PM:

(Oh, and I should say here that I'm sending these design ideas into the discussion stream and don't feel all that proprietary about them. Do minor modifications if you like, or just use them if you like.)

(Or react in horror and do better designs of your own, which is perhaps the best possible outcome.)

#165 ::: David Dyer-Bennet ::: (view all by) ::: May 27, 2015, 11:25 PM:

JJ@163: The missing numbers, of course, are me playing with different fonts. Yeah, #1 may be overly finicky (though I found worse). I've currently got a full Adobe Creative Suite subscription, so I've got the full Typekit collection in addition to my own collection, so I can spend days looking for just the right font :-).

Actually, here's my #2, a bit less loopy.

It's of course trivial to change the font used, if consensus comes down to the basic idea but wants different fonts I can work towards that if nobody else jumps on it.

Not very committed to any particular design, just throwing out ideas since discussion seems to flow better when there's something to discuss.

(And I put up a bigger jpeg of the ribbon while I was at it, same URL as before.)

#166 ::: Avram ::: (view all by) ::: May 27, 2015, 11:28 PM:

I’ve worked up a first draft of a ribbon design, but before I devote more time to this, I’d want to know:

  • What are the final working dimensions of a convention ribbon. (Not just the physical object, but the printable space that’d be visible in actual use.)
  • Has anyone on the Marks Protection Committee said anything official about how much use of the Hugo™ Rocket is allowable?

#167 ::: JJ ::: (view all by) ::: May 27, 2015, 11:40 PM:

Avram, #166: What are the final working dimensions of a convention ribbon. (Not just the physical object, but the printable space that’d be visible in actual use.)

Badge Ribbons usually come in thin (4" x 1-5/8") and thick (4" x 2"). I'm not sure how much white space is required from the edges.

#168 ::: Keith "Kilo" Watt ::: (view all by) ::: May 27, 2015, 11:41 PM:

[SWAG]

T-shirt design #5 is my favorite (though maybe with a different font). I don't know how much we want to look like the U.S. Constitution, since the name is pointing out that the Hugo is coming from -all- of us (and not a small subset) and isn't really referencing American history.

That said, as I mentioned, I have absolutely no artistic taste whatsoever, so take that with a huge block of halite. :)

Kilo

#169 ::: JJ ::: (view all by) ::: May 27, 2015, 11:49 PM:

Keith "Kilo" Watt, #168: I don't know how much we want to look like the U.S. Constitution, since the name is pointing out that the Hugo is coming from -all- of us (and not a small subset) and isn't really referencing American history.

I think Americans might see the visual allusion, and anyone else would just say, oh, look at the interesting Latin script.

Is there a font that's frequently associated with Latin?

#170 ::: JJ ::: (view all by) ::: May 27, 2015, 11:55 PM:

Oh, I like this font.

#171 ::: P J Evans ::: (view all by) ::: May 28, 2015, 12:08 AM:

170
Something like Trajan or Weiss Roman then. (Goudy Oldstyle, Goudy Handtooled, Zapf Calligraphy...)

#172 ::: David Dyer-Bennet ::: (view all by) ::: May 28, 2015, 12:19 AM:

Avram@166: Marco (thank you JJ@167) gives dimension details on their site. For the "narrow" ribbons,

Standard Imprint Area: .9"h x 3.4"w

Product Size: 4"w x 1-5/8"h

#173 ::: Kevin Standlee ::: (view all by) ::: May 28, 2015, 12:19 AM:

1. I am in the process of driving from my home in Fernley NV to Spokane WA for a planning meeting coming up this weekend for this year's Worldcon. That means that I'm not hovering over my computer waiting for the Standlee Signal to go up. Instead, I'm driving my minivan up US-95 tomorrow. So don't expect quick turnarounds on messages.

2. I am the current chairman of the WSFS Mark Protection Committee, the permanent body charged by WSFS to manage and protect WSFS's registered service marks, including Worldcon, The Hugo Award, the Hugo Logo, and the design of the Hugo Award trophy rocket. These marks are licensed to each Worldcon committee for their use, and they can sub-license those marks to their official merchandisers (Off World Designs, for this year's Worldcon).

3. The MPC can and sometimes (but rarely) does directly license uses of the marks.

4. I find myself very uneasy with the idea of a WSFS mark being used in the way y'all are discussing it here, regardless of who gets any money involved from it.

I will point the WSFS MPC at this discussion and ask the committee to express their opinion about whether the uses being discussed here are things the MPC approves of, with or without any licensing of the marks. I'll let you know what comes of that, but again, remember that I (and many other members of the current Worldcon's committee and staff) are traveling this weekend are holding meetings to plan the current Worldcon, not sitting at our computers.

#174 ::: gaukler ::: (view all by) ::: May 28, 2015, 12:35 AM:

[SIGNING] Add my name to the list, please:
Mark Shier

#175 ::: Cally Soukup ::: (view all by) ::: May 28, 2015, 01:19 AM:

[SWAG] [Ribbons]

DD-B @162: the ribbon with the Hugo rocket on it is problematic, because of the half-toning. I seriously wouldn't trust the printer to get that right. I mean, they might, but line drawings are probably best.

Avram @166: I like the banner a lot.

#176 ::: David Dyer-Bennet ::: (view all by) ::: May 28, 2015, 01:55 AM:

Cally@175: There's no halftone in the ribbon design. What there is, is two ink colors. For a ribbon, reassigning a bit will get a recognizable image down to one ink color, I just didn't bother to do it for the draft.

#177 ::: Avram ::: (view all by) ::: May 28, 2015, 01:58 AM:

Kevin Standlee @173, am I correct in reading that your uneasy not just with the use of the rocket, but also with the word “Hugo”?

If so, everybody, should we maybe come up with an alternative title? I think “palma” is Latin for award, but someone who actually knows the language should double-check that “E pluribus palma” doesn’t mean something terrible.

#178 ::: Steven desJardins ::: (view all by) ::: May 28, 2015, 02:21 AM:

I'll be attending the Business Meeting.

#179 ::: P J Evans ::: (view all by) ::: May 28, 2015, 02:28 AM:

[SWAG]
I would say that swag should be way down on the list of Necessary Things. Ribbon or button, maybe - but remember that not everyone is going to be happy with these proposals.

#180 ::: David Dyer-Bennet ::: (view all by) ::: May 28, 2015, 02:37 AM:

P J Evans@179: in some ways, it's precisely because there isn't any clear consensus on what to do yet that flagging our position is valuable. It will get people to ask "What's that?" or "Why do you support that rather than 4/6" or whatever, thus leading to (hopefully polite and useful) discussion. (And people choosing to wear it should be prepared to be asked things like that!)

#181 ::: JJ ::: (view all by) ::: May 28, 2015, 02:44 AM:

My idea for a ribbon design. (Please forgive me, my Photoshop is broken and I had to do this in Paint. Please also forgive me the crappy Photobucket site, I don't have a webspace.)

#182 ::: Soon Lee ::: (view all by) ::: May 28, 2015, 03:01 AM:

[SWAG]

According to WSFS, "World Science Fiction Society", "WSFS", "World Science Fiction Convention", "Worldcon", "NASFiC", "Hugo Award", and the distinctive design of the Hugo Award Rocket are service marks of the World Science Fiction Society, an unincorporated literary society.

How about keeping it simple: just the phrase (no rocket) "E pluribus Hugo" with an appropriate font, so the focus is on the phrase?

We should be fine so long as we don't use the phrase "Hugo Award" is my understanding, so we needn't change the name for the proposed rule change either.

(This is all a fun digression, but by way of trying to drag the conversation back to the matter at hand, are there any substantial matters we need to address?)

#183 ::: Buddha Buck ::: (view all by) ::: May 28, 2015, 07:27 AM:

[SIGNING] Supporting member, will not be attending. My name is my name.

#184 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 07:55 AM:

@173: [MPC] Thanks for responding. We all understand if you're not around here 24/7.

As you can see, we're discussing several ideas in the "swag" department, which we'd like clearance or otherwise for.

Money:
-Giving items at cost or free
-Giving items with a small markup, which would go 50% to the MPC and 50% to electology.org (after paying the artist a nominal fee; with no payment or free shirts to the organizers. I'd be happy to pay for Keith's shirt, though.)

Content:
-Use of the phrase "E PLURIBUS HUGO" itself
-Use of the list of past winners
-Use of a recognizable rocket silhouette as one small element of a larger design
-Use of a design that's basically just a rocket
-Use of Futura font, as in the official Hugo logo

I understand that it's not just up to you, but if you're consulting with the committee, we'd like explicit permission or otherwise for reasonable combinations of the above.

Others: is there anything I missed, that we need to ask?

[SUBSTANCE]

I actually think we've got consensus on all the substantive issues. We're still wordsmithing, but that's coming along pretty well, too.

#185 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 09:05 AM:

Here's my first attempt at a design.

(I said earlier we should pay the artist, but if we use something I did, I do not want to be paid.)

#186 ::: David Harmon ::: (view all by) ::: May 28, 2015, 09:06 AM:

My mini-reviews of the designs seen so far:

David Dyer-Bennet #147 et al:
#1: Ow, I've got a serif in my eye!
#2: Not quite as bad as #1, but that's still pretty loopy.
#3: Is that meant to be an interference pattern? Why? Without that, it would be servicable.
#5: Better, it's straightforward. meant to be high on the shirt?

For the ribbon designs, DDB #162 is servicable, but I like JJ #181 better.

#187 ::: David Harmon ::: (view all by) ::: May 28, 2015, 09:12 AM:

A thought about fonts: Maybe use commonplace web fonts, since this has been hashed out on the web? What fonts are most of you getting from Making Light? I'm not sure what actual font is in use, but my browser is telling me the comment boxes have font families: "verdana,arial,sans-serif".

#188 ::: Kevin Standlee ::: (view all by) ::: May 28, 2015, 09:15 AM:

I'm checking in here before checking out from the hotel and setting forth across the wastelands of northern Nevada and eastern Oregon (and if any of you follow my LJ, you'll know that I have vehicle issues that make me nervous about this).

The Mark Protection Committee needs some time to think about this. Not all of its members live with continuous e-mail feeds to their brains like some folks.

However, I'm pretty sure that a ribbon that says

E PLURIBUS HUGO

and that does not include the Hugo Award Logo in any form, would be just fine.

You have to understand that what WSFS (and the MPC) wants to avoid is the appearance that any official entity (either WSFS as a whole or a seated Worldcon committee) is giving official sanction to any particular proposal before the Business Meeting. Licensing the use of the official marks including the logo would imply approval of the proposal.

Imagine some hypothetical proposal that would give a certain Northern European publishing house the right to name all of the Hugo Award finalists. It's technically just as legal as EPH. Now imagine its supporters wanting to produce ribbons and shirts with Hugo Award logos on them to show that they are the True Supporters of Hugo. See the problem?

#189 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 09:19 AM:

@188: I realize you're probably already gone. But if not: do you think my @185 counts as violation? It's arguably just a list of the finalists...

#190 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 09:20 AM:

Sorry, I meant "just a list of past winners".

#191 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 09:26 AM:

Actually... if you'd rather not comment on specific designs, I understand. What I really want is some indication of how expansive you would intend to be about what counts as being the trademark. I totally understand that you probably don't want to officially approve anything that has the trademark. But is that just "anything that unequivocally has a picture of the Hugo rocket", or "anything that hints at a Hugo rocket", or something in between?

#192 ::: Cally Soukup ::: (view all by) ::: May 28, 2015, 09:30 AM:

[SWAG]

Jameson @184: Usual convention culture (at least in the Midwest US, which is what I'm familiar with) is to give away ribbons for free, except for the special case of ribbons that support a money-requiring cause such as a Worldcon bid, and even then they're often free. Since the Business Meeting is free, it's my feeling that charging anyone (except us) for the ribbons would be counter-productive.

I haven't seen anybody give away buttons in long enough that I don't know what the current culture there is. Instead of giving out buttons, it might be a good idea to have known attendees (who are willing to talk about it) wear buttons that say "Ask me about E Pluribus Hugo" and have them be the ribbon distributors. People might even ask....

#194 ::: Yarrow ::: (view all by) ::: May 28, 2015, 10:10 AM:

Jameson Quinn @ 191 and 193: I really like the design! And I'm really leery of treading so close to the Hugo mark, no matter what the Mark Protection Committee says. It seems to be inadvertently hinting at the thing Kevin's worried about, an implication that the WSFS is behind the proposal.

The standalone phrase E Pluribus Hugo does not have that implication, to my eye.

#195 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 10:25 AM:

@194: How about just "E PLURIBUS HUGO" (or "E•PLURIBUS•HUGO") on the front, and the suggestively-shaped list of winners on the back? It's not directly pairing the two things, but... well, I'd really like to actually wear that list on a shirt.

In fact, if we did the front/back thing, we could use different fonts — a "colonial" font for "E pluribus Hugo", and Futura for the list.

I know, two sides of the same shirt is still connected. But ...

(latest version... I added the retro awards because, Farenheit 451; and I tinkered with the ordering and spacing a bit.)

#196 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 10:44 AM:

Personally, I'd even rather have the list on the back, and the slogan with a disclaimer on the front ("Not endorsed by WSFS"), than a shirt with just the slogan and no list or disclaimer. I know, I made it so I'm biased; and I should be modest about it; but I can't help thinking that list is just cool.

#197 ::: Kevin Standlee ::: (view all by) ::: May 28, 2015, 10:45 AM:

[Last chance before I cast off from the hotel. Next online late tonight in Pendleton, if I'm lucky.]

The versions that don't explicitly use the Hugo Logo itself are better than those that do. WSFS doesn't claim ownership of every rocket-shaped design there is (that would be silly), only the specific implementation of the Hugo logo and the Hugo Award trophy rocket, both of which are registered service marks.

#198 ::: Joshua Kronengold ::: (view all by) ::: May 28, 2015, 10:56 AM:

I would love to have a shirt with that design, but I think that to avoid the appearance of favortism, we should wait on the spaceship-suggestive design until -after- this Worldcon (alas). Probably after next Worldcon, too. If at that point the measure is now part of the WSFS constitution, I could see the MPC looking on it substantially more favorably.

Until then.. I'm not sure of a good approach that uses the literal many -> one approach without skirting too close to someone's mark. The little works making the word Hugo -might- work, but I'm not convinced.

I would much rather have a ribbon with "E Pluribus Hugo" and only that on it. If we want to order -two- ribons, I could see also providing "Out of Many, a Hugo" which people could wear as an alternative or explaination.

It's worth noting that usually badge ribons just have a block-printed phrase on them--I've seen other things done with the space, but it's comparatively rare.

#199 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 11:02 AM:

So if I'm reading @197 right, he doesn't want to comment on specific designs, but he is not now ruling out the possibility of the more subtle rocket-type things.

With that in mind, I think my latest is clearly in bounds, and maybe actually more timid than it needs to be. I've tinkered with the shape a bit more (and there's probably little more that can be done), and I've proposed splitting the shape and the slogan across front and back.

#200 ::: Keith "Kilo" Watt ::: (view all by) ::: May 28, 2015, 11:04 AM:

[SWAG]

I think Jameson's 195 t-shirt design is pretty cool, too.

I think Cally@192's suggestion is really good: Primarily go with the horizontal ribbon with just the "E Pluribus Hugo" text (in whatever font) for freebies to help generate support, and then make some "Ask Me" buttons for those attending who are willing to wear them. As long as MPC says okay, T-shirts for anyone who wants to order one would be a cool extra. I'd wear mine on at either the PBM or on the actual day of proposal presentation. If MPC says the rocket design is out, I'm not sure the t-shirts will do much more than the badge/ribbon combination.

Just my thoughts,
Kilo

#201 ::: Keith "Kilo" Watt ::: (view all by) ::: May 28, 2015, 11:07 AM:

[STANDLEE SIGNAL]

Kevin, whenever you're at your landing point for the day, two quick questions:

1) Will there be any kind of A/V equipment at the business meeting? PowerPoint, etc.?

2) Once we submit the proposal to the address you provided, is it locked in or can it be modified before the deadline? I'm thinking we are almost ready to send the proposal out into the wild, so the business meeting site is an obvious place to link people to. I'd like feedback from people outside the group, and if someone had an awesome suggestion, it'd be nice to be able to go back and add that in.

Thanks,
Kilo

#202 ::: Keith "Kilo" Watt ::: (view all by) ::: May 28, 2015, 11:10 AM:

[PROPOSAL]

All: I haven't seen any suggested changes to the proposal in a while. Are we happy with that part? For me, I'm pretty satisfied -- I think we've done a great job on it. Let's take a quick poll and see if there are other outstanding issues.

I'll upload the current FAQs later on this morning, and we'll see if we can iron out the last wrinkles in those.

Thanks much!
Kilo

#203 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 11:27 AM:

Obviously, even the Standleeest of Standlees can't give a definitive ruling without consulting the committee. But I think that even then, he's not going to say "yes" to anything, but rather just be very specific about what he says "no" to, so we're going to have to read between the lines a bit. In that context, I read his latest as being cautiously positive about the possibility of the suggestively shaped list.

#204 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 11:29 AM:

No issues for me with the proposal. We do have a few things to iron out in the FAQ.

#205 ::: Cally Soukup ::: (view all by) ::: May 28, 2015, 11:41 AM:

Yes, I seriously doubt that the MPC would have a problem with a rocket-shaped list of winners, as opposed to the actual "mark" which is the Hugo novel graphic.

I think it would be a good idea to get our proposal on the Business Meeting page as soon as is reasonably possible, in order to a) cause buzz, and b) possibly head off less carefully-thought-out ideas.

With that in mind, has the FAQ changed since the last time it was posted here? Because you're right, we should be focusing on that.

#206 ::: Cassy B. ::: (view all by) ::: May 28, 2015, 11:44 AM:

[SWAG] I'd buy one of those t-shirts with the rocket made out of Hugo-winners, and I'm not even going to Worldcon...

[FAQ] Could whomever has the most recent draft of the FAQ post it, so we can go over it for clarity? I agree that the actual proposal looks good.

#207 ::: Cassy B. ::: (view all by) ::: May 28, 2015, 11:46 AM:

Oh, and [TIES], I agree that explicitly empowering admins to change ties is opening an unnecessary can of worms, and possibly inviting accusations of favoritism in the future. That's a headache I really don't think future admins need.

#208 ::: Seth ::: (view all by) ::: May 28, 2015, 12:35 PM:

[Attending] I generally attend the Business Meeting, and certainly plan to do so this time.

#209 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 12:47 PM:

My wife, veteran of kern-picking the title page of her thesis, showed me a better way to do the spacing.

#210 ::: junego ::: (view all by) ::: May 28, 2015, 01:03 PM:

Keith "Kilo" Watt @202
[PROPOSAL]

"All: I haven't seen any suggested changes to the proposal in a while. Are we happy with that part? For me, I'm pretty satisfied -- I think we've done a great job on it. Let's take a quick poll and see if there are other outstanding issues."

Sorry, RL grabbed and I just got back!

I have a mod to the proposal, mostly polishing it up. We want this to be really good language and procedure to, hopefully reduce on-the-fly editing at the meeting.

Can you give me until tomorrow morning, please? I still have a RL issue to deal with.

#211 ::: Keith "Kilo" Watt ::: (view all by) ::: May 28, 2015, 01:06 PM:

[FAQs]

Here's the latest FAQ section:

FAQ’s:
1. Can you explain the system in plain language?
The important thing to remember is that nothing changes in how you nominate. If you think a nominee is Hugo-worthy, then nominate it. That’s all. There’s no need to rank your choices at the nomination stage, and there’s no good reason not to nominate something you think even might be Hugo-worthy. All we are doing at this stage is throwing names into a hat. The final Hugo voting system, which actually chooses the winner, is unchanged. We could, in theory, simply put everyone’s nominations on the final ballot, but that would make for a very long ballot indeed. We therefore need to narrow the nomination list down. This system narrows down the list by eliminating the least popular nominees until only five (under current rules) finalists remain. Here are the basic steps to the elimination process:
a. You have one nomination “point” for each category that will be divided equally among the nominees you choose to nominate in that category. So, if you nominate two nominees in a category, each will get half a point; if you nominate three nominees, each will get one-third of a point, and so on. (We call this the Calculation Phase.)
b. All the points given to each nominee from all nomination ballots are added together. The two nominees that got the least number of points are eligible for elimination. One of these nominees is the least popular and will be eliminated. (We call this the Selection Phase.)
c. To determine which of these two nominees is least popular, we compare the total number of nominations they each received (that is, the number of nomination ballots on which each nominee appears). The nominee that received the fewest total number of nominations is the least popular and now completely vanishes from the nomination process as though it never existed. (We call this the Elimination Phase.)
d. We start over for the next round and repeat the process, however, if one of your nominees was eliminated, then you now have fewer nominees on your nomination ballot. This means that each nominee gets more total points, since you aren’t dividing your point among as many nominees. For example, if one of your five nominated nominees was eliminated, your remaining nominees now get one-fourth of a point each instead of one-fifth of a point. If four of your nominated nominees are eliminated, your remaining nominee now gets your full point.

2. So what kinds of strategy should I use when nominating? Shouldn’t I just nominate one work if I want to give it the best chance to win?
Under SDV-LPE there’s no good reason to use any strategy at all. The system was crafted specifically to make most strategies irrelevant. For example, suppose you wanted to give your full support behind one nominee that you really feel strongly about. You can do that, particularly if there isn’t anything else you feel is Hugo-worthy. On the other hand, you generally won’t hurt your candidate by nominating something else that you feel is worthy. If one nominee gets eliminated, then your full support will go towards the remaining nominee, just as if you had never nominated the others. In general, the best strategy is simple: nominate as many nominees as you feel are worthy.

3. Why are there Selection and Elimination Phases? Couldn’t you simplify it by doing it all in one step per elimination?
The Selection Phase looks for nominees with the fewest points, and the Elimination Phase eliminates the one with the fewest nominations. Looking at the two different measures of popularity adds robustness to the system. With non-slate nominations, as occurred with most of the Hugo Awards to date, the nominees with the fewest points will tend to be the same ones as those with the fewest number of nominations. But while it might be possible for strategic blocs to manipulate one of these two measurements (as occurred in 2015), it would be very difficult to manipulate both in the same direction. A strategy that focuses on nominations for the Elimination Phase (such as bloc voting) will be stymied by the Selection Phase; a strategy that focuses on points for the Selection Phase (such as nominating only one nominee) will turn out to be counterproductive in the Elimination Phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.

4. E Pluribus Hugo?
The designers felt this name accurately reflects what the Hugo nomination system should be trying to achieve. Fandom is the Many. Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in SF. This is a defining characteristic of fandom, and this system embraces it. Out of all of these many tastes in nominees, this system allows a single set of nominations to slowly emerge as the most popular candidates for the Hugo Award: “Out of the Many, a Hugo.”

5. But surely that’s not what the system is formally called?
Okay, sure. Formally, in the context of election theory, this system would be called “single divisible vote with least popular elimination” or SDV-LPE for short. E Pluribus Hugo is both cooler and geekier, as everything is better with a little Latin.

6. How does this system eliminate slate or bloc voting?
It doesn't, exactly, nor should a nominee be automatically eliminated just because it appears on a slate. On the other hand, any slate which nominates a full set of five nominees will find that each of its nominations only count 1/5 as much. With “non-slate” nominating, some of your nominees will be slowly eliminated, so your remaining nominees get more and more of your support. Since slate nominees tend to live or die together, they tend to eliminate each other until, in general, only one slate nominee remains. With a large enough support behind the slate (five times as much), the slate may still sweep a category; however, if that many voters support the slate, they arguably deserve to win, and no fair and unbiased system of nomination will prevent that. The answer in that case is, simply, to increase the general pool of voters. Regardless, with SDV-LPE, slates will never receive a disproportionate share of the final ballot, as occurred in the 2015 Hugos.

7. What if there are multiple slates (slate wars, “parties”, etc.)?
As with a single slate, the nominees on each slate will tend to eliminate each other until only the most popular one remains. The end result is that even multiple slates are unable to sweep the nominations.

8. What happens if a broadly popular nominee is nominated by a group of unrelated people?
If it is broadly popular, the system will still select that nominee for the final ballot.

9. What happens if a broadly popular nominee also appears on a slate?
Even if it is on a slate, if the nominee garners support from individuals then the system will select that nominee for the final ballot. In general, slates neither help nor hurt any given nominee.

10. Why re-run the system if someone declines a nomination? Why not just take the nominee that was eliminated in the next to last round?
We can do that, and in fact have an amendment for that ready if that’s what the people decide they want to do. The “re-run, merge, and extend” (RME) system described in the proposed language has only one major difference between it and taking the “6th place” nominee. Under RME, if a slate nominee makes the final ballot and that nomination is declined, the declined nomination will tend to be replaced by another nominee from the slate. Under “6th place”, if a slate nominee makes the final ballot and that nomination is declined, the declined nomination will tend to be replaced by a non-slate nominee. Either way works, but since our goal is for all sections of fandom to be fairly represented on the final ballot, RME seemed more in line with that philosophy.

11. Isn't it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws. The objective is to choose a system whose flaws are not in an area of concern to the electorate.

12. What are SDV-LPE’s flaws?
In rare cases, it is possible that eliminating both members of a tie could change the final ballot slightly from what it would be if the tie were broken in such a way that only one member was eliminated. The change is usually in the least popular of the finalists, and requires the two nominees in question be very close in popularity. This situation was extremely rare using realistic data (in fact in never occurred in any of our situations with realistic data), so the statistical probability of how ties are handled affecting the results is extremely small. There are a number of additional tie breakers that could be used if deemed it necessary in the future; however, simulations that have been run by the designers of this system show that it really shouldn’t be required.

13. What are SDV-LPE’s benefits?
Simply put, it reduces the power of bloc voting without eliminating the chance that nominees that do appear on slates will make it to the final ballot. Conversely, it makes it very difficult for slates to prevent non-slate nominees from appearing on the ballot.

14. Couldn't slates just recommend a single nominee for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate has enough supporters that is certainly a viable possibility – it’s also completely fair. It does not force all other nominees off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn't a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a nominee deserves a Hugo nomination, then it should, in fact, be represented on the final ballot.

15. What happens with a large field with no stand-out favorites when a slate votes?
Even in this case, we were unable to find a simulation in which no non-slate nominees appeared on the final ballot. Slate nominees did receive a larger proportion of nomination slots than they did otherwise, however, again, this could be considered a fair and valid result. If there was no general favorite, then members really had no collective preference.

16. How do the results of this system compare to the results under the current nomination system?
Statistical tests showed that this system and the current system overlap in about (conservatively) 4.5 out 5 candidates. In the absence of slates, it generally gives identical results to the current nomination system.

17. I think we should just increase the number of nomination slots on the final ballot to (for example 6), and decrease the number of slots a voter can vote for to a smaller number (for example, 4). Wouldn't that be simpler and easier?
Unfortunately, this simply means that the largest slate will receive four of the nominations and the next largest will receive the remaining two. It doesn't solve the problem of forcing nominees off the ballot that had a chance to win the final election. In general, we want fandom to nominate as many nominees as they feel are Hugo-worthy, since under SDV-LPE there is no strategic reason not to do so. For this reason, most of the designers of this system would prefer that members not be limited to nominating only four nominees. Keep in mind, however, that SDV-LPE will work with this (or most any other) change as well, so one does not preclude the other.

18. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
This could work. The problem is that now you have a small group of people who serve as literal gatekeepers to the Hugo Awards. In spite of the word on the Internet, this has never been the case in the past. Establishing it now means that those groups who believed it existed in the past will now be correct. Ultimately, human judgement is fallible. The fairness of a committee’s decisions will forever be subject to opinion. The end result is that the prestige of the Hugo Awards will forever be tarnished.

19. I think we should use [insert other mathematical voting system].
We considered essentially every applicable type of voting system currently in the literature, guided by two experts in the field. It should be kept in mind, however, that the goals and requirements for choosing a set of representatives in a political situation are different from those for choosing a set of Hugo finalists. Some of these systems do, in fact, have positive properties that speak for them. None of them were as simple or as intuitive as SDV-LPE, yet SDV-LPE meets all of the stated goals for a Hugo nomination system.

20. Won’t SDV-LPE be complicated to code and implement?
Actually, no. One of our non-experts coded a full simulator for the system in a matter of days. A full web-based app would not be much more difficult to handle.

21. Wasn't this system just designed by Social Justice Warriors to block the Good Stuff?
It is true that much of the discussion for this system occurred on Teresa and Patrick Nielsen Hayden’s “Making Light” discussion board. Other than serving as occasional moderators, however, TNH and PNH had no real input in the discussions of the system. Those of us who worked on the system were very clear that our goal was not to keep the Sad/Rabid Puppies off of the Hugo ballot, and that any system which specifically targets any type of nominee is inherently wrong and unfair. One of the members of the group is a retired US Naval officer, a combat veteran, a certified Navy marksman, a Christian, and considers Robert Heinlein to be the greatest science fiction author who has ever lived. In short, he is exactly the Puppies’ demographic. But any slate, of any sort, be it a Sad Puppy or a Happy Kitten of Social Justice, breaks the Hugo Award because a small percentage of voters can effectively prevent any other nominee from appearing on the final ballot. This is a major flaw in the Hugo nomination system, and it is a flaw that must be fixed if the integrity of the award is to be maintained. Politics should play no role whatsoever in whether a nominee is Hugo-worthy or not.

#212 ::: Keith "Kilo" Watt ::: (view all by) ::: May 28, 2015, 01:08 PM:

[PROPOSAL]
junego@210:

No worries, submitting your proposed changes tomorrow is fine.

[FAQs]
The one I'm still not sure if we should include or not is #21. I'd be interested in opinions on that one in particular...

Thanks,
Kilo

#213 ::: David Dyer-Bennet ::: (view all by) ::: May 28, 2015, 01:09 PM:

I like James Quinn's rocket silhouette made of titles design! And the successive versions are each improvements.

Also like JJs ribbon design. Including the second line of text helps make things clearer I think (what we're about). (And the rocket can easily be swapped back to an "i" if that's what we decide we want, doesn't seriously damage the design in my eyes at least.)

#214 ::: Keith "Kilo" Watt ::: (view all by) ::: May 28, 2015, 01:09 PM:

[SWAG]

Jameson: One thing to keep in mind about t-shirts is that printing on the front and back dramatically increases the cost. Personally, I think your #195 is awesome. I like the dots between the words, too.

Kilo

#215 ::: Keith "Kilo" Watt ::: (view all by) ::: May 28, 2015, 01:10 PM:

Seth@208:

Seth, what's your last name? We have two Seth's, so I want to make sure I get the right one. (I think I know, but never hurts to confirm.)

Thanks,
K

#216 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 01:10 PM:

[SHIRTS]

Let's say we get the go-ahead to use some version of the winner-list rocket. That is to say, let's say we get a list of things not to do, and the winner-list is not on that list.

With full respect to Keith, I think that since it was my idea and my design, and I would be putting up the initial money, it's OK for me to do this as a fundraiser. That is, I would charge some reasonable markup that would go 50/50 to the MPC and to Electology.org. (And I'd pay for my own shirt and Keith's, including the markup, and I'd take no cut myself as the organizer/designer). Then the questions are:

1. Lee: what is your estimate on a base and marginal per-shirt cost (ie: something like $X00 to create the design and then $X per shirt for one-sided one-colored, and for two-sided one-colored ..., and for one-sided two-colored ....)?

2. Who here would (probably) buy how many shirts? How many do people think could be sold on site? I would rather err on the low side if anything, so I don't get stuck with too many shirts; but not too low. (I'm guessing the price per shirt, including markup, will probably be under $20.)

Again, all of this is subject to veto by the MPC, but I think it's not unreasonable to make plans for what to do if we aren't vetoed.

#217 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 01:20 PM:

[FAQ] Looking pretty good. The one thing I'd add is something like @70, as an extra question after #2 ("Can you explain voting strategy in this system in more nerdy detail?")

#218 ::: David Dyer-Bennet ::: (view all by) ::: May 28, 2015, 01:42 PM:

Comments on the latest draft I see posted (May 27 04:54 pm):

3.8.8.1 specifically refers to "all of the nominations on that ballot in that category that were not eliminated in previous rounds", but that's slightly over-specific; 3.9.1 for example specifies another way that nominations can be eliminated. While the current language is clear enough to follow (the 3.9.1 language specifies it's doing something like virtual eliminations), I'd be happier if 3.8.8.1 wasn't overly specific in the first place; perhaps change "in previous rounds" to just "previously" in the bit of 3.8.8.1 I quote?

3.8.8.4: In "none of the nominees from that round shall be eliminated" the undefined concept "nominees from that round" is introduced. Since it's saying don't eliminate, it's not a really serious ambiguity; whatever it is you thought you might eliminate, don't! But still, perhaps change the last sentence to "If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then do not eliminate any nominees this round; all remaining nominees shall appear on the final ballot, extending it if necessary."

3.9: First, in the merge of the re-run finalist list with the original finalist list, I think you need to specify that the ineligible or declined nomination is removed from the original list. Otherwise it appears again in a strict implementation of the current language.

Second, we specify doing this when a finalist is deemed ineligible -- without specifying when that can happen. A finalist could be discovered to be ineligible after the final ballot was published, for example; are we intending to specify here how that must be handled? It could even be discovered after the voting deadline -- all votes are in, and somebody discovers an obscure prior publication that disqualifies a finalist. The issue of what to do when a work is withdrawn or disqualified after the ballot is published is difficult, and I don't think we want to accidentally constrain its handling (and I don't think we want to open the can of worms of how it should be handled; if anything that should be a separate change). (As shown this year, we can also face the issue of somebody who had previously accepted, or failed to respond, responding late and declining, or changing their mind and declining. Again, there's no real guidance in the rules for how to handle that, and again I don't think we should extend accidentally into that area.) Perhaps clarifying when our stated rules apply, plus adding a sentence saying roughly "If the nominee list needs to be revised at any later point, it should be generated by the method described in 3.9.1" or something? That limits how the list is generated while leaving whether it should be in unusual cases up to the admins.

Third, I'm fairly sure it matters whether we do multiple re-runs or make all the changes in a category at once. The intention as I understand it is that there will be an initial run, nominees will be contacted, and if one declines, the re-run and merge will immediately be performed and any new nominees contacted (and repeat until we have enough acceptances). Not that we might, say, get two "declines" on the same day and so we make only one re-run with both of those nominations removed. If this is correct, we should say so. (If my intuition is wrong about it mattering whether it's one or multiple sequential runs, never mind!)

#219 ::: Keith "Kilo" Watt ::: (view all by) ::: May 28, 2015, 01:46 PM:

Jameson@216:

I think that since it was my idea and my design, and I would be putting up the initial money, it's OK for me to do this as a fundraiser.

I am very, very uncomfortable with this for a number of reasons.

First and foremost, our goal is to get the proposal passed so that we can save the Hugo. I really do see it in those terms. I'm opposed to anything that works against that goal. We want to try to convince as many people as possible to support our proposal. Asking people for money to support our proposal is crazy, strategy-wise.

But I realize you aren't asking for money for the proposal, you're trying to raise money for Electology. I don't know anything about your organization, really. I'm sure they do fine work -- especially if you are the type of person they employ. But the whole idea behind the proposal is (to the greatest extent that we can) to remove the politics from the Hugo nomination process. This system was designed by members of fandom, though with the guidance of experts in the field. If the Worldcon members were under the impression that some outside political organization was pushing this proposal, then we're doomed. The fact is that this proposal was not designed by Electology. You did come up with the idea that led to SDV-LPE, but that was an individual and not as representative of your organization. There have been over 3000 messages of refinements and modifications to the system, so I think we're safe in saying that it was made by us and not Electology. We don't, however, want to give fandom the impression that it was otherwise.

I do recognize that it's your design and your idea, but because you are using the proposal's title, that means that there is an explicit support of everyone here for fund-raising for Electology. I would have no problem with you designing an Electology t-shirt and using that for your fund-raiser. But I'm not comfortable co-opting the proposal title in order to raise money for a political organization that has nothing to do with Worldcon or even SF.

So, in short, I'd support a general Electology t-shirt being used as a fund-raiser, but not a shirt related to the proposal. I think this would be a huge negative that we would have to work to overcome. We shouldn't make our job any harder than it has to be.

Kilo

#220 ::: David Dyer-Bennet ::: (view all by) ::: May 28, 2015, 01:51 PM:

FAQ 12: Currently says "... tie breakers that could be used if deemed it necessary..."; I think there's a word missing, perhaps "we" between "if" and "deemed"?

#221 ::: Ian Gillespie ::: (view all by) ::: May 28, 2015, 01:54 PM:

Kilo@100:

Thanks for your excellent reply Kilo. I was keeping in mind the re-adjustment of the weights, but I've now worked my way through a simulation based on your reply and have a better understanding of how this results in slated works being unable/less likely to dominate the final ballot. I also understand we're talking about the nomination process here, but I still feel the two sentences that I've flagged, as written, make the FAQ very confusing (i.e."nothing changes in how you nominate", "the final Hugo voting system... is unchanged").

Here are some ideas I would suggest for the FAQ.

1) Address the issue about what changes and what doesn't more simply and directly:

"E Pluribus Hugo is a system for counting Hugo nomination ballots that guarantees fairness for all involved and ensures a small number of ballots do not dominate all the nomination slots in a given category. Nothing changes about the way you vote, either at the nomination stage or on the final ballot - the only thing that changes is how nomination ballots are counted."

2) When explaining the system itself, eliminate all extraneous information (what one might do in theory). When discussing the mechanics, start with a high-level description and work your way down. Don't get into details such as "points" until you've explained the basic function. This will help easy the reader into the explanation, even for those quite familiar with complex multi-round voting systems.

I'd suggest something like:

"E Pluribus Hugo is similar to the process that's already used in the final round of Hugo voting - with multiple "rounds" of vote counting where the lowest placed finisher is eliminated in each round. The only difference here is that instead of using a ranked, preferential ballot, we use a two step elimination process."

"In the first step, called the 'selection phase', each vote is weighted based on how many other nominees are still in the running from that same ballot. If a voter has had four out of their five nominees eliminated before the current round begins, the full strength of their ballot counts for their one remaining choice, whereas the value of a ballot with five nominees still in the running will be divided between the five. This way, all BALLOTS count equally in each roudn of vote counting."

"The second step in this process, the 'elimination phase', is a good-old head-to-head run off between the lowest placing finishers (either the bottom 2, or several based on a tie). Here, the weights are discarded and the ballots are re-counted to determine the lowest place finisher. This ensures that no work is eliminated purely due to weighting against another work that got more votes."

"The end result of this system that BALLOTS ARE COUNTED EQUALLY in each round, rather than some ballots effectively accounting for five votes in a given round while other ballots account for only one."

3) When giving the more detailed desciption including how points are used, I would demonstrate this with a simply simulation such as:

1: S1, X, S3
2: S1, S2, Y
3: S1, S2, Z
4: A, B, C
5: D, E, A
6: F, G, B
7: H, I, J

The advantage of this particular simulation is that nominees D-J are eliminated in the first round.

1: S1
2: S1, S2
3: S1, S2
4: A, B
5: A
6: B

The slate contenders then effective have a run off amongst themselves where S2 is eliminated even though it got the same number of votes as A and B. The nominees wind up being A, B and S1.

4) When explaining how this system deals with slated, don't use the "strong" version of the question and then explain how it doesn't *eliminate* slate or blocing voting. Frame the question the way it should be framed in the first place and give a simple and direct answer.

"Q: Does this system prevent groups from using slates or blocing voting tactics to game the final Hugo ballot and effective exclude other nominees? How?"

"A: Yes. This system give BALLOTS equal value in the first step of the count process, thus ensuring a small group of ballots do not select all the nominees."

"All voters begin with the strength of their ballot divided between up to five selections. For voters who did not coordinate their choices with others, their less popular nominees will slowly be eliminated. As this happens, the value of their remaining votes will increase. Conversly, those voters who did coordinate their choices (i.e. slate voters), will have the value of each of their votes reduced until some of their selections are eliminated. When a group of ballots all features the same nominees, they will effectively contend against one another until enough of their nominees are eliminated and the weight of each remaining vote increases. Again, this ensures all BALLOTS are counted equally in each round."

#222 ::: Zack ::: (view all by) ::: May 28, 2015, 01:56 PM:

Extremely belated response to Tim Illingsworth @20: I'm not going to be able to attend Sasquan, so while I appreciate the offer to help me draft something, I couldn't take it through the rest of the process myself.

#223 ::: Seth ::: (view all by) ::: May 28, 2015, 01:56 PM:

#215 I'm Seth Breidbart.

Here's another issue: When are ineligible nominations removed? Suppose a number of nominations are made for an ineligible candidate (e.g. published in the wrong year); are they still counted, and if it gets into the top 5 then what? What about candidates ineligible under 3.8.5?

#224 ::: Zack ::: (view all by) ::: May 28, 2015, 02:12 PM:

While I'm here, have some editorial notes:

[FAQ]

The term "SDV-LPE" is used in item 2, but not defined until item 5. I suggest striking the two words "Under SDV-LPE" from the beginning of item 2 (so the first sentence now reads "There's no good reason to use any strategy at all.")

It might also be appropriate to edit out later uses of "SDV-LPE", since item 5 describes it as a technical term that is *not* considered the name of the proposal.

And I also suggest deleting "Actually, no" from item 20 on the grounds that this makes it sound less aggressive.

#225 ::: P J Evans ::: (view all by) ::: May 28, 2015, 02:23 PM:

223
The most obviously ineligible will be taken out right at the start, I would think. The edge cases - those are the ones that show up late, like some this year.

#226 ::: P J Evans ::: (view all by) ::: May 28, 2015, 02:24 PM:

219
THIS.

#227 ::: abi ::: (view all by) ::: May 28, 2015, 02:24 PM:

I'm with Keith @219.

Jameson, I was a little uncomfortable with you using the extra money from the fundraiser to attend Worldcon to go to Electology. I'd have preferred it went to an attend-Worldcon fund like Con or Bust. But I didn't contribute, and the contributors knew what would happen with the surplus when they made the choice to give.

This feels like yet another step toward this entire effort somehow being for the benefit of Electology. It's not. And I don't know enough about Electology to know whether or not I want to support it, and therefore I don't want my support for this proposal to be conflated with it.

I'd be keen to have an E Pluribus Hugo T-shirt. I really like your design. But I'm not going to buy one to support an organization I know nothing about, and which is not associated with the thing I am interested in.

Also, Teresa already had a conversation with you about using Making Light as a platform for promoting Electology, and you came very close to being banned for your boundary-nudging. You've since done the very best thing you could do for your organization in this context, by participating in the effort to refine this proposal. But that doesn't entitle you to springboard that into further efforts to promote Electology here on Making Light.

#228 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 02:27 PM:

@219: First off, thanks for a thoughtful and respectful response. I'm sure you realize that the respect is entirely mutual.

I disagree with you. But I won't go ahead with my plan unless I can get you to at least moderate your opposition. I do trust that you won't simply stand on principle for the sake of showing who's boss, and that you recognize that I've already shown flexibility in dropping the idea of a panel.

One point where I agree with you, and which I think will be a productive basis for this discussion, is: "First and foremost, our goal is to get the proposal passed so that we can save the Hugo.... I'm opposed to anything that works against that goal."

The difference is, I don't think that selling a quality T-shirt at a reasonable markup works against that goal at all. Consider:

1. We'd still have buttons and/or ribbons, with no money changing hands (except for those of us, myself included, who'd put up the printing costs).

2. I think this T-shirt would be an attractive design, and worth the price. I've also promised that I would buy one for you with my own money.

3. By pre-ordering a fixed amount, I'd be lowering the costs, and avoiding the markup from CafePress or whatever. However, I'd also be taking a certain amount of financial risk personally, as I might end up stuck with inventory.

4. I think that the two causes — the MPC and Electology — are both good ones, and reasonably related to the proposal and design. No, I'm not claiming that this proposal belongs to Electology, or to me myself, any more than I'm claiming it belongs to the MPC. But I think most people will understand the connection.

5. I think most people are accustomed to the fact that when they buy things, they're paying more than just the cost of materials. I don't think anybody would be offended by paying (say) $17 instead of $13 for a t-shirt, especially if they know that the extra money is going to a good cause. Some people even might even like it more. (And that includes some people who could, but probably wouldn't, pay $13 and donate $4. People aren't always rational economic agents, even in their giving.) And these days, something like $17 for a nice, well-made T-shirt is not at all out of line with the average price you'd pay.

6. Honestly, I think this design is good enough to be worth selling at the convention; that there are people who would want it, who wouldn't be plugged in to this blog enough to notice the chance to pre-order. However, that means I will have to take risk (as discussed in point 3 above) and also do the work of selling them. I think it's reasonable to expect some payoff — not personal, but in terms of resources going to an organization I care about and believe in — for that work. I also think that my presence, engaged in this activity, will tend to increase, not decrease, support for this proposal, and thus improve its chances of passing.

Again, I hear your point of view, and I'm going to give in to you unless you moderate your opposition. I'm also keeping an open mind (I hope), and so if you have counterarguments for the above, you may be able to convince me I was wrong. But I hope that I can convince you to agree with me, or at least to stand aside from disagreeing, so that I can proceed with this course (assuming of course that the MPC doesn't veto it.)

With great respect,
Jameson

#229 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 02:35 PM:

Just a note, that I wrote @228 before reading @226 and @227. If I had read those, I would not have posted it, at least not until receiving some show of support.

And in specific response to @227: Yes, I did get slapped down early on for promoting my organization. And I'm sorry. In my tepid defense, though, I also had at least two people taking my side; and I can also say that in the online communities where I hang out most, there are definitely moderators, but what I did would not have been outside the norms. I know: when in Rome. And so since those early missteps, I've tried to be more careful, and I hope I have shown myself to be a productive community participant.

#230 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 02:38 PM:

So, to be more explicit: I was willing to respectfully debate Keith, but I am not going to push the issue if it's just me against Keith, PJ, and Abi. This is the last you'll hear from me on that unless somebody else speaks up in favor of a (reasonable) markup.

#231 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 02:46 PM:

When I say "the last you'll hear from me on that", I mean, on the idea of having a markup. I'm still happy to discuss other T-shirt issues.

(sorry for the "and one more thing" posting.)

#232 ::: abi, who is also Idumea ::: (view all by) ::: May 28, 2015, 02:46 PM:

Jameson @229:
Just a note, that I wrote @228 before reading @226 and @227. If I had read those, I would not have posted it, at least not until receiving some show of support.

It's clear that we cross-posted. I get that.

And so since those early missteps, I've tried to be more careful, and I hope I have shown myself to be a productive community participant.

You have certainly been a productive community participant, and your contributions here are valued. You're welcome to continue to hang out here even after the completion of this proposal.

And @230:
This is the last you'll hear from me on that unless somebody else speaks up in favor of a (reasonable) markup.

Just to be clear, I'm speaking ex cathedra in this paragraph. I've touched bases with Teresa as well. Making Light is not a venue for Electology fundraising.

(climbs out of the chair) Also, although Making Light does not own E Pluribus Hugo in any way, shape, manner, or form, I support Keith's leadership decision in this matter.

#233 ::: Cassy B. ::: (view all by) ::: May 28, 2015, 02:51 PM:

I don't have a problem in principle with a small t-shirt markup, but I'd really expect the profits to go to benefit the Hugos Award in some way. Since helping the Hugos is the actual point of the t-shirt...

Per Tammy Coxen @151, the Mark Protection Committee would seem a logical choice. There may be others that are also logical. (Who maintains the Hugo website?)

No offense intended, Jameson, but I know nothing other than what you've said of your organization, and I'm uncomfortable giving money to a charity that I haven't personally vetted. However worthy it may be.

If the shirt is sold at cost, OR if any profits go to a Worldcon and/or Hugo benefitee, then I'll buy one. (I really DO like that rocket-made-of-titles. Nice work.)

#234 ::: Soon Lee ::: (view all by) ::: May 28, 2015, 02:54 PM:

@Jameson Quinn,

You've accrued a lot of good will with your valuable contributions here, which is why I was happy to contribute to your GoFundMe. I don't have a problem with excess from the GoFundMe going to Electology.org either, for the same reason.

But I'm not comfortable with your new idea for the reasons Keith articulated @219.

#235 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 03:01 PM:

@232: I'm tempted to get a bit defensive about the "ex cathedra" statement (I don't think that sharing designs and respectfully discussing possible plans constitutes fundraising, and as for the Gofundme, I explicitly got permission to post about that), but really there's no point; I've dropped it. Respect; you moderators have a tough job and you clearly do it well.

#236 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 03:09 PM:

Obviously, I'm not going to respond to @233 ("but I know nothing other than what you've said of your organization, and I'm uncomfortable giving money to a charity that I haven't personally vetted") by posting a bunch of links promoting the organization. Equally obviously, anybody who wants such links can contact me privately and I'd be glad to oblige.

#237 ::: abi ::: (view all by) ::: May 28, 2015, 03:10 PM:

Let's move on, shall we?

#238 ::: Soon Lee ::: (view all by) ::: May 28, 2015, 03:17 PM:

[FAQ]

I like Seth's addition @67 which was revised by Jameson Quinn @70. I think including it in or just after Question 2 be good.

Re:more nerdy detail
Would it be useful to provide references to SDV-LPE at the end of the FAQ? I'm thinking of papers in peer-review journals (if appropriate ones exist) for those academically inclined.

Visualization:
Is there one available? I'd be interested in one that can show how *my* ballot would be treated under this new system compared with the existing one.

#239 ::: Chris Battey ::: (view all by) ::: May 28, 2015, 03:36 PM:

[RIBBONS]

However complex the shirt design ends up being (I still intend to give the seal-based idea a shot, maybe this evening), I think the ribbons will work best as text only.

We can get slightly larger text on two lines as in http://www.ribbonsgalore.com/opendesign/14046, or slightly smaller text on one line as in http://www.ribbonsgalore.com/opendesign/14047. Anyone have particularly strong opinions?

Either way, I'll do a full template design rather than using their easy-design interface, to make sure we get the best use of the space (their interface doesn't seem to support "make the text as large as you can to fit the space available").

1000 ribbons will cost $180; 500 ribbons will cost $105. (The next volume-purchase rate drop is at 2500 ribbons which lowers the cost to $0.17 each; I'm guessing that's too many.)

I'm comfortable making the initial outlay myself, and then anyone else who's interested in contributing to the cost and distributing ribbons can meet me at the con (probably Wednesday night) and pay me $18 or $21 for 100 ribbons to distribute. (Or $9-$10 for 50 if you'd prefer.)

So who's interested in buying/distributing ribbons? (And how many - 50, 100, or more?) I'll size the order based on interest.

#240 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 03:42 PM:

@239: Put me down for 150. And I agree: text alone is best for ribbons, though we should coordinate fonts with the other swag.

#241 ::: Keith "Kilo" Watt ::: (view all by) ::: May 28, 2015, 03:54 PM:

Honoring abi's recommendation to move on, I did just want to reiterate how much I respect and esteem Jameson personally and am especially grateful for his contributions. He's been invaluble to the discussions.

Kilo

#242 ::: Keith "Kilo" Watt ::: (view all by) ::: May 28, 2015, 03:56 PM:

Chris@239:

I'll take a 100 for sure; hopefully we can find you at the con if we run out, yes?

K

#243 ::: JJ ::: (view all by) ::: May 28, 2015, 04:49 PM:

Chris, #239: So who's interested in buying/distributing ribbons? (And how many - 50, 100, or more?) I'll size the order based on interest.

I'm in for 100.

#244 ::: ULTRAGOTHA ::: (view all by) ::: May 28, 2015, 05:02 PM:

#239 - Chris Battey

[RIBBONS]

So who's interested in buying/distributing ribbons? (And how many - 50, 100, or more?) I'll size the order based on interest.

I will buy 50. We are staying at the Doubletree and will be at the Preliminary Business Meeting (and at the Bujold meet-up which I think is the day before).

If lots of people agree to buy ribbons, I don't think we should all buy if it adds up to more than whatever the optimal number is (1000?). So drop me off the list if we get too many. In that case, I'll just donate to Chris.

#245 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 05:22 PM:

I think 1000 sounds about right. That is, 800 might be too few.

(Note: When I said "my idea" @216, I meant the T-shirts, not the proposal. The proposal is ours, not mine.)

#246 ::: Adam Tilghman ::: (view all by) ::: May 28, 2015, 06:02 PM:

Chris @239 - I'm in for 100 ribbons at-con.

#247 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 06:23 PM:

I note that my design looks better in Inkscape than in the browser. Which means that when I'm done tweaking it, I can convert text to paths, and the silhouette will be even smoother than what you see now.

#248 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 06:29 PM:

So, how do others feel about what Cassy@233 says about markups for MPC only, or for MPC and Con or Bust? I have an opinion there, but don't want to discuss it if it's a dead horse.

#249 ::: Chris Battey ::: (view all by) ::: May 28, 2015, 09:08 PM:

[RIBBONS]

Jameson @ 240: Agreed; we don't need to lock the design until mid-July at the earliest, so we have time to make sure everything coordinates well.

Keith @ 242: Absolutely. I'll probably have a few extra batches just in case, particularly since reaching 1000 doesn't look like it'll be a problem.

ULTRAGOTHA @ 244: There isn't really an "optimal" number except for the points where it becomes cheaper to bump up your order to the next pricing tier (e.g. 850 ribbons costs about the same as 1000, so it's not worth ordering 875-975 - just get a thousand). As another example, 1200 ribbons isn't priced as 1000 ribbons plus 200 ribbons (at those respective rates), it's just 1200 at the rate that applies to any order in the 1000-2475 range.

[SHIRTS]

I would rather avoid explicit fundraising at all if we can, even for the MPC; selling Hugo-related shirts with proceeds earmarked for the MPC would seem to put the MPC in a position to profit from the popularity of our proposal, which looks like a gigantic conflict of interest to me, and having any other organization financially tied to the proposal causes similar problems. I appreciate the desire to try to leverage the economics of shirt-buying in a positive way, and I think if we happen to end up with a small profit it's okay to donate it to the MPC, but I'm really worried about the message this ends up sending.

#250 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 09:20 PM:

It's not so much the economics of shirt - buying that I want to leverage, as the activity of shirt - selling. Without a markup, inventory is too risky, so we can't have extras to sell at the con; and selling is a good way to meet people and discuss the proposal.

#251 ::: Chris Battey ::: (view all by) ::: May 28, 2015, 10:07 PM:

Selling at the con also requires managing all the other logistics that go into selling anything at the con. Are we planning on purchasing table space? Banners? Who's staffing the table? How are we handling payments?

And inventory is risky in any case, unless you're only printing as many shirts as you can afford with the advance orders, which puts a pretty hard cap on how much inventory you can carry at all.

Personally, I'd rather just use something like Teespring to set up a preorder campaign that will close in time for everyone to get their shirts in advance of the convention (and sell as close to at cost as possible), and then use interest in the t-shirts ("I've seen several of those shirts around - what does it mean?") to drive conversation along with badge ribbons, buttons, etc.

We hand out the ribbons and buttons for free - we're putting our own money where our collective mouth is - and nobody has to feel awkward about a literal please-give-me-money sales pitch getting bundled up with (or worse, conflated with!) the please-vote-for-E-Pluribus-Hugo campaign pitch.

#252 ::: Jameson Quinn ::: (view all by) ::: May 28, 2015, 10:16 PM:

I'm a bit of an introvert here, so when you see "daunting logistics" I see "comfortable structure for interacting". But points taken.

#253 ::: Chris Battey ::: (view all by) ::: May 28, 2015, 11:05 PM:

As an extrovert with social anxiety, I'm really sympathetic to the notion of a comfortable structure for interacting. In my case, the idea of someone approaching me to ask me about my shirt and giving me the opportunity to respond with "It's for a proposal we have before the Business Meeting to change the Hugo nomination process..." fits my comfort zone a lot better than trying to sell people things.

And for the same reason, the notion of someone confusing my philosophical interest in getting the proposal passed with a pecuniary interest in selling t-shirts fills me with dread. A fear of having my motivations misunderstood is one of the primary ways in which my anxiety manifests. I also suspect that there will be people who not only are vehemently against this proposal (or any other attempt to address the slate-voting), but will also take any opportunity to smear both the proposal and its proponents; I don't want to give them anything they could use to argue against the proposal's legitimacy.

In any case, while I think my overall points about logistics and about keeping the proposal free of monetary concerns are still valid, I acknowledge that part of my negative reaction to the fundraising ideas are based in my anxiety. Other opinions from the rest of the group would be (as always) very much appreciated...

#254 ::: JJ ::: (view all by) ::: May 28, 2015, 11:46 PM:

Chris Battey, #253: And for the same reason, the notion of someone confusing my philosophical interest in getting the proposal passed with a pecuniary interest in selling t-shirts fills me with dread. A fear of having my motivations misunderstood is one of the primary ways in which my anxiety manifests. I also suspect that there will be people who not only are vehemently against this proposal (or any other attempt to address the slate-voting), but will also take any opportunity to smear both the proposal and its proponents; I don't want to give them anything they could use to argue against the proposal's legitimacy.

This. I would much prefer that we set up a way to pre-order the t-shirts, and publicize that to anyone who wishes to order one.

Taking money at the con for anything related to E Pluribus Hugo, even if it's selling t-shirts for cost, seems to me to be very wrong -- and if I perceive it that way, how will the uninformed con-goer perceive it?

#255 ::: Kevin Standlee ::: (view all by) ::: May 28, 2015, 11:48 PM:

Keith "Kilo" Watt @201:

1) Will there be any kind of A/V equipment at the business meeting? PowerPoint, etc.?
Yes. We've been told by Programming that we will have access to a data projector. The Parliamentarian, Donald Eastlake, will be responsible for managing this. If you end up with a PowerPoint, we'll want you to submit it in advance so that Don knows it works with his equipment. What I don't want is you showing up with a computer that you want to try to plug into the equipment there. Something I hate is meetings being stalled by people fussing with equipment. That includes showing up with a USB stick expecting to plug it into Don's machine during the meeting. If you work witht he Parliamentarian in advance, you'll be much better off.

2) Once we submit the proposal to the address you provided, is it locked in or can it be modified before the deadline?
Any proposal may be withdrawn or modified until the deadline. After the deadline, it doesn't "belong" to the makers anymore, but to the meeting as a whole. Thereafter, it would require a suspension of the rules to withdraw the proposal, and all other changes are amendments, even so-called "friendly" amendments. (There's no such thing as a "friendly amendment," but people who have learned sandlot parliamentary rules think there is.) So yes, it's better to submit it sooner than later IMO.

Cassy B. @233:

Per Tammy Coxen @151, the Mark Protection Committee would seem a logical choice. There may be others that are also logical. (Who maintains the Hugo website?)
The MPC pays for the domain registration and hosting for TheHugoAwards.org, Worldcon.org, and WSFS.org. FWIW, donations to the MPC may be tax-deductible charitable donations on US taxes. (Details upon request.)

TheHugoAwards.org is managed by the Hugo Awards Marketing Committee, which the WSFS MPC appoints to run the site. In fact, of late I make nearly all of the edits myself as I'm one of the HAMC members with keys to the site.

Regarding shirts, inventory, etc: Why not use something like CafePress? You can sell shirts at zero markup if you like.

I'm waiting for more members of the WSFS MPC to weigh in until I say more about specific designs, but I will share the current proposal with them. (You can't expect them to be following this discussion.)

[In Pendleton OR tonight after about ten hours driving from Winnemucca. On to Spokane on Friday.]

#256 ::: jonesnori/Lenore Jones ::: (view all by) ::: May 29, 2015, 12:29 AM:

[SIGNING] and possibly [ATTENDING], Lenore Jean Jones. I'll be at the con and hope to make the meetings. Very good work, y'all.

#257 ::: rnv ::: (view all by) ::: May 29, 2015, 01:19 AM:

Long-time listener, first-time caller. I've only recently started reading these threads closely, since I didn't feel I'd have anything to add or contribute. But I do have a suggestion for FAQ #18.

Shifting the answer into the conditional might help clarify that it's addressing the consequences of establishing a committee, rather than describing an actual current state of affairs. Maybe it's a small thing, but I'm concerned that as it stands, it's the sort of statement, especially if quoted out of context, that could be problematic.

It could be modified something like this:

This could work. The problem is that now then you would have a small group of people who serve serving as literal gatekeepers to the Hugo Awards. In spite of the word on the Internet, this has never been the case in the past. Establishing it now means that would prove those groups who believed it existed in the past will now be correct. Ultimately, human judgement is fallible. The fairness of a committee’s decisions will would forever be subject to opinion,. The end result is that and the prestige of the Hugo Awards will would forever be tarnished.

#258 ::: Cally Soukup ::: (view all by) ::: May 29, 2015, 01:24 AM:

rnv@257: Oh, very good catch! Let's hear it for fresh eyes! At this point, most of us have looked at it so often that we know what was meant rather than what was written....

#259 ::: Sora Battey ::: (view all by) ::: May 29, 2015, 02:11 AM:

Hello, all. I'm the wife of Chris Battey (see comment #19 and others). He asked me to come take a look and put a fresh pair of eyes on the proposal and the FAQ, since I haven't been following the discussion here, but have been reading a bit about the Slate Situation as a whole (and, of course, will be at Sasquan as well). And in case anyone needs it for tallying purposes, I'll be [ATTENDING] the meeting, too. (It seems ridiculous, however, to put my name up as a signee, though I'll certainly be voting for the proposal; I haven't earned any such thing.)

So, having read through both the proposal and FAQ in the Original Post and the Reformatted Proposal in #121 and Revised FAQ at #211, here are my thoughts:

[PROPOSAL]

Overall, I really like the reformatted proposal in #121. It's clear, succinct, and easy to read. The one suggestion I would make is as follows:

3.8.8.4: These rounds continue until

For absolute clarity, it might be good to expand this phrase something more explicit, e.g.:

The phases described in 3.8.8.1, 3.8.8.2, and 3.8.8.3 are repeated in order until

The meaning of the word "rounds" and what a "round" consists of is implied in 3.8.8, but not defined.

[FAQ]

All question numbers refer to revised FAQ numbers from post #211.

First and foremost, I agree with Zack @ #224 that the term "SDV-LPE" should be avoided in the FAQ as much as possible, under the assumption that the FAQ is largely for people who have not been hip-deep in the discussion defining the system on this blog. It is not an intuitive term, nor does it relate directly to the name of the proposal, and adding a large number of new acronyms to the FAQ makes it feel jargon-heavy and confusing.


Q1

there’s no good reason not to nominate something you think even might be Hugo-worthy.

"Even might" reads awkwardly. "Might" alone conveys the same idea without the awkwardness; the emphasis "even" implies isn't necessary.

I would strongly suggest adding a "concrete" example here, as well -- whether you're speaking of Alice, Bob and Carol nominating "Rocket", "Castle", "Tentacles", "Swords", "Blasters", and "Incantations" (or what have you), or using real data from past Hugo nominations (though I'd suggest making sure it's from a couple decades ago, so there's no implied contentiousness).

The explanation given here makes sense to people who already understand what's going on -- but for people who haven't been here and don't deal well with abstractions, giving them a "real" example will help their understanding immensely.


Q10

This question is vital, but the answer is unnecessarily confusing due to the addition of new terms that are used only here and then discarded.

Existing answer:

We can do that, and in fact have an amendment for that ready if that’s what the people decide they want to do. The “re-run, merge, and extend” (RME) system described in the proposed language has only one major difference between it and taking the “6th place” nominee. Under RME, if a slate nominee makes the final ballot and that nomination is declined, the declined nomination will tend to be replaced by another nominee from the slate. Under “6th place”, if a slate nominee makes the final ballot and that nomination is declined, the declined nomination will tend to be replaced by a non-slate nominee. Either way works, but since our goal is for all sections of fandom to be fairly represented on the final ballot, RME seemed more in line with that philosophy.

I'd suggest, with the paragraph breaks for readability:

We can do that, and in fact have an amendment for that ready if that’s what the people decide they want to do.

Our system has only one major difference between it and taking the "6th place" nominee. Under our proposal, if a slate nominee makes the final ballot and that nomination is declined, that nomination will tend to be replaced by another nominee from the slate. In taking the "6th place" nominee, the declined nomination will tend to be replaced by a non-slate nominee.

Either way works, but since our goal is for all sections of fandom to be fairly represented on the final ballot, our existing proposal seemed more in line with that philosophy.


Q15

This is an odd case where the answer makes more sense to me than the question; I had to basically back-solve what the question meant based off of how it was answered.

I'd suggest changing

What happens with a large field with no stand-out favorites when a slate votes?

to something more like

What happens when there are a lot of nominees with no obvious favorites and slate votes are introduced?

#260 ::: Chris Battey ::: (view all by) ::: May 29, 2015, 03:03 AM:

Sora @ 259: Welcome to the thread, love. Thank you.

[SHIRTS]

So I finally got some time to try sketching out my t-shirt idea; here's a draft of the work in progress. I'm envisioning it as white ink on a black shirt. The blue lines were sketched in at the end of the evening (as I realized I wouldn't have time to finish the art tonight), but I'm going to take another pass at some of the white lines as well.

It turns out the image of the US Seal already includes an "E Pluribus Unum" banner, so modifying that (as well as emphasizing it) shouldn't be too hard (though at the moment it's just a placeholder). I replaced the eagle with a phoenix (representing fantasy, as well as the rebirth/renewal inherent in any significant change); the phoenix is clutching a stack of ballots and a Hugo rocket (based on, but not actually reusing, the Hugo trademark itself). Above the phoenix, at the moment, I have a ringed planet (one of the traditional icons of science fiction), but that's still a preliminary idea.

I'm also still trying to decide what to put on the shield (or other device across the phoenix's chest); the US seal has a shield with a pattern reminiscent of the American flag, but I'm hoping for something a bit more symbolic. One idea I considered was a book (or a stack of books), but I don't want to imply that Best Novel is all there is to the Hugos. But a whole stack of mixed media (books, magazines, DVDs, audio?) would be hard to represent...

Anyway - I'd welcome any suggestions on the design, particularly for the spaces above the phoenix and on the central device.

#261 ::: JJ ::: (view all by) ::: May 29, 2015, 03:31 AM:

Chris Battey, #260

You've worked really hard on this, and I appreciate your efforts. But, and this is just my opinion, I think that shirt design is way too busy. (It's also way too U.S.-centric.) Honestly, I wouldn't be willing to shell out for one -- because I think that the busy-ness of the design would get in the way of the shirt's primary purpose: to be a clear promotion of the voting proposal, one that will catch people's eye and maybe get them to ask about it (but at the very least, they should be able to clearly tell for sure what the shirt says).

I'd be happy just for clear text in a nice font, maybe inside a horizontal rocket facing to the viewer's right (something like this).

Also, if we do CafePress or a similar online shop, people will be able to pick their own t-shirt colors, and so you want 2 versions of the design: all-white, and all-black, and whichever contrasts best goes on whichever color of shirt.

Again, this is just my personal opinion, for what it's worth.

#262 ::: David Langford ::: (view all by) ::: May 29, 2015, 05:27 AM:

[SIGNING] if it's any help -- I'm not sure whether sheer numbers eventually become counterproductive? Can't make it to Sasquan, though.

#263 ::: David Harmon ::: (view all by) ::: May 29, 2015, 06:29 AM:

Jameson Quinn #250: I won't be at the actual con, but I'd strongly agree with Chris (and I'd guess some of the prior commenters), that putting "give me money" anyplace within earshot of "here's our proposal for the Meeting" would be a Bad Idea.

Anyone who wants to implement T-shirts for this year's Worldcon, needs to be able and willing to cover a potential loss of money in the cause of the proposal (in other words, they need to be a patron). As soon as you've "gotta make enough to cover the losses", that's already a sale motive and a distraction from the project. That said, this is a multi-year project, which means that shirts not sold this year can be brought to next year's con.

My advice is to do a small batch for the people who are willing to pre-order and pay for them from here, especially folks who'd be wearing them to the con. That first batch ends up as "I was there first" trophies for us, but a generous patron could provide extras to be sold at cost, or given away for publicity, at the con.

The response to "hey what's that shirt mean" needs to be "it's for a proposal for the Business meeting, here's our pamphlet describing it". Only if someone asks "where can I get a shirt", do they get pointed to a dealer's table or a POD link. Barring a heavier layout by that patron, a POD setup is surely what we want for the amorphous "second batch" of shirts, which may or may not be identical to the "I was there first" shirts.

#264 ::: David Harmon ::: (view all by) ::: May 29, 2015, 06:48 AM:

Chris Battey #260: I'm fond of the phoenix archetype myself, but I suspect that representing it in monochrome might be intrinsically iffy.

If that can be managed, I would say ditch the shield/icon entirely, and probably the planet too. That would go some distance towards cleaning up the "busy-ness".

While I sympathize with concern of "US-centrism", we're already riffing off the same source for the title, and stripping decorations as above will help distance it from the original seal.

#265 ::: Jameson Quinn ::: (view all by) ::: May 29, 2015, 08:28 AM:

Clearly, I should bow to the experience of people who actually do cons.

So the options are: pay a community member for a fixed number of shirts, which should be most or all preorders; or go with a one-at-a-time site like cafepress. Since my instincts here seem unreliable, I won't express an opinion on that question. Either way, I've already offered to pay for Keith's shirt and my own, and that offer still stands.

If we're going with a fixed number of shirts, I think we should have at least a few to sell at the con; even without the slightest amount of hawking them, I think there will be some number of people committed enough to find out where and when they have to be with cash in order to get a shirt, so if we have fewer than that number, all the shirts will go. In that case, I am willing to help put up the initial money, to be paid back by those "sells itself" sales.

As for the design: I like my rocket, but am not opposed to the idea of a phoenix. I agree that phoenixes are hard to do in monochrome — red and gold/yellow are the obvious colors. And if we're doing a two-color design, solid colors may be easier; and once you're thinking about a bird made of a limited number of solid colors, Northwest formline art suggests itself. I am certainly not up to doing that, but I'd be interested to see it attempted.

In the mean time, I think the draft of the eagle is well-drawn, but I agree that it's still too busy. My Guatemalan wife tells me to say that it's also "too imperial, but then, a lot of science fiction is imperial, isn't it". Now she's playing with the puppy and saying "quieren ser muy imperiales, quieres ser muy imperial?" in her puppy voice.

#266 ::: Jameson Quinn ::: (view all by) ::: May 29, 2015, 08:34 AM:

Continuing the riff on the phoenix thing and my family: I know that if I asked my daughter, she'd like the idea of a Gryffindor-type shirt. But honestly I see this proposal as more of a Hufflepuff or Ravenclaw solution; not heroically fighting the bad guys, but quietly doing the work and being wonky.

#267 ::: junego ::: (view all by) ::: May 29, 2015, 09:32 AM:

[PROPOSAL]

Long post, sorry. I ended up just re-typing the whole proposal because the corrections, suggested amendments, and my own proposal were too extensive. I tried to avoid this, thinking "Is this just rules wankery?', but realized we really do need the constitution part of the proposal to read as precise, clear and consistent as possible. So here it is.

If someone wants me to put up a sausage-making post showing the exact before/after/why, let me know.

1) I went over the existing constitution more carefully and reformatted some of the sub-sub-sub-sections to conform to using (1) instead of 3.8.8.4.1, etc. See sections 1.2, 1.8.1, 2.1, etc. of the current constitution.

2) I reconciled some of the language inconsistencies that had slipped in...shall/should/will, nominees/nominations, points/total points, etc. and, again, tried to conform to current usage.

3) I reformatted the beginning of sub-section 3.8.8 so that the terms 'point', 'points', 'nominations', 'round' are more clearly defined and consistently used within 3.8.

4) Kilo, I changed the end of a round back to the calculation phase, sorry. Incorporating the suggestions of David Dyer-Bennet on May 28 and my own fix for the definition of terms and adding a preliminary survey to set-up the first round made having the calculation part come next just too clunky. Leaving the defs and the preliminary info within Calculation was also too clunky. (Some of that probably *is* rules wankery 😆. Blame it on the Gov't for conditioning me with 50 page specifications for an electrical plug (for example)).

5) I clarified the issue of replacing ineligible or non-accepting finalists (per David Dyer-Bennet's suggestion) and moved it to within sub-section 3.8 for clarity. I also left the Worldcon Committee the same discretion they have now about whether or not to replace finalists under these conditions (which they used this year by closing the ballot.) I also clarified when they must rerun and merge the finalists (instead of just plain rerunning the tally when a work is found ineligible -before- anyone is notified of being a finalist) by using 'acknowledged" in 3.8.8.7.

Kilo, I would be happy to take-up keeping track of and incorporating suggestions/changes to the official constitution language if it would help you.


======================================
Short Title: E Pluribus Hugo (Out of the Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations), section 3.9 (Notification and Acceptance), and section 3.11 (Tallying of Votes) as follows:

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees finalists receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the process described in sub-section 3.8.8.

[The intervening sub-sections are not shown as there are no proposed changes.]

3.8.8: Finalist selection for each category shall consist of a preliminary survey of nomination ballots followed by a multi-round three phase elimination process as described in this sub-section.

3.8.8.1: Preliminary Survey: All eligible nomination ballots shall be surveyed to produce two criteria, nominations and points, for all nominees. These criteria shall be used in the Selection, Elimination, and Calculation Phases described in 3.8.8.2, 3.8.8.3, and 3.8.8.4 to determine the finalists in each category. The methods for generating these criteria shall be:

(1) Nominations: The total number of nominations for each nominee in a category shall be tallied from all ballots.

(2) Points: A single point shall be assigned to each category of each ballot. That point shall be divided equally among all nominees in that category. All points from all ballots shall be totaled for each nominee in a category.

3.8.8.2: Selection Phase: Each round shall start by selecting the two nominees with the fewest points for the Elimination Phase (3.8.8.3). (See 3.8.8.5 for ties.)

3.8.8.3: Elimination Phase: Nominees chosen in the Selection Phase shall be compared. The nominee with the fewest nominations shall be eliminated and removed from all ballots for the Calculation Phase (3.8.8.4) of all subsequent rounds. (See 3.8.8.5 for ties.)

3.8.8.4: Calculation Phase: Those ballots that contain eliminated nominee(s) from the Elimination Phase shall have their one point redistributed among their remaining nominees (if any) in that category and all points shall be recalculated for all nominees. The next round shall then start at 3.8.8.2.

3.8.8.5: Ties shall be resolved as follows:
(1) During the Selection Phase if two or more nominees are tied for fewest points, all such nominees shall be selected for the Elimination Phase.

(2) During the Selection Phase if one nominee has the fewest points and two or more nominees are tied for second fewest points, then all such nominees shall be selected for the Elimination Phase.

(3) During the Elimination Phase if two or more nominees are tied for the fewest nominations, the nominee with the fewest points at that round shall be eliminated.

(4) During the Elimination Phase if two or more nominees are tied for both fewest nominations and fewest points, then all such nominees tied at that round shall be eliminated.

3.8.8.6: Rounds shall continue until the number of finalists specified in 3.8.1 remain. If an elimination round would reduce the number of finalists to fewer than that specified in 3.8.1, then none of the nominees in that round shall be eliminated and all remaining nominees shall appear on the final ballot, extending the ballot if necessary.

3.8.8.7: After the initial Award ballot is generated, if any finalist(s) are removed for any reason and the Worldcon Committee determines to replace such finalist(s), the nomination tally system (sub-section 3.8) shall be rerun as though the removed finalist(s) had never been nominees. None of the remaining acknowledged original finalists shall be removed as a result of this rerun. The new finalist(s) shall be merged with the original finalists, extending the ballot if necessary.

Section 3.9: Notification and Acceptance.
3.9.1 Worldcon Committees shall use reasonable efforts to notify the nominees finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each nominee finalist shall be asked at that time to either accept or decline the nomination. If the nominee finalist(s) declines the nomination, that nominee finalist(s) shall not appear on the final ballot. The procedure for replacement of such finalist(s) is defined in sub-section 3.8.8.7.

Section 3.11: Tallying of Votes.
3.11.4: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. During the same period a record of all rounds of the selection process for each category shall also be published.

#268 ::: junego ::: (view all by) ::: May 29, 2015, 09:56 AM:

Sora Battey @259

[PROPOSAL]
[SEE JUNEGO @267]

Overall, I really like the reformatted proposal in #121. It's clear, succinct, and easy to read. The one suggestion I would make is as follows:
3.8.8.4: These rounds continue until
For absolute clarity, it might be good to expand this phrase something more explicit, e.g.:
The phases described in 3.8.8.1, 3.8.8.2, and 3.8.8.3 are repeated in order until
The meaning of the word "rounds" and what a "round" consists of is implied in 3.8.8, but not defined.

Hi Sora,

I just reformatted the constitution change again to cover some earlier suggestions. If you could look it over and let us know if it responds to your concerns or creates new ones, it would be appreciated.

#269 ::: Jameson Quinn ::: (view all by) ::: May 29, 2015, 11:08 AM:

3.8.8.5.2: I suggest:

(2) During the Elimination Phase if two or more nominees are tied for the fewest nominations, the nominee with the fewest points at that round shall be eliminated, unless doing so would leave exactly the number of finalists specified in 3.8.1.

....

Honestly, if it were up to me, I'd include 6th place even if it were only 1 nomination below 5th place. It's probably better to err on the side of inclusion.

#270 ::: Ian Gillespie ::: (view all by) ::: May 29, 2015, 11:15 AM:

Soon Lee @ 238

I think visualization is super important here. This system will be difficult for people in the - no doubt extraordinarily packed - room to comprehend. Is PowerPoint allowed? You don't want people saying "This is too complicated. Let's take an extra year to get this right, not rush into something and destroy the Hugos in the process."

Once the FAQ (i.e. your talking points) are complete, it would also be worthwhile to consider contacting "influencers" - contacting prominent attendees of the convention to ask for their support/answer any questions/assure them this system is simple and robust. Having prominent members of the community supporting the system will reassure people it isn't a confusing, over-complicated scheme.

The perception of voting systems like this being "too complicated" - the counting being too complicated, not the voting - has been the major reason for their failure in my experience.

#271 ::: David Dyer-Bennet ::: (view all by) ::: May 29, 2015, 11:40 AM:

junego@267:

I like your new 3.8.8.7. I like having an explicit statement about handling "late" changes to the ballot (and leaving lots of it to the discretion of the committee). I like the sentence explicitly (and slightly redundantly; that's fine) saying that original finalists aren't removed. "Acknowledged" isn't right, though -- it's too strong. I'm reasonably sure that the intention is not just to protect finalists who have accepted, but in fact to protect all finalists who have been notified (even if they haven't responded yet). I think what's wanted is not yanking back finalist status from anybody where it's known beyond the (very closed) administration committee (remember, lack of response is treated as acceptance). (Details of how to write it deferred until it's clear that's in fact what's wanted.)

3.9.1 has some clumsy singular/plural workarounds that I think can be cleaned up. How about (second sentence): "Each finalist shall be asked at that time to either accept or decline the nomination. If a finalist declines the nomination, that finalist shall not appear on the final ballot." (Changing definite article to indefinite and removing the optional plural in two places.)

Oh, and our "points" are infinitely divisible; I think they fall in the category of things which call for "less" rather than "fewer" (so, in 3.8.8.2, "...with the least points...") (same as percentage, which seems to be a commonly-discussed situation). (3.8.8.3 should however continue to refer to "the nominee with the fewest nominations..." since those are countable.) Also in 3.8.8.5(2) and 3.8.8.5(4).

(And thanks for acknowledging my input; in this case it would have been clear on detailed examination that you were influenced by it, but just mentioning that you've seen it would possibly keep me from repeating myself in the case where you didn't take my advice. :-) )

#272 ::: junego ::: (view all by) ::: May 29, 2015, 11:42 AM:

[FAQ]

Upthread (sorry, don't remember the name) someone mentioned that the large section in the proposal that covers TIES, 3.8.8.5, makes it appear that there is some problem/weakness in the new voting system. Does anyone else think we should add a response to this concern?

A possible:

Q?: Why is so much space in the proposal spent on ties?
A: One consequence of nominees challenging each other to reach the next round is that ties have to be broken so that one nominee can be chosen. With two measuring criteria of the amount of support a nominee has garnered, total nominations and points, nominees can be tied in four ways. We needed to define those ties and create clear rules on how to handle them. The different ways to break ties were thoroughly discussed. Dozens of simulations were run using real and created data sets to help decide which rules would maintain our goal of supporting the diversity of opinion within fandom. It turned out that all the simulations showed that there was almost no difference in outcomes no matter how we broke ties, so we chose the simplest rules.

#273 ::: Kevin Standlee ::: (view all by) ::: May 29, 2015, 11:49 AM:

In general, the feedback I'm getting from the members of the WSFS MPC is that we don't want any sort of licensing agreement (it would make it look like WSFS was officially sanctioning any particular proposal); we don't need money that badly. (But again, anyone who is looking for a US tax deduction -- the MPC's legal entity is a 501(c)(3) charity -- and/or wants to do WSFS a good turn can contact me regarding a direct donation to the MPC.)

E PLURIBUS HUGO is not something you need WSFS's permission to use. If you use the words HUGO AWARD anywhere in your design (like in a subheading), please put a ® "bug" by it. (Hugo Award is a registered service mark of the World Science Fiction Society.) The acknowledgement of the WSFS service mark does not imply endorsement by WSFS.

Please do not use the Hugo Award Logo in any of your designs. Vaguely rocket-shaped designs, as long as they don't actually reproduce the two that WSFS owns (the logo and the image/design of the actual trophy rocket) aren't likely to attract official complaint from WSFS. Including a "not sanctioned by WSFS" note may be helpful IMO. I can't be more specific than that because it's an intellectual property law minefield, and while I'm Chairman of the WSFS MPC, I'm a computer programmer, not a lawyer.

#274 ::: David Dyer-Bennet ::: (view all by) ::: May 29, 2015, 11:59 AM:

Kevin@273: Just to be clear, what makes it an issue here is that we're talking about merchandise, right? If for example this discussion thread were titled "Discussion of Hugo Award Nominationg Procedures" and used the rocket logo at the top of the page, there'd be no grounds for MPC to object, right?

(Just to be clear -- I fully agree with the MPC's desire not to have WSFS appear to be sponsoring anything, and I would want us to cooperate with that even if the law allowed us to ignore it.)

#275 ::: Jameson Quinn ::: (view all by) ::: May 29, 2015, 01:30 PM:

@272: I like it, but "dozens" should be "hundreds".

#276 ::: Zack ::: (view all by) ::: May 29, 2015, 01:45 PM:

ddb @271 re infinitely divisible points: I think "with the lowest point total" might be even clearer.

#277 ::: David Dyer-Bennet ::: (view all by) ::: May 29, 2015, 02:31 PM:

Zack@276: Not sure about clearer, but to my ear it sounds better, and it's no less clear at least. So I endorse that change to my suggestion.

#278 ::: junego ::: (view all by) ::: May 29, 2015, 03:50 PM:

Jameson Quinn @269

3.8.8.5.2: I suggest:
(2) During the Elimination Phase if two or more nominees are tied for the fewest nominations, the nominee with the fewest points at that round shall be eliminated, unless doing so would leave exactly the number of finalists specified in 3.8.1.


Honestly, if it were up to me, I'd include 6th place even if it were only 1 nomination below 5th place. It's probably better to err on the side of inclusion.

Hey, Jameson, I'm confused about your point. Is this comment about the latest draft @267? The place where 3.8.1 is now mentioned is 3.8.8.6. 3.8.8.5 (2) refers to a tie at Selection and the words you quote are from 3.8.8.5(3) and are referring to how to break any such tie, even at 20th place.

Could you rephrase please?

If there

#279 ::: Jameson Quinn ::: (view all by) ::: May 29, 2015, 04:02 PM:

@278: Yes, I'm referring to the draft @267. My point is that if you are comparing the works with 5th and 6th place point totals, in order to do the very last elimination, and the two works have the same number of nominations, you should just leave them both as finalists, rather than breaking the tie using points.

In other words: the tiebreaker using points is only used when necessary for the method to proceed to the next round. If it's already on the last round, just stop.

#280 ::: Keith "Kilo" Watt ::: (view all by) ::: May 29, 2015, 04:24 PM:

junego@267:

Some general comments...

I would be happy to take-up keeping track of and incorporating suggestions/changes to the official constitution language if it would help you.

Thanks, but it's easier for me if I'm keeping up with the various changes. Appreciate the offer, though!


1) I went over the existing constitution more carefully and reformatted some of the sub-sub-sub-sections to conform to using (1) instead of 3.8.8.4.1, etc.

I like this a great deal. The extended numbering of subsections was bothering me, and I think this is an elegant and effective way to handle it.


2) I reconciled some of the language inconsistencies that had slipped in...shall/should/will, nominees/nominations, points/total points, etc. and, again, tried to conform to current usage.

I also strongly agree with this. I think someone found a few more instances as well. We'll need to do a good fine-toothed pass over the proposal to make sure we've caught all of these. Same goes for the fewer/least wording as well.


I clarified the issue of replacing ineligible or non-accepting finalists (per David Dyer-Bennet's suggestion) and moved it to within sub-section 3.8 for clarity.

This is a good idea; I like it. We may need to do a bit of reconciling with 3.9, since that's where declined nominations are dealt with currently. Let me see what I can come up with, but I agree having it in the tallying section is useful.


3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees finalists

Small nit here: They actually aren't finalists yet. We select five nominees to -be- finalists, not select five finalists (from among which set of finalists?) to do anything.


Blame it on the Gov't for conditioning me with 50 page specifications for an electrical plug (for example)).

Oh, I hear you. Most of my career was at NASA, and I was a naval officer before that (which is worse, believe me). But I think we may have gone too far in that direction. As Kevin has pointed out, the "legislative history" (FAQ's, commentary, and the like) are just as important in determining how the rules are implemented. I think we need to make sure that the amendment matches the style and reads like the rest of the constitution and not like a flight manual. We actually -don't- want to go the route of goverment sepcifications. I think we may need to swing somewhere in the middle here.

My biggest problem is in your 3.8.8.1. Having so many phases makes the system seem much more complicated than it is. In reality, there's not much going on at this point, and to refer to it as a "Preliminary Survey" seems to imply a lot more is happening than actually is. When it comes down to it, this phase is exactly what is happening in the calculation phase, so breaking it out is confusing. I may can be convinced to put the calculation phase at the end -- though that's not how it actually works, so it sort of violates the prescriptive mode that you're advocating. I still am of a mind that it's simplest to describe the point allocation in 3.8.8.1 and then in 3.8.8.4 just reference that we go back to 3.8.8.1. It's worth pointing out in 3.8.8.4 that the point totals will have changed (the number of nominations will not, unless we had used something conjoined twins for ties), though.

I'm also not sure the definitions of nominations and points (your 3.8.1) needs quite so much space. I think all of this could be tightened up into a single paragraph.


Keep in mind that one of the advantages of the system is that the basic form of the system is pretty simple. We don't want to hide that simplicity in a proposal that makes everything seem more complex than it is. For that reason, I think shorter, tighter paragraphs are a better idea, though not to the point that there is confusion about what we are trying to say. Somewhere between the extremes is probably the correct path, I think. Ultimately, as someone said up-thread, most people are going to skim the legalese and read the FAQ, so -that's- where I think our detailed explanations should go, not in the proposal itself. Since the "legislative history" is part of the rule, we don't really have to worry -quite- so much about rules lawyering as we would otherwise.


All:
Most of this is just wordsmithing, but there is one substantive difference that the group needs to discuss. Specifically, do we want to mention the calculation part of the system before we mention the elimination or after? I can see some arguments either way, but so far, it still seems to me that things flow better with doing calculation first -- particularly since, as I said, that's how it's actually done.

Let me know what you all think, and I'll see what I can come up with doing it both ways.

Thanks,
Kilo

#281 ::: Chris Battey ::: (view all by) ::: May 29, 2015, 04:24 PM:

[SHIRTS:DESIGN]

Thanks for the feedback, everyone. A couple thoughts on specific points:

Seal design: As David Harmon mentioned, the US-centrism is, for better or worse, baked in to the proposal's name already. The design is certainly busy, but then so is the seal itself. However, the similar design used as part of the Presidential seal uses a single banner for the motto (rather than the split banner my current design uses), which I could expand to take up the entire upper part of the design (and make the E PLURIBUS HUGO text one of the focal points of the design). I'm going to leave the shield in for the next revision (probably with the ringed-planet icon), but if the design still looks too busy I'll take it out.

Phoenix/other bird designs: I'm personally confident in my ability to represent a phoenix in monochrome, though I acknowledge I'm not there yet. I'm looking at options for halftone gradients, and I'll price out color shirts as well. I don't want to appropriate the Pacific Northwest native art style (despite Sasquan already doing so); I have neither a cultural claim to the style nor any experience in working with it.

[SHIRTS:LOGISTICS]

My current plan (if there are enough people who like the final design) is to set up a Teespring campaign, which will let me set a minimum goal for purchases; that minimum plus the desired markup (i.e. zero in this case) will determine the price. (If significantly more than the minimum number of shirts is purchased I may up with a small profit; if that is the case I'll apply it towards the ribbon order to reduce final costs to everyone.)

If that doesn't work for some reason, the alternative is (as Kevin mentioned) a CafePress store with a zero-markup design. Teespring will be more cost-effective if more than a small number of people want a shirt, and their quality is better, but CafePress is certainly more flexible for one-offs.

#282 ::: junego ::: (view all by) ::: May 29, 2015, 04:30 PM:

David Dyer-Bennet @271
junego@267:

I like your new 3.8.8.7. I like having an explicit statement about handling "late" changes to the ballot (and leaving lots of it to the discretion of the committee). I like the sentence explicitly (and slightly redundantly; that's fine) saying that original finalists aren't removed. "Acknowledged" isn't right, though -- it's too strong. I'm reasonably sure that the intention is not just to protect finalists who have accepted, but in fact to protect all finalists who have been notified (even if they haven't responded yet). I think what's wanted is not yanking back finalist status from anybody where it's known beyond the (very closed) administration committee (remember, lack of response is treated as acceptance). (Details of how to write it deferred until it's clear that's in fact what's wanted.)

Yes, the consensus from the last thread was that no one should be disqualified by a rerun after they have been notified. If you have a suggestion for clearer wording of that intention without writing a paragraph it would be appreciated,

3.9.1 has some clumsy singular/plural workarounds that I think can be cleaned up. How about (second sentence): "Each finalist shall be asked at that time to either accept or decline the nomination. If a finalist declines the nomination, that finalist shall not appear on the final ballot." (Changing definite article to indefinite and removing the optional plural in two places.)

Our intent was to propose only the minimum required changes to the constitution. We were advised that would be best wrt the BsMtg. I agree with what you're saying, but adding this to the proposal is beyond our brief.

After reading the document closely in the last week, I'd like to redline huge chunks of it! Of course it shows that it's a 50+ year old document that's been somewhat kluged over the decades. FI, instructions for the final vote for the Hugo is half in the correct section & half over in a completely unrelated section with language that is...obscure is a kind way to put it.

Oh, and our "points" are infinitely divisible; I think they fall in the category of things which call for "less" rather than "fewer" (so, in 3.8.8.2, "...with the least points...") (same as percentage, which seems to be a commonly-discussed situation). (3.8.8.3 should however continue to refer to "the nominee with the fewest nominations..." since those are countable.) Also in 3.8.8.5(2) and 3.8.8.5(4).

I disagree. We are in fact counting the points, for our purposes they are only divisible by 5 (not infinitely), and I'd prefer to use the same adjective to describe the smallest amount of counted things for both criteria. I do appreciate the input, though.

(And thanks for acknowledging my input; in this case it would have been clear on detailed examination that you were influenced by it, but just mentioning that you've seen it would possibly keep me from repeating myself in the case where you didn't take my advice. :-) )

Well, *I* wouldn't shut up if I thought I had a point, whether or not I was mentioned. You (possibly?) are a better person than I am. :^]

#283 ::: Jameson Quinn ::: (view all by) ::: May 29, 2015, 04:47 PM:

I think calculation (point allocation?) should go first, not last. I understand and respect that junego has found that so far they've been able to make it flow better by putting it last. But I'm confident that as a group, if we put our heads together, we'll be able to overcome that hurdle, and get it to flow well with the calculation step first.

#284 ::: junego ::: (view all by) ::: May 29, 2015, 05:10 PM:

Kilo @280


All:
Most of this is just wordsmithing, but there is one substantive difference that the group needs to discuss. Specifically, do we want to mention the calculation part of the system before we mention the elimination or after? I can see some arguments either way, but so far, it still seems to me that things flow better with doing calculation first -- particularly since, as I said, that's how it's actually done.

Well, I'll defend my proposal (although I'm not unalterably married to it or anything).

The way I am describing it is like walking the admins through the steps, sorta...the computer actually does almost all the work.

First they gather up the ballots and create a baseline from which to launch the process (Preliminary Survey). Generating the nominations number will only be done once so saying that this subsection (Calculation) is part of all subsequent rounds also means extra verbiage to explain that recalculating nominations is not part of the rounds. this is part of the 'klunk' I mentioned.

Second, the admins start the elimination rounds by selecting, then eliminating, then reapportioning (Calculation) the points. Rinse, repeat. *This* is the simple, repeatable process I was trying to show.

Your instincts are correct that generating these two numbers is the first thing, but we only repeat one recalculation with every round, though. Having the three steps of the repeated process clearly delineated seems clearest and simplest to me.

#285 ::: junego ::: (view all by) ::: May 29, 2015, 05:14 PM:

Jameson Quinn @283
"I think calculation (point allocation?) should go first, not last. I understand and respect that junego has found that so far they've been able to make it flow better by putting it last. But I'm confident that as a group, if we put our heads together, we'll be able to overcome that hurdle, and get it to flow well with the calculation step first."

Calculation IS FIRST, it's right there in the Preliminary Survey. RECALCULATION of one number is what each round consists of.

#286 ::: Keith "Kilo" Watt ::: (view all by) ::: May 29, 2015, 05:17 PM:

junego@284:

Having the three steps of the repeated process clearly delineated seems clearest and simplest to me.

Here we agree. But it seems to me that your last idea has four steps (albeit one that is not repeated) instead of three. In reality, the same calculation is done in your preliminary step as in the recalculation step, so it seems to me that making it look like they are different just adds phantom complexity that isn't actually there.

Kilo

#287 ::: David Dyer-Bennet ::: (view all by) ::: May 29, 2015, 05:20 PM:

junego@282:

Okay, suggested 3.8.8.7 wording change—maybe it's easier than I thought, it almost works to just change "acknowledged" to "notified". Or, a step further, "None of the remaining original finalists who have been notified shall be removed as a result of this rerun."

3.9.1: Oops, was I fixing stuff that we hadn't added? Never mind! We should leave that bag closed, as you say.

"less" vs. "fewer": We're not counting the points, we're calculating them; we've even named the entire phase for that activity ("calculation phase").

(And I believe we can end up with increments finer than 1/5 point; while that's the limit for a nominating ballot, there's a special case where the admins move a work from its default category to a different category, and nominations for the work are required to be preserved even if the nominator had already used all five slots in the new category themselves. And while in practice I can't really believe in more than one work being moved into the same category that way, it's perfectly legal.)

So no, I think our point values must be thought of as at least in "rational number" space.

Well, *I* wouldn't shut up if I thought I had a point, whether or not I was mentioned. You (possibly?) are a better person than I am. :^]

Well, perhaps I'm attempting to pass for that hypothetical better person, anyway. Now and then.

Seriously, I think some things are more important than others (not, I believe, unique to me!), and I'm more stubborn about things I think are more important.

#288 ::: Keith "Kilo" Watt ::: (view all by) ::: May 29, 2015, 05:29 PM:

David@287:

Your argument has me persuaded: We are "calculating" points not "counting" them, so it would be "least" and not "fewer". That's a good explanation, thanks.

K

#289 ::: Tim Illingworth ::: (view all by) ::: May 29, 2015, 05:49 PM:

KW@280: you said: The extended numbering of subsections was bothering me

I'd really suggest fixing this by making 3.8.8 its own section. The usual was to do this is:
"Insert new section 3.A after Section 3.8 as follows:
Section 3.A Nomination Elimination Procedure: "
and the rest of the text.

I think the fact that we're in sub-sub-sections means we need a break to make it more readable.

#290 ::: Keith "Kilo" Watt ::: (view all by) ::: May 29, 2015, 06:07 PM:

Tim@289:

That sounds like a good idea to me. Does anyone see any issues with that?

K

#291 ::: David Dyer-Bennet ::: (view all by) ::: May 29, 2015, 06:17 PM:

Order of description and calculation, number of steps, etc. (many previous messages):

Having a "calculation phase" which combines both counting nominations (which remain constant thereafter) and calculating points (which must be re-done each iteration) is the basis of the problem I think.

This is also a place for making it "as simple as possible, and no simpler." Just possible we may not all agree on exactly where that boundary is :-).

The obvious solution is to break out the counting to a separate step, leaving us with:

3.8.8.1: Startup: The total number of nominations for each nominee in the category shall be tallied from all ballots.

3.8.8.2: Point calculation phase: Each ballot shall have its one point redistributed equally among all remaining nominees on that ballot.

3.8.8.3: Selection phase: Select the two nominees with the fewest points for the Elimination Phase (3.8.8.4). (See 3.8.8.6 for ties.)

(Current 3.8.8.3 "elimination phase" renumbered 3.8.8.4 but text unchanged)

3.8.8.5: Rounds shall continue (going back to point calculation at 3.8.8.2) until the number of finalists specified in 3.8.1 remain. If an elimination round would reduce the number of finalists to fewer than that specified in 3.8.1, then none of the nominees in that round shall be eliminated and all remaining nominees shall appear on the final ballot, extending the final ballot if necessary.

(Current 3.8.8.5 "ties" renumbered 3.8.8.6 but text unchanged.)

(Current 3.8.8.7 kept as is)

(I have heroically resisted the tempation to refactor this as a "while" loop rather than the current "goto" looping construct.)

(There's something of an inconsistency about when we do and don't refer to the fact that we're acting within a category all the time; we probably need to clarify that up front and remove it in each detail step? I have not, I think, made it any better in my proposed refactor.)

I added "final" qualifying the ballot that's extended if elimination would take us below the goal number of finalists. The nominating ballot is frequently just called a "ballot", so we need to somehow distinguish references to the final ballot what is the output of our process.

Possibly "selection phase" and "elimination phase" should be combined into one numbered section; selection phase is very very simple (all the complexity is by reference to handling ties), so it adds to the step count without isolating complexity.

#292 ::: Keith "Kilo" Watt ::: (view all by) ::: May 29, 2015, 06:24 PM:

David@291:

One thing to consider if it helps: in the code we actually do calculate number of nominations and total points each round. While under the current version of SDV-LPE the number of nominations won't change, there are some variations where it will. In particular, the "conjoined twins" method of tie breaking requires that the number of nominations be recounted each round (since the twins get half a nomination as well as half their usual number of points). We decided against using that method, but it's not inconceivable that someone will want to do it that way, now or 15 years from now. It would make that eventuality simpler if we leave it so that both measures are calculated each round. The code does this now just to maintain that flexibility. There's no real reason not to recount, it's just that number won't change in the current version.

Kilo

#293 ::: David Harmon ::: (view all by) ::: May 29, 2015, 06:26 PM:

Re: Calculation Phase: ISTM that it's incidental that the nomination count for a nominee "doesn't change"... until it drops to 0 for an eliminated one. I think it makes more sense to have the calculation phase described once up front, and a GOTO 20 at the end of the phase.


#294 ::: David Dyer-Bennet ::: (view all by) ::: May 29, 2015, 06:30 PM:

Keith@292: Ah, modern computer software design! When it's not worth the effort to move an invariant out of the loop. That's probably true; the size computational task this poses is so trivial in today's terms that it really doesn't, I suspect, matter the slightest little bit. Probably can't measure it with a manual stopwatch.

However, I may have made a tactical error by talking in terms of code, because my primary reason for the separation is to make it clearer. In the old description, calculation happens in two places, and that's IMHO confusing.

#295 ::: junego ::: (view all by) ::: May 29, 2015, 06:33 PM:

Keith "Kilo" Watt @286
junego@284 said:
Having the three steps of the repeated process clearly delineated seems clearest and simplest to me.

Kilo said:
"Here we agree. But it seems to me that your last idea has four steps (albeit one that is not repeated) instead of three. In reality, the same calculation is done in your preliminary step as in the recalculation step, so it seems to me that making it look like they are different just adds phantom complexity that isn't actually there."

That's a fair point.

OTOH cramming the description of the two numerical criteria and how they are generated and how the points are reaportioned after each elimination and what constitutes a round -all under the one section, then carving out the subset of that explanation as the part of the round that is repeated seemed unnecessarily complex/confusing to me.

It made more sense to me to separate out the definition of terms and how the criteria are used into a different section. Since I already had a different section and the clearest way to describe the "round" was to put it in three repeating steps (with repeated language), I opted to use the new section to say, "hey, here's the baseline we start from."

We can certainly put Calculation next and make it work, if a bit awkwardly istm. It would be a two part - a) create the first iteration of the things just described; b) after the first elimination, reapportion points, recalculate and repeat this step for every round.

You still need the extra section or *something* to define terms, etc. I don't think cramming all that into "calculation" section is a good way to do it.


#296 ::: Joshua Kronengold ::: (view all by) ::: May 29, 2015, 06:48 PM:

It seems that the things forming a rocket-shaped icon would work as long as it doesn't exactly duplicate the Hugo rocket (which wouldn't be hard). With a disclaimer, that would make a pretty nifty shirt design.


Re Section 3.11.4:

3.11.4: ...During the same period a record of all rounds of the selection process for each category shall also be published.

It seems to me that this could be onerous. Might it not make sense to specify "at least the last ten rounds"? (to duplicate the 15 item requirement that we're removing).

#297 ::: David Dyer-Bennet ::: (view all by) ::: May 29, 2015, 06:57 PM:

Keith@290: What Tim suggests is the way I've seen proposals that add sections worded in the past, now that he mentions it (I've watched this online for years but never participated in-person at the convention). So it's probably the way we should do it. Tim and Kevin Standlee (not involved in this advice, though he's been in this thread) are among the people most commonly involved in helping other people draft proposals (and not just ones they support), so their advice on how to word the proposal should be taken very seriously.

#298 ::: junego ::: (view all by) ::: May 29, 2015, 07:19 PM:

David Dyer-Bennet @287
junego@282

"less" vs. "fewer": We're not counting the points, we're calculating them; we've even named the entire phase for that activity ("calculation phase").
(And I believe we can end up with increments finer than 1/5 point; while that's the limit for a nominating ballot, there's a special case where the admins move a work from its default category to a different category, and nominations for the work are required to be preserved even if the nominator had already used all five slots in the new category themselves. And while in practice I can't really believe in more than one work being moved into the same category that way, it's perfectly legal.)
So no, I think our point values must be thought of as at least in "rational number" space.

If I cut a whole bunch of pies into pieces for a picnic and then went to find out how many pieces of pie each family ended up with, but translated the number back into the equivalent of how many whole pies that would be for each family, I'm still just counting pie. I wouldn't say the Smith's had lesser pies (pieces of pie) than the Jones', I'd say they had fewer. That's the exact equivalent of what we're doing here.

#299 ::: junego ::: (view all by) ::: May 29, 2015, 07:33 PM:

Keith "Kilo" Watt @290
[PROPOSAL]

Tim@289said:
"this by making 3.8.8 its own section. The usual was to do this is: "Insert new section 3.A after Section 3.8 as follows:
Section 3.A Nomination Elimination Procedure: "
and the rest of the text."

Kilo said:
"That sounds like a good idea to me. Does anyone see any issues with that?"

This sounds fine to me.

I think we'd have to rename 3.8 though since the current title is "Tallying of Nominations"? Everything else in the section is about the preliminary work -before- the ballots are counted.

#300 ::: junego ::: (view all by) ::: May 29, 2015, 07:56 PM:

David Dyer-Bennet @291

The obvious solution is to break out the counting to a separate step, leaving us with:
3.8.8.1: Startup: The total number of nominations for each nominee in the category shall be tallied from all ballots.
3.8.8.2: Point calculation phase: Each ballot shall have its one point redistributed equally among all remaining nominees on that ballot.

Part of the problem is that someone (can't remember who) told the group a while back that too many steps might make the proposal look too complicated, which could alone cause excess resistance at the business meeting.

One possibility would be to define a round as one sub-section and use the (1), (2), (3) format that seems to be more acceptable to fans. Maybe?

Do similar formatting tweaks to make it seem less complex.


#301 ::: junego ::: (view all by) ::: May 29, 2015, 09:21 PM:

#299 ::: junego ::: (view all by) ::: May 29, 2015, 07:33 PM:
Keith "Kilo" Watt @290
[PROPOSAL]

Tim@289said:
"this by making 3.8.8 its own section. The usual was to do this is: "Insert new section 3.A after Section 3.8 as follows:
Section 3.A Nomination Elimination Procedure: "
and the rest of the text."

Kilo said:
"That sounds like a good idea to me. Does anyone see any issues with that?"

junego said:
"This sounds fine to me.

I think we'd have to rename 3.8 though since the current title is "Tallying of Nominations"? Everything else in the section is about the preliminary work -before- the ballots are counted."
*************************************************

I may have to take this back. There are already Sections 3.9 through 3.13. We'd have to renumber all of these, unless we made our proposal 3.14 and jumped all the way to bottom of the queue. I don't think either of these is a good option.

That sounded so good for a minute or ten.

#302 ::: Tim Illingworth ::: (view all by) ::: May 29, 2015, 10:04 PM:

Junego@301: the renumbering will be taken care of automatically if this is ratified - see Rule 4.3.

#303 ::: Keith "Kilo" Watt ::: (view all by) ::: May 29, 2015, 10:06 PM:

junego@301:

Kevin or Tim can correct me if I'm wrong, but I think the "nitpicking and flyspecking" committee is empowered to renumber sections if needed. So I think with Tim's wording, we're probably okay if we want to do it. And as DavidDB@294 pointed out, if they recommend a certain wording, that carries a lot of weight, as far as I'm concerned. I'm certainly not opposed to it, unless someone thinks it's a bad idea for some reason.

K

#304 ::: Keith "Kilo" Watt ::: (view all by) ::: May 29, 2015, 10:06 PM:

..aaand Tim and I cross-posted. :) Thanks Tim!

K

#305 ::: Tim Illingworth ::: (view all by) ::: May 29, 2015, 11:12 PM:

And, while looking at other bits of the Constitution, I noticed in Section 6.4:
"If two or more candidates are tied for elimination during this process, the candidate that received fewer first-place votes shall be eliminated. If they are still tied, all the tied candidates shall be eliminated together." which looks like option 2a to me...

#306 ::: Keith "Kilo" Watt ::: (view all by) ::: May 29, 2015, 11:37 PM:

So basically we are making tie breaking in the nomination phase procedures match the procedures outlined in 6.4. That's a strong argument indeed for 2a.

I think we should point that out in the FAQ...

Kilo

#307 ::: David Dyer-Bennet ::: (view all by) ::: May 30, 2015, 01:21 AM:

jonego@298: if you have given me three pieces of pie on my plate of various sizes, and I want more or less, I will ask for more, or for less; I will not ask for "fewer". The pieces aren't the essence, the quantity is the essence.

#308 ::: David Dyer-Bennet ::: (view all by) ::: May 30, 2015, 01:30 AM:

junego@300: yeah, "simpler" is actually a subjective judgment.

I find linear composition of non-interacting segments to be simple, even if there are quite a few of them. Complexity comes from choices. For me.

Overall, I wasn't fully happy with the representation of the cycle of counting in my last attempt (that's part of what I was referring to about refactoring it to a "while" loop).

Maybe:

x.y.A: Repeat the following steps until x.y.A.4 says to stop:

x.y.A.1: Count nominations
x.y.A.2: calculate points
x.y.A.3: Select two lowest point-holders (or more, see "tie" rules below); eliminate the one with the fewest nominations (see "tie" rules below).
x.y.A.4: If we're down to the number of finalists specified above, stop.

(that's not quite right, at least not complete, but it's to show the idea of putting the steps of the loop neatly together on their own sublevel, and putting a clear controlling statement about the looping at the start so people understand they're entering the loopy part.

#309 ::: Xopher Halftongue ::: (view all by) ::: May 30, 2015, 02:12 AM:

Well, I couldn't follow the in-depth theoretical discussion that led to this. I just don't have that kind of head for numbers...or number-y things. The discussion in the regular threads made my head swim.

However, I am not confused by this (well, I'm trusting you all on some of the details). I believe I understand the process in general, and I see why it should make slates drastically less effective.

If adding my name (Christopher Hatton) to the signatory list is neutral-to-positive, please do. If not, I will try to show up at the WSFS meeting to vote for it.

#310 ::: junego ::: (view all by) ::: May 30, 2015, 02:28 AM:

David Dyer-Bennet @307

jonego@298: if you have given me three pieces of pie on my plate of various sizes, and I want more or less, I will ask for more, or for less; I will not ask for "fewer". The pieces aren't the essence, the quantity is the essence.

By this reasoning points and nominations are the same, too. It is the quantity of both that determine who survives each round. Does the nominee want more or less total points; more or less total nominations? We calculate how many total nominations per nominee there are by counting/adding them. We do exactly the same thing for total points.

In the last thread, part of the discussion of how we were going to describe this process was whether we should use 1 divided point, 100 points, or 60 points. Mathematically they are identical and for the purposes of what we're tallying, they are identical. Grammatically they look fairly indistinguishable, too.

I think "fewest" is the correct way to describe the number of points and the number of nominations that are received.

#311 ::: junego ::: (view all by) ::: May 30, 2015, 02:41 AM:

Tim Illingworth @302
Junego@301: the renumbering will be taken care of automatically if this is ratified - see Rule 4.3.

Thanks for the info. That makes your idea a great solution to part of our problem.

#312 ::: junego ::: (view all by) ::: May 30, 2015, 03:07 AM:

#308 ::: David Dyer-Bennet @308

junego@300: yeah, "simpler" is actually a subjective judgment.
.................
x.y.A: Repeat the following steps until x.y.A.4 says to stop:
x.y.A.1: Count nominations
x.y.A.2: calculate points
x.y.A.3: Select two lowest point-holders (or more, see "tie" rules below); eliminate the one with the fewest nominations (see "tie" rules below).
x.y.A.4: If we're down to the number of finalists specified above, stop.
(that's not quite right, at least not complete, but it's to show the idea of putting the steps of the loop neatly together on their own sublevel, and putting a clear controlling statement about the looping at the start so people understand they're entering the loopy part.

I like the concept, that's kind of what I was trying for. How to format and where to put all the explainy bits (like what reapportioning and re-adding the points means) is part of the problem. It's fixable, just need to work on it a bit more.

#313 ::: Keith "Kilo" Watt ::: (view all by) ::: May 30, 2015, 03:09 AM:

Regarding less/fewer:

According to the NY Times style book, you use less with a singular noun or a quantity considered to be a single bulk amount (as opposed to a set of discrete things that you can lay out and count). Fewer is used exclusively with plural nouns.

So, you would definitely say "fewest number of nominations", since you can lay the ballots out and count them. Points are a little more ambiguous. If you are talking about individual points that can be counted, then you would clearly say "fewest points". If they are not discrete objects and are being treated as a bulk quantity, however, then you would say "least points".

The unambiguous way to deal with it, according to NYT's guidance, would be to say "lowest point total" since you then clearly have a singular noun.

Can we agree on that compromise?

Kilo

#314 ::: Aan ::: (view all by) ::: May 30, 2015, 07:38 AM:

Chris Battey @ #281

Non-US PoV here. The Latin phrase by itself does not come over as US-centric; that it's used in the USA doesn't appropriate it. (In fact, that use barely registers in my awareness.)
The seal design however does very much come over as US-centric (the imagery is only familiar from US movies and such, and has militaristic connotations), and because of that would personally give me a negative inclination toward any proposal which featured it heavily.

Note that last word, "heavily". I'm not out-of-hand opposed to this as a T-shirt design, but I do think that care should then be taken to not expand its use. (As opposed to Jameson's "rocket made up out of Hugo titles" design, which I think could be plastered all over any material having to do with the proposal without any ill effects.)

#315 ::: junego ::: (view all by) ::: May 30, 2015, 08:01 AM:

Keith "Kilo" Watt @313

Regarding less/fewer:
According to the NY Times style book, you use less with a singular noun or a quantity considered to be a single bulk amount (as opposed to a set of discrete things that you can lay out and count). Fewer is used exclusively with plural nouns.
So, you would definitely say "fewest number of nominations", since you can lay the ballots out and count them. Points are a little more ambiguous. If you are talking about individual points that can be counted, then you would clearly say "fewest points". If they are not discrete objects and are being treated as a bulk quantity, however, then you would say "least points".
The unambiguous way to deal with it, according to NYT's guidance, would be to say "lowest point total" since you then clearly have a singular noun.
Can we agree on that compromise?

Kilo and everyone,

I apologize for the derail and grammer neepery. Go ahead and call them "least points" and "fewest nominations". It won't occur to most people to question the difference and it isn't worth any more time and argument.

#316 ::: Jeffry Herman ::: (view all by) ::: May 30, 2015, 09:00 AM:

Joshua Kronengold@296:

> "Re: It seems to me that [Section 3.11.4] could be onerous. Might it not make sense to specify "at least the last ten rounds"? (to duplicate the 15 item requirement that we're removing)."

I wondered about this too. It's also my understanding that it's been generally considered a little inappropriate to publish a list that effectively reveals all of the works and people that only got one or two nomination votes.

(Apologies if this has been discussed before somewhere and I missed it; I realize I am late to the discussion and have tried to catch up on everything but I could easily have missed some.)

#317 ::: Jameson Quinn ::: (view all by) ::: May 30, 2015, 09:08 AM:

So, here are three more versions of the rocket idea: simple, disclaimer, fat.

#318 ::: Jameson Quinn ::: (view all by) ::: May 30, 2015, 09:19 AM:

I agree with @296/ @316.

#319 ::: Jameson Quinn ::: (view all by) ::: May 30, 2015, 09:48 AM:

And yet another design variation: 3lines.

#320 ::: JJ ::: (view all by) ::: May 30, 2015, 10:24 AM:

Jameson Quinn, #317, #319:

For some reason, the "simple" one looks best. Is the font size/thickness different on that one?

My preference would be that design, but facing right (forward) rather than left (backward), and the two other plumes could be added to that if necessary.

#321 ::: David Harmon ::: (view all by) ::: May 30, 2015, 10:39 AM:

Jameson Quinn #317,#319: I also like the "simple" best. For the disclaimer, "not endorsed by" feels off, not to mention it becomes wrong if we win. :) I'm not sure what the best phrasing would be, but Kevin used "not sanctioned by".

#322 ::: Jameson Quinn ::: (view all by) ::: May 30, 2015, 10:52 AM:

@320: The reason I have the rocket going "backwards" is that if you have it "forwards", the motto reads as going into the rocket, not coming out of it. I understand it's a tradeoff there, and certainly if "forwards" has more support I'll go with that, but I personally prefer "backwards".

#323 ::: Jameson Quinn ::: (view all by) ::: May 30, 2015, 11:13 AM:

Here's an attempt at forwards. As with the "simple" above, I converted fonts to paths, so it should look good even if you don't have the "Futura" font on your system. As with "fat" above, I set a minimal between-line space, leading to a slightly chubbier rocket shape and larger font sizes. As with "3lines", I put the motto on 3 lines, which I think helps make it so the eye isn't going upstream on the rocket exhaust; I think this helps a lot compared to the corresponding 1-line version.

#324 ::: Keith "Kilo" Watt ::: (view all by) ::: May 30, 2015, 11:26 AM:

Jameson@323:

Honestly, I think your design at 195 total rocks. It's by far my favorite.

L

#325 ::: Jameson Quinn ::: (view all by) ::: May 30, 2015, 12:00 PM:

@324: I like @195 too. But when I imagine it on a T-shirt, I um... well, I wonder if there might not be unfortunate associations, which the diagonal designs somewhat lessen. Also, the diagonal designs allow for bigger fonts, because the diagonal distance on a shirt is bigger than the vertical one.

I'm sorry, I know that for some people once I say that about the associations, I can't unsay it. But for me, it's not so extreme an issue that it becomes intolerable; it's just a kind of a thing I'd rather not encourage, and I think the diagonal design is enough to avoid it.

#326 ::: Jameson Quinn ::: (view all by) ::: May 30, 2015, 12:04 PM:

My personal favorite so far is @323. But again, I'm happy with whatever the group here likes best, whether it's one of my designs or not.

#327 ::: P J Evans ::: (view all by) ::: May 30, 2015, 12:10 PM:

316
The list of totals that was released in 1984 is here - and that's as close to the original format as I could come. (It's handwritten on quadrille paper.)

#328 ::: David Goldfarb ::: (view all by) ::: May 30, 2015, 12:12 PM:

I also like #323. You need to capitalize the "sky" in Farmer in the Sky, by the way.

#329 ::: JJ ::: (view all by) ::: May 30, 2015, 12:27 PM:

Jameson Quinn, #325: I wonder if there might not be unfortunate associations, which the diagonal designs somewhat lessen. Also, the diagonal designs allow for bigger fonts, because the diagonal distance on a shirt is bigger than the vertical one.

I agree (I had already had the same thought). And I take back my suggestion to have the rocket angled upward to the right. I think it looks better angled upward to the left, with E Pluribus Hugo on one line, as in your "simple" version in #317.

#330 ::: Keith "Kilo" Watt ::: (view all by) ::: May 30, 2015, 01:38 PM:

Jameson@325:

Yeah, I can see what you're saying there. Still my favorite, but other than that, 317 simple is my second favorite.

Kilo

#331 ::: Cally Soukup ::: (view all by) ::: May 30, 2015, 02:12 PM:

[SHIRT]

I like the 317 "simple" design best.

#332 ::: Jameson Quinn ::: (view all by) ::: May 30, 2015, 03:24 PM:

OK, here's a tweaked "simple", with fixed "Sky" and disclaimer: 1Line.

#333 ::: David Dyer-Bennet ::: (view all by) ::: May 30, 2015, 03:36 PM:

Very much liking the evolution of the Quinn t-shirt design.

I don't find the direction I read the slogan relative to the rocketship to matter; I don't really get different feelings from "coming out" vs. "going in" directions.

The many reasons for slanted rather than straight up seem to be enough to settle it for me.

Really don't like the disclaimer. Possibly some of it is the wording of this particular disclaimer, not absolutely sure. I see no need for such a disclaimer.

I intend to buy a t-shirt when they're offered (won't be at Sasquan, so it would have to be if they're offered online).

#334 ::: abi ::: (view all by) ::: May 30, 2015, 03:38 PM:

I like the Quinn design. If it's available online, I'll buy one; if not, I'm going to suborn someone at Sasquan to buy it for me and send it over.

#335 ::: David Dyer-Bennet ::: (view all by) ::: May 30, 2015, 03:41 PM:

junego@315 and everyone: So far as that goes, I started the less/fewer neepery, so if one is warranted I should be the one apologizing.

I'm perfectly happy with "lowest point total". And feel that I've had my say on less/fewer here and won't bring it back up. But might well continue talking about it if others do :-) .

#336 ::: Tina Black ::: (view all by) ::: May 30, 2015, 03:42 PM:

I could never vote to approve a system that "empowers admins" -- that is a gateway to uneven treatment that need never be opened.

Do I need to point out that we do not need a system that would allow abuse?

I didn't think so.

#337 ::: P J Evans ::: (view all by) ::: May 30, 2015, 03:57 PM:

336
A bit late to that fight.

(Admins have power already: they fix messes that other people make.)

#338 ::: Keith "Kilo" Watt ::: (view all by) ::: May 30, 2015, 04:07 PM:

Tina@336:

Just to be clear, we actually decided against giving admins any specific discretion, so your thoughts are in line with the current version of the proposal.

Kilo

#339 ::: David Dyer-Bennet ::: (view all by) ::: May 30, 2015, 04:07 PM:

Jeffry@316: Interesting thought about the current wording requiring giving out all the details about the people who got 3 just nominations.

The data released on the final count is essentially what our current wording mandates for our nomination count; it's everything (short of raw ballots). It shows who was eliminated first, and by how many votes, and who placed below "no award". I don't really see how "getting only 3 nominations" is more embarrassing than placing below "No award".

The idea of showing the last n rounds (for n=10 perhaps, but whatever) works fairly well with this counting method; it shows the last 10 eliminations, which means that in addition to the 5 finalists it shows you the 10 works that came close but were eliminated, and the order that happened in. I don't think that's too bad a choice. Normally the last 10 rounds eliminate 9 works, since the last round terminates without eliminating anything. But ties may mean more eliminations than that are actually shown.)

(As a separate proposal, one might consider changing the release rules to require that in addition to the counting report on the final ballot and whatever we mandate on nominations, the full set of nominations/ballots be released in anonymized form. This is kind of a radical transparency requirement, plus it provides rich data for researchers. I'm not currently sitting down to write up such a proposal.)

#340 ::: Jeffrey Smith ::: (view all by) ::: May 30, 2015, 04:15 PM:

Jameson:

This lurker would like to add his admiration for the t-shirt design. I personally don't consider the Retro Hugos "real" Hugos and wouldn't include them, but it's not my design. However, the fact that you added them later shows.

These titles aren't necessarily in the design where you would have placed them had they been there in the beginning. In particular, Fahrenheit 451 and Farmer in the Sky are spread out too much, and would fit better higher in the rocket.

Also, the spaces above and below the Retro winners are slightly wider than the spaces in the original design.

Really nitpicky, I know. It's just that the design is so good...

The other thing I don't like, the ampersand in Harry Potter, I can see is unavoidable.

#341 ::: Tina Black ::: (view all by) ::: May 30, 2015, 04:37 PM:

#338 Keith -- That is good. I can't see writing in powers for altering a straight vote. I would rather see an extra contender than see an administrative cut.

#342 ::: Ginger ::: (view all by) ::: May 30, 2015, 05:07 PM:

[SHIRTS]

I'll buy a shirt or three, of whichever design is deemed the er, winner.

#343 ::: felice ::: (view all by) ::: May 30, 2015, 06:27 PM:

Joshua Kronengold @296: "Re Section 3.11.4: It seems to me that this could be onerous. Might it not make sense to specify "at least the last ten rounds"? (to duplicate the 15 item requirement that we're removing)."

I would suggest we keep the existing "at least the fifteen highest vote-getters", and supplement rather than replace it with enough rounds of SDV-LPE to show all the eliminations from said list. I think some voters would find it reassuring to continue to receive the same nomination data we currently get; showing all the rounds would take up a lot of space (we're dealing with hundreds of works in some categories) and is an unnecessary break with tradition; and showing only the last few rounds without the plain nomination data risks people thinking something they liked was eliminated unfairly earlier in the process.

Shirts: I think the current design is visually excellent, but it feels a bit odd conceptually: "support SDV-LPE - look at all the great works that won Hugos without its help!" But I don't have any better ideas.

#344 ::: Jeffrey Smith ::: (view all by) ::: May 30, 2015, 06:36 PM:

"support SDV-LPE - look at all the great works that won Hugos without its help!"

"Support SDV-LPE -- let's keep the tradition going!"

#345 ::: felice ::: (view all by) ::: May 30, 2015, 07:11 PM:

Jeffrey Smith @344: "Support SDV-LPE -- let's keep the tradition going!"

Yeah, that works :)

#346 ::: Jameson Quinn ::: (view all by) ::: May 30, 2015, 08:21 PM:

Here's a reshuffled version addressing @340.

As far as I can see, there seems to be pretty good consensus about almost all the aspects of this — concept, direction, etc. The one thing which I think is worth debating is the disclaimer. What do people think about that? I'm going to abstain from opining on that, because in the end this is about community norms, and I'm barely a newcomer to the community.

@343: I'd give you 2:1 odds that, if we had all the data, every one of the works on the shirt would have been among the finalists that year, and so presumably would have won. Obviously we don't have all that data, but that's consistent with the data we do have. So yes, I think the answer in @344 is the right one.

#347 ::: Keith "Kilo" Watt ::: (view all by) ::: May 30, 2015, 09:00 PM:

Jameson@346:

It's probably an MPC call, I guess. I will say that personally I really don't like the disclaimer, though...

K

#348 ::: JJ ::: (view all by) ::: May 30, 2015, 09:32 PM:

Kevin Standlee, #197: The versions that don't explicitly use the Hugo Logo itself are better than those that do. WSFS doesn't claim ownership of every rocket-shaped design there is (that would be silly), only the specific implementation of the Hugo logo and the Hugo Award trophy rocket, both of which are registered service marks.

I interpret this as Kevin saying that the rocket made out of titles is acceptable and not a violation of their Mark Protection.

I like the reshuffled version. Yes, Retro Hugos are different; but frankly, I think in some ways they're harder-earned, because they tend to be works which have stood the test of time. If you go back through the list of winners, some of them now are WTF? whereas the Retros tend to be much less so.

#349 ::: JJ ::: (view all by) ::: May 30, 2015, 09:34 PM:

Also, what I meant but did not actually say is that I think Kevin is saying a disclaimer is not needed for the rocket made out of titles.

#350 ::: David Dyer-Bennet ::: (view all by) ::: May 30, 2015, 10:27 PM:

Yes, Kevin is coming as close to that as he's going to (personally saying that something is okay is something you won't see, I think; what you'll see is Kevin saying some things aren't okay, and not saying that about others).

Really, people, there's no chance the rocketship of titles is infringing the Hugo trademark. Really, there's just not. The only authoritative answer is to pay an expert attorney to examine our exact situation and tell us what they think; that's very expensive, and I do not recommend it. But we don't need it; we're not close to areas that infringe, and the people who decide if WSFS is going to object to something (at least two of them) have looked at what's going on and have not said there's a problem. It's not going to get any clearer than that I don't think.

#351 ::: Yarrow ::: (view all by) ::: May 30, 2015, 11:18 PM:

As somebody who had concerns about the rocket design when Jameson first presented it, I'd like to say that those concerns have been addressed, and that I also prefer the design without the disclaimer.

(My previous concerns, I think, were actually about using the design to raise money, which I understand is now off the table.)

I love the design and the idea behind it!

#352 ::: Kevin Standlee ::: (view all by) ::: May 31, 2015, 12:16 AM:

David Dyer-Bennet @274:

Kevin@273: Just to be clear, what makes it an issue here is that we're talking about merchandise, right?
Yes.

#353 ::: Jameson Quinn ::: (view all by) ::: May 31, 2015, 01:19 AM:

How about we have the disclaimer on some kind of temporary extra homemade ribbons that we wear for the con? Which can also say "ask me"...

#354 ::: Cassy B. ::: (view all by) ::: May 31, 2015, 09:32 AM:

Jameson Quinn @353, honestly, I think we've gotten as clear a "go-ahead" as we can reasonably expect absent paying a lawyer. It looks like having the disclaimer is not necessary. And frankly, I think the t-shirt looks better without it. (It's a good design; I really like it.)

#355 ::: Shane ::: (view all by) ::: May 31, 2015, 10:51 AM:

[T-shirt]
Jameson Quinn@353. Lovely, I'll take two. Just hope it ships to Australia. My favorite design was the vertical with E Pluribus Unum on three lines next to it. Set the rocket to one side if that helps. But I'd buy any of them.

Also, what Aan said about US-centric. The phrase doesn't feel that way, the seal does.

#356 ::: Shane ::: (view all by) ::: May 31, 2015, 11:16 AM:

[FAQ]
I'm strongly allergic to the phrasing of 21 ("Isn't this just made by SJWs to stop SP"). It seems at once defensive and doesn't address the problem. (Apologies to the author).

Best to say that 1. this proposal gives a voting bloc results in proportion to their numbers, they can't sweep the field but nor will they be disenfranchised. They're basically ineffective. And 2. This effect applies just as well on hidden or informal groups as it does on explicit slates.

You can then address TNH, PNH and Kilo's bonafides in another para, if you want. I'd prefer maybe a brief note of where it was discussed and a restatement of our principles.

This reiterates what I said at the link below but I didn't read the whole thread after. Sorry if it has been addressed.

http://nielsenhayden.com/makinglight/archives/016246.html#4132846

Also, if there's been other discussion of Q21 I'd like to go read it, if anyone has a link.

#357 ::: JJ ::: (view all by) ::: May 31, 2015, 11:25 AM:

Instead of this:

21. Wasn't this system just designed by Social Justice Warriors to block the Good Stuff?

How about this:

21. Wasn't this system just designed to keep certain people out of the Hugo nomination process?

#358 ::: Jameson Quinn ::: (view all by) ::: May 31, 2015, 01:34 PM:

@357: your suggestion replaces "isn't it designed by bad people to do bad things" with "isn't it designed to do bad things". That loses something essential.

#359 ::: Keith "Kilo" Watt ::: (view all by) ::: May 31, 2015, 02:10 PM:

Shane@356

I'm actually of two minds about this one as well. On the one hand, if we don't address it, we're sort of ignoring the elephant in the room. On the other hand, if we do address it, then it's bringing the politics into the discussion that we've so far successfully avoided. I'm not sure which is the best approach.

When I wrote that FAQ entry, I was writing in the style of posts I've actually seen on the various blogs. So the phrasing isn't inauthentic. But do we need to be authentic? Maybe not. Maybe there's a way to say the same thing in a "non-Puppy" mode. But there is something to be said for addressing an issue head-on.

Kilo

#360 ::: Jameson Quinn ::: (view all by) ::: May 31, 2015, 03:37 PM:

I think we should rephrase the question so it still has "bad guys", but without loaded terms like SJW. So: "Isn't this just an attempt by groups who have got lots of Hugos in recent years to maintain their dominance?"

#361 ::: Steve Wright ::: (view all by) ::: May 31, 2015, 03:57 PM:

Or you could, perhaps, neutralize it by generalizing it - "Isn't this just an attempt by one clique to keep out another?"

#362 ::: junego ::: (view all by) ::: May 31, 2015, 04:04 PM:

Keith "Kilo" Watt @359
Shane@356

I'm actually of two minds about this one as well. On the one hand, if we don't address it, we're sort of ignoring the elephant in the room. On the other hand, if we do address it, then it's bringing the politics into the discussion that we've so far successfully avoided. I'm not sure which is the best approach.
When I wrote that FAQ entry, I was writing in the style of posts I've actually seen on the various blogs. So the phrasing isn't inauthentic. But do we need to be authentic? Maybe not. Maybe there's a way to say the same thing in a "non-Puppy" mode. But there is something to be said for addressing an issue head-on.

I agree that we need to address the elephant, but I think using "SJW" is inflammatory. It's the insult that's been directed at a swathe of fandom and it's also associated with other toxic movements, primarily GamerGate. It might be less confrontational not to point to the Pups by name, either. So maybe we should kinda sneak up on the elephant? 😏

Here's a possible alternate way to approach the issue.

21. Wasn't this system designed by one side of the "culture wars" to keep the other side out of the Hugos?
The short answer is "No!" The long answer is that, although it's true that the blog owners of "Making Light" (who were named by some as part of one side in the present imbroglio) allowed most of the discussion for this system on their site, they played no active role in influencing our decisions. The threads were open to anyone who wanted to talk about a solution to the problem of a coordinated nominating minority disenfranchising the majority and dominating the final ballot. This is a major flaw in the current Hugo nomination system, and it's a flaw that we feel must be fixed if the integrity of the award is to be maintained. Early on participants reached consensus that any new system *must* be fair to all members.. We have shown, through many hundreds of simulations, that this system promotes neither advantage nor disadvantage for anyone.

#363 ::: junego ::: (view all by) ::: May 31, 2015, 04:16 PM:

Jameson Quinn @360 and
Steve Wright @361
[FAQ]

Q21
"Isn't this just an attempt by groups who have got lots of Hugos in recent years to maintain their dominance?"

and

"Isn't this just an attempt by one clique to keep out another?"

I think some permutation of these would be a better way bring up about the elephant.

#364 ::: Shane ::: (view all by) ::: May 31, 2015, 05:34 PM:

I don't have an issue naming the Puppies, and accepting the name SJW*, I think our answer should be "No, this is what you said you had a problem with. We took that in good faith and addressed it head on."

@363 I'd amend junego's third sentence:

"...disenfranchising the majority and dominating the final ballot. This is what the Puppies said was the major flaw in the current Hugo nomination system, and what they formed to address. They have demonstrated in public how serious it is, and it's a flaw that we feel must be fixed if the integrity of the award is to be maintained. Early on participants reached consensus that any new system *must* be fair to all members.

* for this question. The term SJW, and the claims about our actions and motives, are arrant nonsense.

#365 ::: P J Evans ::: (view all by) ::: May 31, 2015, 05:45 PM:

364
That is not an improvement.

#366 ::: Shane ::: (view all by) ::: May 31, 2015, 05:55 PM:

Oh! my #356 is wrong. What I get for posting at 2am local - I quoted the question only to show which one I was referring to (and mucked it, my tablet can't quite take Making Light length threads). I think the question is phrased quite well, including down to saying "the Good Stuff". I meant my issue was with the answer.

#367 ::: junego ::: (view all by) ::: May 31, 2015, 06:14 PM:

[FAQ]

Here are some suggestions for focussing Q1 a bit.

1. Can you explain the system in plain language?
First, and importantly, this system makes absolutely no changes in how members nominate (no ranking or complex strategies required). The new system changes only the way nominations are tallied in order to create a more robust measure of which nominees have the broadest and deepest support among WSFS members. The final Hugo voting system, which actually chooses the winner, is also unchanged.

The new system narrows down the nominations proposed by members' ballots by knocking out the least supported works in repeated elimination rounds until only five (under current rules) finalists remain. Here are the basic steps as applied to one category:

a. All the nominations for all nominees are added up, just like in the current system. This number, called 'nominations' in the proposal, is one of the criteria used to determine the final nominees with the broadest support.

b. Next, one point is divided equally among all the nominees in a category on each member's ballot. If there are 2 nominees on a ballot each gets 1/2 point, if 3 nominees each gets 1/3 point, and so on.

c. All the points for each nominee from all nomination ballots are then added together. This number, called points [or whatever] in the proposal, is the second of the two criteria used to determine support.

d, Now the 'rounds' of elimination begin. Nominees are ranked from highest to lowest by their points totals (the criterion from step c.) The two nominees with the lowest points (the last two in rank) are compared to each other. The nominee with the lowest number of nominations (the criterion from step 1) is eliminated because it is the nominee with the least support.

e. After a nominee is eliminated it is removed (or ignored for the purposes of calculations) from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redivided among their remaining nominees. So if there were 4 nominees getting 1/4 point each, there are now 3 nominees getting 1/3 point each. All points are re-added for all nominees. (Note that as less supported nominees are eliminated the total points for remaining nominees increase.) New ranking is generated based on the new total points for each nominee.

f. This process (steps d., e., and f., called 'rounds' in the proposal) is repeated until the designated number of finalists remain.

#368 ::: junego ::: (view all by) ::: May 31, 2015, 07:09 PM:

#364 ::: Shane ::: (view all by) ::: May 31, 2015, 05:34 PM:

[FAQ]. Q21

I don't have an issue naming the Puppies, and accepting the name SJW*, I think our answer should be "No, this is what you said you had a problem with. We took that in good faith and addressed it head on."

I'm not intractably opposed to naming names, but... My understanding, which may be wrong, is that the FAQ becomes part of the permanent record. I'm not sure that SJW and Puppy will make a lot of sense to anyone in 10 or 20 years (if anyone even read the darned thing ;). I think it would be better to describe the situation in more generally understandable terms.

Secondarily, I really don't want to increase tensions. Regardless of who 'started it' (and imo it was the Pups and their claims are overwhelmingly bunk), we want to get this passed. If there are people who may have bought into some of the more reasonable sounding claims (they haven't liked some of the winners and their fav authors haven't made the cut, so they're frustrated) and/or used the Sad Pup slate more like a recommend list (which I believe is true for some people), we want to gain as much support as possible, or at least neutrality from that grouplet.

Maybe change it to:

"...disenfranchising the majority and dominating the final ballot. In recent years we've seen the effects of coordinated nominating and accusations of secret coordinated nominating. These instances have publicly demonstrated how serious this major flaw is in the current Hugo nomination system. It's a flaw that we feel must be fixed if the integrity of the award is to be maintained. Early on participants reached consensus that any new system *must* be fair to all members...

#369 ::: JJ ::: (view all by) ::: May 31, 2015, 11:09 PM:

junego, #362: 21. Wasn't this system designed by one side of the "culture wars" to keep the other side out of the Hugos?

I have a problem with any permutation which seems to make a tacit admission that there are "two sides" here. There aren't.

#370 ::: P J Evans ::: (view all by) ::: May 31, 2015, 11:28 PM:

360i agree - any wording which goes that way implies that the juvenile canines have a valid point. We don't want to imply that - because they don't.

#371 ::: Keith "Kilo" Watt ::: (view all by) ::: May 31, 2015, 11:46 PM:

Here's my as-yet-unformed thinking... If it comes up at the business meeting (which I actually find unlikely), we can verbally give a response that is something like #21. The real problem is that we may be maligned on some sites without our being given a chance to point out that we -aren't- making this change for political reasons. As #21 intimates, just by virtue of the fact that this was developed on ML's boards means there are those who will dismiss it out of hand. But while there's really not any chance of convincing those particular people otherwise no matter we do, my real concern is the viral nature of the internet. Are there otherwise well-meaning people who will be turned off if they are told this is political, when in fact it's not?

I guess the real decision is do we need #21 at all?

Kilo

#372 ::: Chris Battey ::: (view all by) ::: June 01, 2015, 12:57 AM:

[SHIRTS]

Did a bit more work on the composition for the seal design. It's a lot cleaner now, layout-wise, though I still haven't revised the lineart to represent the phoenix. Current draft.

I like how Jameson's design is coming along. (I really hope the disclaimer isn't necessary.) Are you planning on offering it as white-on-black (or another dark-colored shirt)? White shirts generally don't last long for me. I still recommend either offering it through Teespring (or else Cafepress) as an advance-order-only item.

[RIBBONS]

So far I have requests for 500 (150 for Jameson, 100 for Keith, 100 for JJ, 50 for ULTRAGOTHA, 100 for Adam) plus whatever I get to round out the order. Anyone else?

#373 ::: JJ ::: (view all by) ::: June 01, 2015, 02:39 AM:

My preference would be something like this:

21. Wasn't this system just designed to keep certain people out of the Hugo nomination process?
The purpose of this change to the Hugo nominations process is to prevent the ability of any slate with a minority of nominators, regardless of motivation, to dominate the final voting ballot. This change makes the process more fair for all participants.

#374 ::: Shane ::: (view all by) ::: June 01, 2015, 04:08 AM:

P J et al. Mileage obviously varies. I see it as pre-empting their best, or at least favourite, attack. You've busted your hams making a system which does just that, may as well say so.

They've declared us their foe. They don't want to save the Hugos, they want to burn them down. They will cry loudly about how we're a secret "cabal" only out to pick on them. I'd like a succinct statement I can link to address and explode that claim.

The target of such a link is the person who says "I just came in, but there's good points on both sides".

That said. I think I'm the only person who's expressed this view. Unless anyone else speaks up for it, let it drop.

#375 ::: junego ::: (view all by) ::: June 01, 2015, 07:34 AM:

Shane @374
P J et al. Mileage obviously varies. I see it as pre-empting their best, or at least favourite, attack. You've busted your hams making a system which does just that, may as well say so.
They've declared us their foe. They don't want to save the Hugos, they want to burn them down. They will cry loudly about how we're a secret "cabal" only out to pick on them. I'd like a succinct statement I can link to address and explode that claim.
The target of such a link is the person who says "I just came in, but there's good points on both sides".
That said. I think I'm the only person who's expressed this view. Unless anyone else speaks up for it, let it drop.

I think we *should* address the so-called elephant in the room. It seems disingenuous to act like we went to all this trouble because...?

OTOH I agree about toning down the rhetoric and the point about implicitly condoning the "there are two sides" meme. So my vote is to modify Q21 but definitely address the issue of why we are doing this.

#376 ::: junego ::: (view all by) ::: June 01, 2015, 07:49 AM:

JJ @369

junego, #362: 21. Wasn't this system designed by one side of the "culture wars" to keep the other side out of the Hugos?
I have a problem with any permutation which seems to make a tacit admission that there are "two sides" here. There aren't.

Others have suggested better formulations of this question, but I think it's correct to ask it. The answer to the question should make it clear that those of us (that's the collective 'us', I had nothing to do with the heavy lifting in this endeavor) who worked on this proposal aren't on any 'side'. I don't think it would be correct for us to dive into the controversy/politics directly, though, by making claims about it. We may not *feel* neutral about what has happened, but we have *acted* as neutrally as possible in creating this new nomination system.

#377 ::: David Harmon ::: (view all by) ::: June 01, 2015, 07:54 AM:

Chris Battey #372: Maybe a list of people who want a ribbon for themselves, but are out-of-context for distributing them? (Here!)

Shane #374: The problem is that the "war" is one they created. They want us to buy into it, because then they can recruit people to "their team". We want to undercut their dualism, and refuse to dignify them as "enemies".

This falls into the topic of tribal definitions -- we want to give their "tribe" as little recognition as possible. We face a constant temptation to reify them as "our enemies", because that seems to strengthen our own tribal boundaries (and we're feeling kinda sensitive about those lately). The problem is, it's actually reinforcing their tribal boundaries. At this, point, SFF has joined the mainstream, and our proper boundaries are those we want society at large to defend.¹ So we want to cast this not as "their tribe against our tribe", but as "marauders against civilization".

¹ Including our "interest-group" boundary, but that's no longer a sharp line marking an single tribe.

#378 ::: Jameson Quinn ::: (view all by) ::: June 01, 2015, 08:00 AM:

Isn't this just an attempt by one clique to keep out another?

This system was designed to keep any significant group from being shut out of the list of Hugo finalists. That means a minority should not win more than their fair share of slots, as happened in 2015. It also means that a minority should not win less than their fair share. Under this system, even an 80% majority should not be able to keep the remaining 20% from getting a finalist they support. If some hypothetical cabal were trying to maintain it's dominance, this would be about the last system they'd come up with, because it works exactly against that purpose.

#379 ::: Jameson Quinn ::: (view all by) ::: June 01, 2015, 08:01 AM:

In the suggested FAQ#21 I just posted, the third and last "should" should be a "would".

#380 ::: David Harmon ::: (view all by) ::: June 01, 2015, 08:04 AM:

Chris Battey #372: This looks better -- indeed, with the simpler layout, even the phoenix's flames are more recognizable. But I think the shield needs to be smaller, with "softer" edges (that is, not a box of heavy, mostly-straight lines).

#381 ::: David Goldfarb ::: (view all by) ::: June 01, 2015, 11:40 AM:

I like Jameson's @378, but note that the "it's" should be "its".

#382 ::: junego ::: (view all by) ::: June 01, 2015, 12:33 PM:

Jameson Quinn @378

[FAQ]
Q21

Isn't this just an attempt by one clique to keep out another?
This system was designed to keep any significant group from being shut out of the list of Hugo finalists. That means a minority should not win more than their fair share of slots, as happened in 2015. It also means that a minority should not win less than their fair share. Under this system, even an 80% majority should not be able to keep the remaining 20% from getting a finalist they support. If some hypothetical cabal were trying to maintain it's dominance, this would be about the last system they'd come up with, because it works exactly against that purpose.

I like this one, too. It's more concise, yet covers the elephant ("in 2015"), emphasizes neutrality by pointing out that a 20% minority deserves representation, and refers to the secret cabal in an appropriately dismissive way while emphasizing that the new system prevents such a group from actually succeeding. Good wordsmithing, Jameson!

#383 ::: Chris Battey ::: (view all by) ::: June 01, 2015, 12:58 PM:

David Harmon @ 372, 378:

When we arrange a time to distribute the bundles of ribbons, anyone who just wants one to wear can show up too. Besides, it'll be good to meet our collaborators in person! (Maybe we should set up a specific gathering early in the con?)

The blocky shield is another artifact from hewing too closely to the original seal; I agree that it jumps out against the curved/wavy lines all around it. I'll see what it looks like with a softer design.

#384 ::: Cally Soukup ::: (view all by) ::: June 01, 2015, 01:03 PM:

Me, I'm hoping that one of my meatspace friends will pick up a ribbon or two for me, since I won't be going to Sasquan.

#385 ::: P J Evans ::: (view all by) ::: June 01, 2015, 01:06 PM:

382
I object to the idea that a minority, even one of 20%, deserves representation. No, that's not part of the Hugo, which is about what people like most.
That is, though, the precise crap that we're dealing with this year: a minority that decided to force its opinions on everyone else. (That they're showing themselves to be poor judges of quality is on them.)

#386 ::: David Langford ::: (view all by) ::: June 01, 2015, 01:48 PM:

Quoted in #382: If some hypothetical cabal were trying to maintain it's dominance ... well phrased, but please lose the apostrophe!

#387 ::: Jameson Quinn ::: (view all by) ::: June 01, 2015, 03:16 PM:

@386: I'm sorry about that apostrophe; now your pointing it out, it grate's on my nerves too. (Just kidding!)

@385: The proposed FAQ text says nothing about "deserving"; that was junego speaking for themself. (And yes, spellcheck, "themself" is a word; it's the neuter singular reflexive pronoun.)

#388 ::: junego ::: (view all by) ::: June 01, 2015, 03:37 PM:

P J Evans @385

I object to the idea that a minority, even one of 20%, deserves representation. No, that's not part of the Hugo, which is about what people like most.
That is, though, the precise crap that we're dealing with this year: a minority that decided to force its opinions on everyone else. (That they're showing themselves to be poor judges of quality is on them.)

I'm sorry if some of the words in my comment on Jameson's proposed language got your hackles up. There was no deeper message intended other than that a fair system should not let the majority completely overwhelm a minority any more than letting the minority dictate to the majority.

What is your opinion wrt the proposed language and whether or not we should even include this question in the FAQ?

#389 ::: David Dyer-Bennet ::: (view all by) ::: June 01, 2015, 04:01 PM:

I think there has to be some sort of public answer to the obvious questions. Currently the FAQ is the place where we're putting that sort of thing, and that seems reasonable to me.

The obvious puppy talking points, therefore, need to be addressed in the FAQ.

If it helps anybody, maybe we might want to specifically label questions that are about the political issue rather than about the proposal itself? If nothing else, those won't confuse people in the future (if people don't know enough to recognize immediately that certain questions are responses to contemporary political issues).

#390 ::: Shane ::: (view all by) ::: June 01, 2015, 05:17 PM:

#378 ::: Jameson Quinn ::: less the offending apostrophe.

Loveley. That answers my concern.

#391 ::: David Harmon ::: (view all by) ::: June 01, 2015, 05:23 PM:

Chris Battey #383: The "context" problem I mentioned is that I won't be at the con... It'd still be nice to get a ribbon.

If you want to discuss options by e-mail, my address is my name as above, separated by an underscore, at Earthlink.

#392 ::: Buddha Buck ::: (view all by) ::: June 01, 2015, 06:28 PM:

[FAQ Q21]

My only concern with Jameson's most recent wordcrafting is the suggestion of a group having a "fair share". To me, it supports the idea of parties/cabals splitting the nominations between them.

#393 ::: P J Evans ::: (view all by) ::: June 01, 2015, 06:46 PM:

392
[FAQ Q21]
This, yes. The reason for this particular proposal is not to make sure minorities get their fair share, but to make sure that they can't get all their choices on, and block the majority's choices. (Which is what already has happened.)

#394 ::: Chris Battey ::: (view all by) ::: June 01, 2015, 08:08 PM:

Ah, understood. For those of you not at the con, I will happily mail small amounts of ribbons out once I have them.

I have created a Google form here to collect requests for mailed ribbons. The data gathered by the form goes into a spreadsheet where only I have viewing permissions. Nevertheless, if you're not comfortable putting your information there you can email me instead - my mail domain is gmail dot com and my username there is darthparadox.

#395 ::: Cally Soukup ::: (view all by) ::: June 01, 2015, 08:25 PM:

Thank you Chris; I really appreciate that!

#396 ::: David Harmon ::: (view all by) ::: June 01, 2015, 08:52 PM:

Chris Battey #394: Thank you! Note sent.

#397 ::: JJ ::: (view all by) ::: June 01, 2015, 09:45 PM:
21. Isn't this just an attempt by one clique to keep out another?
This system was designed to keep any significant group from being shut out of the list of Hugo finalists. That means a minority should not win more than their fair share of slots, as happened in 2015. It also means that a minority should not win less than their fair share. Under this system, even an 80% majority should not be able to keep the remaining 20% from getting a finalist they support. If some hypothetical cabal were trying to maintain it's dominance, this would be about the last system they'd come up with, because it works exactly against that purpose.
I strongly object to this wording: the part about "fair share", and the way that it's worded defensively.
#398 ::: Cally Soukup ::: (view all by) ::: June 01, 2015, 10:39 PM:

Instead of "fair share", how about "proportional number"?

#399 ::: Keith "Kilo" Watt ::: (view all by) ::: June 01, 2015, 10:47 PM:

[FAQ21]

I think this is the key message of #21:

"But any slate, of any sort, be it a Sad Puppy or a Happy Kitten of Social Justice, breaks the Hugo Award because a small percentage of voters can effectively prevent any other nominee from appearing on the final ballot. This is a major flaw in the Hugo nomination system, and it is a flaw that must be fixed if the integrity of the award is to be maintained. Politics should play no role whatsoever in whether a nominee is Hugo-worthy or not."

As others have noted, it's not really about giving anyone their fair share of the final ballot; it's about not letting any group -prevent- someone from making the final ballot. That, really, is the problem with slates and why we had to fix it. So I think however we word this FAQ should stay focused on that aspect. While related in some sense, the two perspectives aren't really saying the same thing.

Kilo

#400 ::: Lydy Nickerson ::: (view all by) ::: June 01, 2015, 11:24 PM:

I dislike the attempts to wash out some of the detail for Q21. It is true that we are doing this because of the Sad Puppy Slate, and it is true that they will accuse us of having done so to persecute them, and it is guaranteed that they will point to his particular venue, Making Light, as proof that this initiative is really all about oppressing them. And I think we should actually address this. I like the initial formulation of the question and answer.

I should also note that this particular bias of mine got me and mine in boat loads of trouble that we probably didn't need, 15 years ago, when we tried to remake Minicon. That I retain this bias may well be an example of failing to learn. But I really hate being mealy-mouthed about these sorts of things.

#401 ::: Joshua Kronengold ::: (view all by) ::: June 01, 2015, 11:30 PM:

I'd like 50 ribbons that said "E Pluribus Hugo", if that's being made.

Re shirt designs, I'm not interested in the seal design, if other are, go them.

It occurs to me that much better than a rocket made of Hugo winners would be a rocket made of first names. After all, the whole point of "E Pluribus Hugo" as a proposal title is that out of all of -us- (the many) comes a single Hugo winner (per category). We wouldn't want to use full or last names as they are too personal; too identifiable -- but just using the most common X male and female first names in alphabetical order [mostly], I think, would get the message across nicely.

#402 ::: JJ ::: (view all by) ::: June 02, 2015, 12:49 AM:

Joshua Kronengold, #401: It occurs to me that much better than a rocket made of Hugo winners would be a rocket made of first names. After all, the whole point of "E Pluribus Hugo" as a proposal title is that out of all of -us- (the many) comes a single Hugo winner (per category). We wouldn't want to use full or last names as they are too personal; too identifiable -- but just using the most common X male and female first names in alphabetical order [mostly], I think, would get the message across nicely.

I'd want a t-shirt that will be meaningful anywhere I wear it, including at Worldcon when E Pluribus Hugo is a distant memory.

Plus, putting the names of the people who proposed and supported the rules changes strikes me as saying that it's about the people, rather than what it's really about, which is the award.

#403 ::: Tom Whitmore ::: (view all by) ::: June 02, 2015, 01:05 AM:

(nitpick) Also, please note that the quoted bit of 21 in 397 indicates that the original uses the wrong form of "its"/"it's": it's the possessive, or "its", rather than "it is", or "it's", that should be used there. (/nitpick) (That kind of misuse can get me to vote against something I'm waffling on, because it indicates the authors aren't paying attention to detail -- and detail is crucial on this kind of motion.)

#404 ::: Soon Lee ::: (view all by) ::: June 02, 2015, 04:15 AM:

[FAQ21]

I think it's either two related questions we are trying to address or one question with a two part answer. Here's my rough attempt at it using some of the previous phrasing. Thoughts?

Isn't this just an attempt by one group to keep out another?

No. This system is designed to fix the flaw in the current Hugo nominating process that allows a small percentage of nominators to dominate the final ballot as happened in 2015 with the Sad Puppies & Rabid Puppies slates. This is a major flaw that must be fixed if the integrity of the award is to be maintained. This system dilutes the effect of slate or bloc nominating so if some hypothetical cabal were trying to maintain its dominance, this would be about the last system they'd come up with, because it works exactly against that purpose.

While it is true that much of the discussion for this system occurred on Teresa and Patrick Nielsen Hayden’s “Making Light” discussion board, Teresa and Patrick Nielsen Hayden had no real input in the discussions of the system. Those of us who worked on the system were very clear that our goal was not to keep anyone off of the Hugo ballot, and that any system which specifically targets any type of nominee is inherently wrong and unfair. Our goal was to create a system resistent to slate or bloc nominating, a system that allows a consensus final ballot to emerge from all the nominations submitted.

#405 ::: Shane ::: (view all by) ::: June 02, 2015, 06:47 AM:

[FAQ21]

#399 ::: Keith "Kilo" Watt :::

Hi Kilo, I agree that is the key message of #21, except that I do want to explicitly state we cover the possibility of informal or hidden groups.

Soon Lee @404 that's lovely. And yes I think you're right.

#406 ::: Tammy Coxen ::: (view all by) ::: June 02, 2015, 07:33 AM:

I'll also take 50 ribbons to hand out.

#407 ::: David Goldfarb ::: (view all by) ::: June 02, 2015, 08:26 AM:

Tom@403: You do realize that you're the third person to complain about that? (After me and Dave Langford.) I mean, if we're going to talk about attention to detail and all....

#408 ::: David Harmon ::: (view all by) ::: June 02, 2015, 08:41 AM:

Soon Lee #404: I like the wording, but there are a couple of run-on sentences in there, that really need to be broken up. Both sentences starting with "This system..." And "While it is true..." duplicates PTNH's names in full. My suggestions:

"E Pluribus Hugo" is designed to fix a flaw in the current Hugo nominating process, which became prominent in 2015 due to its use by the Rabid Puppies & Sad Puppies slates. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories.

[That's also a bit sharper than what you had, though I could have said "misuse". However, it also focuses on what exactly was wrong with their actions.]

"E Pluribus Hugo" dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, this would be about the last system they'd come up with.

Much of the discussion for this system did happen on the “Making Light” discussion board, owned by Teresa and Patrick Nielsen Hayden. However, the owners had no real input in the discussions of the system.

#409 ::: Xopher Halftongue ::: (view all by) ::: June 02, 2015, 09:11 AM:

Soon Lee 404: I like it. I'd suggest adding a comma after 'ballot' in the first full sentence ("...a small percentage of nominators to dominate the final ballot, as happened in 2015...") and after 'nominating' in the last sentence of that paragraph ("This system dilutes the effect of slate or bloc nominating, so if...").

#410 ::: Cally Soukup ::: (view all by) ::: June 02, 2015, 10:04 AM:

Much of the discussion for this system did happen on the “Making Light” discussion board, owned by Teresa and Patrick Nielsen Hayden. However, the owners had no real input in the discussions of the system.

Did happen? Why not just "happened" or "took place"?

#411 ::: David Harmon ::: (view all by) ::: June 02, 2015, 10:35 AM:

Cally Soukup #410: Remnant of the implied reference to the conflict implied by "While it is true".

#412 ::: Tom Whitmore ::: (view all by) ::: June 02, 2015, 12:39 PM:

No, David Goldfarb @307, I didn't. Because I haven't been reading this thread other than very lightly, and said so at least once before. And it wasn't corrected in what I saw.

I firmly believe it's better to be corrected three times than not at all. Especially when it's me being corrected. Thank you for letting me know that others are as nitpicky as I am.

#413 ::: Soon Lee ::: (view all by) ::: June 02, 2015, 03:09 PM:

[FAQ21]

Thanks for the feedback; I've incorporated some of it & tweaked it some more. Here's the revised version for comment.


Isn't this just an attempt by one group to keep out another?
No. This system is designed to fix the flaw in the current Hugo nominating process that allows a small percentage of nominators to dominate the final ballot, as happened in 2015 with the Sad Puppies & Rabid Puppies slates. This is a major flaw that must be fixed if the integrity of the award is to be maintained. "E Pluribus Hugo" dilutes the effect of slate or bloc nominating, so if some hypothetical cabal were trying to maintain its dominance they would not propose this system. This system works against that purpose.

It is true that much of the discussion for this system took place at Teresa and Patrick Nielsen Hayden’s “Making Light” discussion board. However, the owners had no real input in the discussions of the system. Those of us who worked on the system were very clear that our goal was not to keep anyone off of the Hugo ballot, and that any system which specifically targets any type of nominee is inherently wrong and unfair. Our goal was to create a system resistant to slate or bloc nominating, a system that allows a consensus final ballot to emerge from all the nominations submitted.

#414 ::: Soon Lee ::: (view all by) ::: June 02, 2015, 03:19 PM:

I'm off to work so won't be replying until tonight (New Zealand time). I'm also amused that this FAQ21 version was first posted as comment 404.


(I find amusement where I can)

#415 ::: Keith "Kilo" Watt ::: (view all by) ::: June 02, 2015, 03:34 PM:

JJ@160:
Have you an e-mail you can rot13, or are you one of the numerous Keith Watts on Facebook?

So sorry I missed this one! I think people have found that if they Google my handle here they get my work email fairly readily. If you have problems, let me know!

Kilo

#416 ::: abi ::: (view all by) ::: June 02, 2015, 03:39 PM:

I can connect people discreetly if they need it. If your email is not the one you use on your post entries, feel free to reach me at this username at this domain.

#417 ::: Steve Halter ::: (view all by) ::: June 02, 2015, 04:22 PM:

Caught up yet again. I rather like the simple Quinn shirt without the disclaimer.

#418 ::: Keith "Kilo" Watt ::: (view all by) ::: June 02, 2015, 05:03 PM:

[FAQ]

I've been working with updating the FAQ, so here is a new entry based on #67,#70, and #117. Jameson, could you please check this for accuracy? There were some parts in your #70 that I think weren't clear, but your comments in #117 (if I'm interpreting them correctly) cleared it up. Let me know...

Kilo
==================================
3. Surely strategy can make some difference, can't it?

Here's a specific example:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. Suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all could submit nomination ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its nominations compared in the Elimination Phase (you have no way to increase the number of nominations that Castle receives if it is selected for the Elimination Phase). If you push it up to 5th place, it can still be eliminated by Tentacles if Castle has fewer nominations. If you could somehow push it all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters to such a high level of precision before the election. Maybe Rocket isn’t so popular after all and needs your nominations to get over the top. Maybe Tentacles has a broad enough base of support that your nominations for Castle aren’t enough to eliminate it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire; and only one, highly-specific way for it to work.

So nominating what you really think is Hugo-worthy is your best strategy.


#419 ::: Annie Y ::: (view all by) ::: June 02, 2015, 05:16 PM:

Keith "Kilo" Watt @ #418

[FAQ]

Well - you may also add that this is not different from the current process where the choice to add Tentacles to your ballot may give it this one additional vote that allows it to inch over Castle and send Castle off the ballot while leaving it out would have made Castle and Tentacles draw and put both on the ballot.

Still catching up with what else happened last week in this thread so will probably be posting again in a bit.

#420 ::: David Harmon ::: (view all by) ::: June 02, 2015, 05:36 PM:

Soon Lee #413: Looks good!

Keith "Kilo" Watt #418: How about: "to such a high degree of precision" --> "so precisely".

#421 ::: felice ::: (view all by) ::: June 02, 2015, 06:48 PM:

Keith "Kilo" Watt @418: "Maybe Rocket isn’t so popular after all and needs your nominations to get over the top."

Would it be worth mentioning the effect of large scale use of this strategy? Ie, if lots of people assume Rocket is safe and leave it off their ballot in order to give more help to other works, then this strategy could cause Rocket to lose. Or it could drop Rocket's points enough that it ends up competing with and eliminating Castle, even though both actually had more nominations than Tentacles.

#422 ::: P J Evans ::: (view all by) ::: June 02, 2015, 07:08 PM:

471
That would seem to require coordination - which would be a bad idea. Remember that you aren't supposed to know what other people are putting in the pot.

#423 ::: Michael I ::: (view all by) ::: June 02, 2015, 07:15 PM:

PJ Evans@422

It wouldn't necessarily require coordination. Absent actual coordination, it would require a substantial number of nominators coming to the same conclusion about the likely preferences of the other nominators.

#424 ::: Keith "Kilo" Watt ::: (view all by) ::: June 02, 2015, 07:25 PM:

[PROPOSAL/FAQ]

All: Okay, I think I've got the next drafts of the proposal and FAQs ready to go. I'll post each in a separate message. Notice that there has been quite a bit of renumbering thanks to the many suggestions here, so when you refer to a number, please use the new ones below.

Thanks much!

Kilo

#425 ::: P J Evans ::: (view all by) ::: June 02, 2015, 07:32 PM:

423
That still requires information about what a large number of other people are doing - and that isn't easily available. (It shouldn't be available at all.)

#426 ::: Keith "Kilo" Watt ::: (view all by) ::: June 02, 2015, 07:35 PM:

[PROPOSAL]

Short Title: E Pluribus Hugo (Out of Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations), section 3.9 (Notification and Acceptance), and section 3.11 (Tallying of Votes) as follows:

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the process described in section 3.A.

Insert new section 3.A after Section 3.8 as follows:
Section 3.A Nomination Elimination Procedure

3.A.1: The finalist selection process shall be conducted in rounds consisting of three phases:

(1) Calculation Phase: For each category, the total number of nominations (the number of ballots on which each nominee appears) shall be tallied for each surviving nominee. Next, a single “point” shall be assigned to each category on each nomination ballot, and that point shall be divided equally among all of the nominations on that ballot in that category that were not eliminated previously. Finally, all points from all nomination ballots shall be totaled for each nominee in that category. These two numbers, total nominations and total points, shall be used in the Selection and Elimination Phases.

(2) Selection Phase: Each round shall start by selecting the two nominees with the lowest point totals for the Elimination Phase. (See 3.A.3 for ties.)

(3) Elimination Phase: Nominees chosen in the Selection Phase shall be compared, and the nominee with the fewest total nominations shall be eliminated and removed from all ballots for the Calculation Phase of all subsequent rounds.

3.A.2: The phases described in 3.A.1 are repeated in order until the number of finalists specified in 3.8.1 remain. If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then instead no nominees will be eliminated this round, and all remaining nominees shall appear on the final ballot, extending it if necessary.

3.A.3: Ties shall be handled as described below:
(1) During the Selection Phase, if two or more nominations are tied for the lowest point total, all such nominees shall be selected for the Elimination Phase.
(2) During the Selection Phase, if one nominee has the lowest point total and two or more nominees are tied for the second-lowest point total, then all such nominees shall be selected for the Elimination Phase.
(3) During the Elimination Phase, if two or more nominees are tied for the fewest number of nominations, the nominee with the lowest point total at that round shall be eliminated.
(4) During the Elimination Phase, if two or more nominees are tied for both fewest number of nominations and lowest point total, then all such nominees tied at that round shall be eliminated.

3.A.4: After the initial Award ballot is generated, if any finalist(s) are removed for any reason, the nomination elimination system shall be rerun as though the removed finalist(s) had never been nominees. None of the remaining original finalists who have been notified shall be removed as a result of this rerun. The new finalist(s) shall be merged with the original finalists, extending the final ballot if necessary.

Section 3.9: Notification and Acceptance.
3.9.1 Worldcon Committees shall use reasonable efforts to notify the nominees finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each nominee finalist shall be asked at that time to either accept or decline the nomination. If the nominee a finalist(s) declines the nomination, that nominee finalist(s) shall not appear on the final ballot. The procedure for replacement of such finalist(s) is defined in subsection 3.A.4.

Section 3.11: Tallying of Votes.
3.11.4: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. During the same period, the results of the last ten rounds of the selection process for each category (or all the rounds if there are fewer than ten) shall also be published.

Submitted by: Keith “Kilo” Watt, Jameson Quinn, Tammy Coxen, Yoana Yotova, Joshua Kronengold, Christopher Battey, David Gallaher, Adam Tilghman, CJ Cabourne, Steven Halter, P.J. Evans, David Goldfarb, Seth Gordon, Ginger Tansey, Steve Wright, Catherine Faber, Andy Holloway, Duncan J. Macdonald, Claudia Beach, Derry Earnshaw, Jason Skiles, Soon Lee, David Harmon, Lydy Nickerson, Abigail Sutherland, Lee Billings, Oskari Rantala, Seth Breidbart, Chris Suslowicz, Fragano Ledgister, Lori Coulson, Jeffry Herman, Mark Shier, Buddha Buck, Lenore Jean Jones, David Langford, Christopher Hatton

Commentary: “E Pluribus Hugo” – the name of this proposal is also its goal: recognizing the many opinions within fandom as to what nominees might be worthy of the Hugo. It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members: They still list the nominees, unranked, they feel are Hugo-worthy (up to the maximum permitted) in any categories they choose, just as they have in the past. In the past, we have counted the number of nominations each nominee received, and the top five nominees were put on the final ballot. However, because SF fandom typically nominates a variety of different nominees, it was easy for an organized slate to make it so that no other nominees made the final ballot.

Using this system, fandom isn't penalized for nominating a wide variety of nominees. If you nominate something that ends up not having a chance to make the final ballot, then your remaining choices automatically get more of your support instead of just being wasted. In other words, you can safely nominate anything you feel is Hugo-worthy. If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot. In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which nominees should be voted on to be the final Hugo winner.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no minority of members – of any sort – can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Award.


[NB: If I have forgotten a signer, please let me know!]

#427 ::: Keith "Kilo" Watt ::: (view all by) ::: June 02, 2015, 08:03 PM:

FAQ’s:

1. Can you explain the system in plain language?
First, and most importantly, E Pluribus Hugo makes absolutely no changes in how members nominate (no ranking or complex strategies are required). The new system changes only the way nominations are tallied in order to create a more robust measure of which nominees have the broadest and deepest support among WSFS members. The final Hugo voting system, which actually chooses the winner, is completely unchanged.

The new system narrows down Worldcon members' nominations by knocking out the least supported works in repeated elimination rounds until only five (under current rules) finalists remain. Here are the basic steps, as applied to one category:

a. All the nominations for all nominees (that is, the number of ballots on which each nominee apears) are added up, just like in the current system. This number, called "total nominations" in the proposal, is one of the criteria used to determine the final nominees with the broadest support.

b. Next, one point is divided equally among all the nominees in a category on each member's ballot. If there are two nominees on a ballot in a category, each gets 1/2 point; if there are three nominees, each gets 1/3 point, and so on.

c. All the points for each nominee from all nomination ballots are then added together. This number, called "point total" in the proposal, is the second of the two criteria used to determine the final nominees with the broadest support.

d. The nominees with the lowest point totals are compared to each other. The nominee with the fewest number of nominations is eliminated, because it is the nominee with the least support.

e. After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed among their remaining nominees. So, if there were four nominees getting 1/4 point each, there are now three nominees getting 1/3 point each, and so on. All points are re-added for all nominees. As less-supported nominees are eliminated, the total points for remaining nominees increase.

This process is repeated until the designated number of finalists (five in the curent version of the constitution) remain.


2. E Pluribus Hugo?
The designers felt this name accurately reflects what the Hugo nomination system should be trying to achieve. Fandom is the Many. Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in SF. This is a defining characteristic of fandom, and this system embraces it. Out of all of these many tastes in nominees, this system allows a single set of nominations to slowly emerge as the most popular candidates for the Hugo Award: “Out of the Many, a Hugo.”


3. But surely that’s not what the system is formally called?
Okay, sure. Formally, in the context of election theory, this system would be called “single divisible vote with least popular elimination” or SDV-LPE for short. E Pluribus Hugo is both cooler and geekier, as everything is better with a little Latin.


4. Why are there Selection and Elimination Phases? Couldn’t you simplify the system by doing it all in one step per elimination?
The Selection Phase looks for nominees with the lowest point total, and the Elimination Phase eliminates the one with the fewest nominations. Looking at the two different measures of popularity adds robustness to the system. With non-slate nominations, as occurred with most of the Hugo Awards to date, the nominees with the fewest points will tend to be the same ones as those with the fewest number of nominations. But while it might be possible for strategic blocs to manipulate one of these two measurements (as occurred in 2015), it would be very difficult to manipulate both in the same direction. A strategy that focuses on nominations for the Elimination Phase (such as bloc voting) will be stymied by the Selection Phase; a strategy that focuses on points for the Selection Phase (such as nominating only one nominee) will turn out to be counterproductive in the Elimination Phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.


5. Shouldn’t I just nominate one work if I want to give it the best chance to win?
The system was crafted specifically to make most strategies irrelevant. For example, suppose you wanted to give your full support behind one nominee that you really feel strongly about. You can do that, particularly if there isn’t anything else you feel is Hugo-worthy. On the other hand, if you nominate four other things that you think might be Hugo-worthy, you aren't hurting your favorite. This is because if your other nominees get eliminated, then your full support will go towards the remaining nominee, just as if you had never nominated the others. If enough people agree with your other four, some of them might make it, too. If enough people don't agree with your favorite, there's nothing you can do to get it on the final ballot, because you will never have more than one point. In general, the best strategy is simple: nominate as many nominees as you feel are worthy.


6. Surely strategy can make some difference, can't it?
Under E Pluribus Hugo there’s no good reason to use any strategy at all. The important thing to remember is that nothing changes in how you nominate. If you think a nominee is Hugo-worthy, then nominate it. That’s all. There’s no need to rank your choices at the nomination stage, and there’s no good reason not to nominate something you think might be Hugo-worthy. Here's a specific example:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. Suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all could submit nomination ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its nominations compared in the Elimination Phase (you have no way to increase the number of nominations that Castle receives if it is selected for the Elimination Phase). If you push it up to 5th place, it can still be eliminated by Tentacles if Castle has fewer nominations. If you could somehow push it all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters so precisely before the election. Maybe Rocket isn’t so popular after all and needs your nominations to get over the top. Maybe lots of people will assume Rocket is safe and cause it to end up losing as a result of leaving it off of their ballots. Maybe Tentacles has a broad enough base of support that your nominations for Castle aren’t enough to eliminate it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire, and only one, highly-specific way for it to work.

Nominating what you think is Hugo-worthy really is your best strategy.


7. Isn't it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws. The objective is to choose a system whose flaws are not in an area of concern to the electorate.


8. What are E Pluribus Hugo’s flaws?
In rare cases, it is possible that eliminating both members of a tie could change the final ballot slightly from what it would be if the tie were broken. The change is usually in the least popular of the finalists, and requires the two nominees in question be very close in popularity. Hundereds of simulations were run using real and created data sets to help decide which tie-breaking methods would maintain our goal of supporting the wide range of opinions within fandom. It turned out that all the simualtions showed that there was almost no difference in outcomes, no matter how we broke ties. We have chosen, therefore, to break ties in a manner consistent with section 6.4 of the Worldcon constitution. There are a number of additional tie breakers that could be used if it were deemed necessary in the future; however, simulations that have been run by the designers of this system show that it really shouldn’t be required.


9. What are E Pluribus Hugo’s benefits?
Simply put, it reduces the power of bloc voting without eliminating the chance that nominees that do appear on slates will make it to the final ballot. Conversely, it makes it very difficult for slates to prevent non-slate nominees from appearing on the ballot.


10. How does this system eliminate slate or bloc voting?
It doesn't, exactly, nor should a nominee be automatically eliminated just because it appears on a slate. On the other hand, any slate which nominates a full set of five nominees will find that each of its nominations only count 1/5 as much. With “non-slate” nominating, some of your nominees will be slowly eliminated, so your remaining nominees get more and more of your support. Since slate nominees tend to live or die together, they tend to eliminate each other until, in general, only one slate nominee remains. With a large enough support behind the slate (five times as much), the slate may still sweep a category; however, if that many voters support the slate, they arguably deserve to win, and no fair and unbiased system of nomination will prevent that. The answer in that case is, simply, to increase the general pool of voters. Regardless, with E Pluribus Hugo, slates will never receive a disproportionate share of the final ballot, as occurred in the 2015 Hugos.


11. Couldn't supporters of slates just recommend a single nominee for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate has enough supporters that is certainly a viable possibility – it’s also completely fair. It does not force all other nominees off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn't a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a nominee deserves a Hugo nomination, then it should, in fact, be represented on the final ballot.


12. What if there are multiple slates (slate wars, “parties”, etc.)?
As with a single slate, the nominees on each slate will tend to eliminate each other until only the most popular one remains. The end result is that even multiple slates are unable to sweep the nominations.


13. What happens if a broadly popular nominee is nominated by a group of unrelated people?
If it is broadly popular, the system will still select that nominee for the final ballot.


14. What happens if a broadly popular nominee also appears on a slate?
Even if it is on a slate, if the nominee garners support from individuals then the system will select that nominee for the final ballot. In general, slates neither help nor hurt any given nominee.


15. What happens when there are a lot of nominees with no obvious favorites and nomination slates are introduced?
Even in this case, we were unable to find a simulation in which no non-slate nominees appeared on the final ballot. Slate nominees did receive a larger proportion of nomination slots than they did otherwise, however, again, this could be considered a fair and valid result. If there was no general favorite, then members really had no collective preference.


16. How do the results of this system compare to the results under the current nomination system?
Statistical tests showed that this system and the current system overlap in about (conservatively) 4.5 out 5 candidates. In the absence of slates, it generally gives identical results to the current nomination system.


17. Why re-run the system if someone declines a nomination? Why not just take the nominee that was eliminated in the next to last round?
We can do that, and in fact have an amendment ready if that’s what the people decide they want to do. Our system has only one major difference between it and taking the “6th place” nominee. Under our proposal, if a slate nominee makes the final ballot and that nomination is declined, that nomination will tend to be replaced by another nominee from the slate. In taking the “6th place” nominee, the declined nomination will tend to be replaced by a non-slate nominee. Either way works, but since our goal is to not prevent any section of fandom from being represented on the final ballot, the method of handling declined nominations outlined the proposal seemed more in line with that philosophy.


18. I think we should just increase the number of nomination slots on the final ballot to a larger number (for example, 6), and decrease the number of slots a voter can vote for to a smaller number (for example, 4). Wouldn't that be simpler and easier?
Unfortunately, this simply means that the largest slate will receive four of the nominations and the next largest will receive the remaining two. It doesn't solve the problem of forcing nominees off the ballot that had a chance to win the final election. In general, we want fandom to nominate as many nominees as they feel are Hugo-worthy, since under E Pluribus Hugo there is no strategic reason not to do so. For this reason, most of the designers of this system would prefer that members not be limited to nominating only four nominees. Keep in mind, however, that E Pluribus Hugo will work with this (or most any other) change as well, so one does not preclude the other.


19. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
This could work. The problem is that then you would have a small group of people serving as literal gatekeepers to the Hugo Awards. In spite of the word on the Internet, this has never been the case in the past. Establishing it would prove those groups who believed it existed in the past would then be correct. Ultimately, human judgement is fallible. The fairness of a committee’s decisions would forever be subject to opinion, and the prestige of the Hugo Awards would forever be tarnished.


20. I think we should use [insert other mathematical voting system].
We considered essentially every applicable type of voting system currently in the literature, guided by two experts in the field. It should be kept in mind, however, that the goals and requirements for choosing a set of representatives in a political situation are different from those for choosing a set of Hugo finalists. Some of these systems do, in fact, have positive properties that speak for them. None of them were as simple or as intuitive as E Pluribus Hugo, yet E Pluribus Hugo meets all of the stated goals for a Hugo nomination system.


21. Won’t E Pluribus Hugo be complicated to code and implement?
One of our non-experts coded a full simulator for the system in a matter of days. A full web-based app would not be much more difficult to handle.


22. Isn't this just an attempt by one group to keep out another?
No. E Pluribus Hugo is designed to fix the flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the award is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nomintions. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.

It is true that much of the discussion for this system occurred on the “Making Light” discussion board, owned by Teresa and Patrick Nielsen Hayden, however they had no real input in the discussions of the system. Those of us who worked on the system were very clear that our goal was not to keep anyone off of the Hugo ballot, and that any system which specifically targets any type of nominee is inherently wrong and unfair. Our goal was to create a system resistent to slate or bloc nominating, a system that allows a consensus final ballot to emerge from all the nominations submitted.

#428 ::: Jameson Quinn ::: (view all by) ::: June 02, 2015, 08:14 PM:

@418: That looks pretty good. Let me suggest some edits:

[FAQ]
3. Surely strategy can make some difference, can't it?

Yes; it is impossible to create a voting system where strategy never makes a difference. But in this system, there's probably no good way to know ahead of time that a strategy is safe and effective. Here's a specific example:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. Suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all could submit nomination ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its nominations compared in the Elimination Phase (you have no way to increase the number of nominations that Castle receives if it is selected for the Elimination Phase). If you push it up to 5th place, it can still be eliminated by Tentacles if Castle has fewer nominations. If you could somehow push its point total all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters to such a high level of precision before the election. Maybe Rocket isn’t so popular after all and needs your nominations to get over the top. Maybe Rocket is popular, but too many other voters strategically ignore it just as you do, so it loses. Maybe Tentacles has a broad enough base of support that your nominations for Castle aren’t enough to eliminate it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire; and only one, highly-specific way for it to work.

So nominating what you really think is Hugo-worthy is your best strategy.

#429 ::: Jameson Quinn ::: (view all by) ::: June 02, 2015, 08:18 PM:

@426: I suggest:

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no group of members – of any sort, minority or majority – can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Award.

#430 ::: Jameson Quinn ::: (view all by) ::: June 02, 2015, 08:19 PM:

@427: That looks good, except for the changes I suggest in @428. Specifically, I'm satisfied with that version of FAQ#22.

#431 ::: Soon Lee ::: (view all by) ::: June 02, 2015, 08:26 PM:

I think we are in excellent shape. I would want to do a closer read to ensure consistent usage of "nominee" & "finalist".

[PROPOSAL]
3.11.4: I'm still not entirely clear what this is trying to say: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee

Commentary: fandom slowly converges to a consensus as to which nominees should be voted on to be the final Hugo winner. -> delete the word "final"?

[FAQ]

Q8: Hundereds of simulations were run -> Hundreds
Q8: It turned out that all the simualtions -> simulations

Q22: E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nomintions. -> nominations
Q22: Our goal was to create a system resistent -> resistant

#432 ::: Keith "Kilo" Watt ::: (view all by) ::: June 02, 2015, 09:41 PM:

Soon Lee@431:

3.11.4: I'm still not entirely clear what this is trying to say: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee

This is part of the original constitution, so isn't a part we've messed with. I believe it is referring to the actual final voting phase (as opposed to the nomination phase), so we should probably just leave it alone.

Thanks for catching the typos!

Kilo

#433 ::: Joshua Kronengold ::: (view all by) ::: June 02, 2015, 09:49 PM:

JJ@402: You seem to have missed all my points, so I'll try again:

Rather than use a list of Hugo winners making up a Rocket (which seems clever, but actually doesn't have anything to do with the meaning of E Pluribus Hugo in context, and anyway, we don't have permission to use those book titles making it a bit shady IMO), we should use common first names of people (drawing from lists of common first names).

Not specific people associated with Worldcon or this proposal. Just first names, chosen with an eye for them being names that people are likely to have (so common names, maybe with some curation to remove names too similar to one another).

The meaning of E Pluribus Hugo is not "out of all these books, there is one Hugo" -- much less "out of all these Hugo winners, there is one Hugo". Instead, it's "out of all these -people-, there is one Hugo" -- so ideally the 'pluibus' should be people.

#434 ::: David Dyer-Bennet ::: (view all by) ::: June 02, 2015, 10:50 PM:

Really hate the "random first names" variant of the t-shirt design, sorry.

No issue using the titles that I can see. I'm sure there are even multiple books under some of those titles -- because titles can't even be protected by copyright. And titles of Hugo winners have quite a lot to do with the nominating / voting process, really.

#435 ::: Cally Soukup ::: (view all by) ::: June 02, 2015, 11:06 PM:

[shirt]

I like the titles, myself. These are the works that have arisen out of the "pluribus" in years past. These, collectively, are the Hugo. A bunch of people's first names have to be explained.

One could make exactly the same shirt with the registered Hugo logo, but that runs into legal (not to mention ethical) issues. This design does the same thing without implying Hugo Mark Committee approval.

#436 ::: felice ::: (view all by) ::: June 02, 2015, 11:16 PM:

For 3.11.4, I'd suggest something like this:

The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. Additionally, the point totals for each round of the selection process shall also be published, in each category starting from the first round in which any of the previously listed works are eliminated.

#437 ::: Lydy Nickerson ::: (view all by) ::: June 02, 2015, 11:22 PM:

Keith, Sasquan knows me as Lydia Nickerson. Generally, I do go by Lydy, but I suspect that it would be better if it matched the legal name I registered under.

I like the rewording of Q.22. It doesn't elide all the detail, but it's cleaner and less defensive. Thank you.

#438 ::: Lydy Nickerson ::: (view all by) ::: June 02, 2015, 11:24 PM:

In re: t-shirts, I vastly prefer the titles to the names. Because what the Hugos is about is, you know, the work, not the people. (Ok, and you can quibble about the fan-writer Hugo being about people, but moving right along...) I like the design whole bunches, and I do want one, even though I can't be at the convention.

#439 ::: JJ ::: (view all by) ::: June 03, 2015, 12:59 AM:

Joshua Kronengold, #433:

You seem to have missed all my points, so I'll try again

The meaning of E Pluribus Hugo is not "out of all these books, there is one Hugo" -- much less "out of all these Hugo winners, there is one Hugo". Instead, it's "out of all these -people-, there is one Hugo" -- so ideally the 'pluibus' should be people.

Actually, no, the E Pluribus Hugo stands for "out of all these stories, we narrow it down to one for a Hugo".

This is not about the people (with the exception of the people categories).

I would like to have a shirt with a meaningful graphic that will still be meaningful several years from now.

#440 ::: JJ ::: (view all by) ::: June 03, 2015, 01:24 AM:

Joshua Kronengold, #433:

You seem to have missed all my points, so I'll try again

The meaning of E Pluribus Hugo is not "out of all these books, there is one Hugo" -- much less "out of all these Hugo winners, there is one Hugo". Instead, it's "out of all these -people-, there is one Hugo" -- so ideally the 'pluibus' should be people.

Actually, no, the E Pluribus Hugo stands for "out of all these possibilities, we narrow it down to one for a Hugo".

I would like to have shirt with a meaningful design which will still be meaningful several years from now.

#441 ::: JJ ::: (view all by) ::: June 03, 2015, 01:25 AM:

<dammit zapped by a server error and an entry I will swear was not there when I rewrote and reposted.>

#442 ::: Soon Lee ::: (view all by) ::: June 03, 2015, 06:10 AM:

[PROPOSAL]

How about this for second-last paragraph?

In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which nomineesfinalists should be voted on to be the final Hugo winner.

I had a look at the Sasquan member list which doesn't yet list Jameson Quinn & Jason Skiles. All the other names match except for the following that aren't exact matches but are easy enough to work out who they're meant to be, so no issues there:
Lydy Nickerson is listed as Lydia Nickerson.
Duncan J. McDonald. There are two Duncan MacDonalds listed by Sasquan so I assume the J. is to disambiguate.
P.J.Evans is listed as P J Evans.


#443 ::: Soon Lee ::: (view all by) ::: June 03, 2015, 06:30 AM:

[FAQ]

Q2: Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in SF. -> 'SFF' or 'SF & Fantasy'.

Q17:Either way works, but since our goal is to not prevent any section of fandom from being represented on the final ballot, the method of handling declined nominations outlined the proposal -> outlined in the

Q18: ...and decrease the number of slots a voter can vote for to a smaller number -> the number of slots a member can nominate to a smaller number

Q18: It doesn't solve the problem of forcing nominees off the ballot -> doesn't solve the problem of slates forcing nominees off the ballot

Q22: Those of us who worked on the system were very clear that our goal was not to keep anyone off of the Hugo ballot -> keep anyone off the Hugo ballot

#444 ::: Annie Y ::: (view all by) ::: June 03, 2015, 06:33 AM:

Soon Lee @443

Q2 - SF as in "Speculative Fiction" here not Science Fiction?

#445 ::: Soon Lee ::: (view all by) ::: June 03, 2015, 06:52 AM:

Annie Y #444:

"Speculative Fiction" would work for me too. I just want to be unambiguous.

But maybe stick to the way the WSFS Constitution phrases it, "Hugo Awards are given for work in the field of science fiction or fantasy"?

#446 ::: junego ::: (view all by) ::: June 03, 2015, 07:39 AM:

[NEW PROPOSAL MODS]

My suggestions will be in [bold brackets] or or bold strikethru

3.A.1: The finalist selection process shall be conducted in rounds consisting of three phases:

(1) Calculation Phase: For each category [First], the total number of nominations [from all eligible ballots] (the number of ballots on which each nominee appears)shall be tallied for each surviving nominee [in each category.] Next, a single “point” shall be assigned to each category on each nomination ballot[.] and [T]hat point shall be divided equally among all of the nominations [nominees] on that ballot in that category that were not eliminated previously [and shall be redivided as needed after Elimination Phase below]. Finally, all points from all nomination ballots shall be totaled for each nominee in that category [and shall be retotaled as needed after Elimination Phase below] These two numbers, [total nominations total and total points total, shall be used in the Selection and Elimination Phases.

(2) Selection Phase: [Each round shall start by selecting [T]he two nominees with the lowest point[s] totals shall be selected for the Elimination Phase. (See 3.A.3 for ties.)

(3) Elimination Phase: Nominees chosen in the Selection Phase shall be compared, and the nominee with the fewest total [lowest nominations total] shall be eliminated and removed from all ballots for the Calculation Phase of all subsequent rounds. [(See 3.A.3 for ties.)

3.A.2: The phases described in 3.A.1 are repeated in order until the number of finalists specified in 3.8.1 remain. If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then instead no nominees will be eliminated this round, and all remaining nominees shall appear on the final ballot, extending it if necessary.

#447 ::: junego ::: (view all by) ::: June 03, 2015, 08:20 AM:

[FAQ]

After all the final language is nailed down, I'd suggest that the FAQs be reordered so that common topics are clustered together. Examples: all the "how does it work?.(Qs 1, 4 & 17 -BTW did the idea to include an answer to why TIES are defined shot down?); followed by the "what happens if? "Why don't we just do x?", etc.

I'm inclined to putting the "isn't this one group keeping out another group?" first to declare our principles up front and/or have a question of "why is this change needed?"

#448 ::: Duncan J Macdonald ::: (view all by) ::: June 03, 2015, 08:47 AM:

Soon Lee @442:

There are two Duncan MacDonalds listed by Sasquan so I assume the J. is to disambiguate.

Yes. I didn't see any location asking for a middle name / initial when I registered, so I didn't indicate one in the First Name field. I use the "J." to separate me from other, more plebian Duncan Macdonalds.

#449 ::: Cassy B. ::: (view all by) ::: June 03, 2015, 08:47 AM:

junego @446, [FAQ} I'd prefer to keep in at least one explicit mention that this process takes place independently in each category. Otherwise it might be misunderstood to mean that every person has one point split up amongst everything they've nominated, and the short story nominations dilute the novella nominations.

We don't want people to make bad assumptions, when a simple "For each category" or "in each category" will clarify it.

#450 ::: junego ::: (view all by) ::: June 03, 2015, 09:33 AM:

[NEW FAQ MOD]

I've tried to sharpen the points people endorsed upthread and tune-up the language a bit. Also added my idea about the "why do this?" question.

*************************************************
22. Why is this change needed? Isn't this just an attempt by one group to keep out another?

There is a weakness that can be exploited in the current Hugo nominating process that was highlighted by the Sad and Rabid Puppy slates of 2015. This flaw allowed a small coordinated bloc of nominators to dominate the final ballot, shutting out all or most other nominations in entire categories. This is a major defect that must be fixed if the integrity of the award is to be maintained. E Pluribus Hugo accomplishes that goal by diluting the effect of slate or bloc nominating and greatly reducing the ability of any group to control the nominations.

Contrary to the idea that anyone is being excluded, if some hypothetical secret cabal were trying to maintain its dominance, they would not propose this system because E Pluribus Hugo works against both overt and covert attempts to control the nomination process. Those of us who worked on the system were very clear that our goals were to create a system that resisted bloc or slate nominating, treated all individual nominators equally and allowed a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.

Although it's true that much of the discussion for this system occurred on the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, they had no appreciable input into the goals, ideas or decisions about this proposal.
***************************************************

[This is a question/suggestion? Would the blog admins/owners be willing to create a link to these discussion threads on the front page that we could direct people to, or would that be too fraught with potential problems?]

#451 ::: junego ::: (view all by) ::: June 03, 2015, 09:43 AM:

Cassy B. @449
junego @446, [FAQ} I'd prefer to keep in at least one explicit mention that this process takes place independently in each category.

I didn't remove any references to each "category." I moved the one phrase from the beginning of the sentence to the end for consistency, i.e. "First...," "Next,..." "Finally,..." as the beginnings of the sentences that describe the steps in the Calculations section.

Do you think it should be referenced in other places?

#452 ::: David Harmon ::: (view all by) ::: June 03, 2015, 10:19 AM:

junego #450: There is already a link to the current thread at the top of the front page. What the mods could still do, would be put an index to the previous threads above the cut.

#453 ::: Keith "Kilo" Watt ::: (view all by) ::: June 03, 2015, 10:48 AM:

junego@446:

A couple of comments:

In the Calculation Phase description, I think we need the definition of total number of nominations in parenthesis, since this is the first time that term appears. Also, I think we do want "surviving", as not all works will still be there when we repeat this phase in the next round (I'm okay with another word that means the same thing). I think "redivided as needed" is quite confusing, especially since you removed the reference to "surviving" -- why wouldn't I need to redistribute the points? (Yes, I know the answer is because some ballots haven't changed, but that's not obvious -- and there's no reason not to recalculate -everyone's- ballots, since that's what the code actually does anyway.) Finally, I think "nominations total" is less clear than "total number of nominations". This may be in response to the fewer/lesser debate, but really, that one was unambiguous, so I think should be left alone.

Overall, I think there is too much uneeded foreshadowing in the way this phase is written here. It just comes out being much more complex than it needs to be. Once people have read all three phases, it's pretty clear what's going on and what the process is.

Kilo

#454 ::: Joshua Kronengold ::: (view all by) ::: June 03, 2015, 12:00 PM:

Ok, suggestion withdrawn re shirts. I'd happily get a shirt with the title design in any case.

I also don't much like including Retro Hugos (I don't like Retros or consider them a serious award--as the field is far -smaller- by the time the retros are considered; the electroate far less well informed about what the good works were in that year--and the award thus far less meaningful). But if people want to include them, I won't object that much.

#455 ::: David Dyer-Bennet ::: (view all by) ::: June 03, 2015, 12:52 PM:

For some time I've been unhappy about backing away quite so enthusiastically from PNH and TNH in FAQ 22. Making the point that this isn't their personal proposal is fine, since it's true, and it may also be useful, but emphasizing their low level of input bothers me.

In practical terms, it's open to challenge on the argument that being the moderator, and just having formed much of the local culture, is influence.

How about something making the first point, and then saying that the full discussion is publicly visible, and while very long, it lets people see exactly why decisions were made and whose ideas were influential? Emphasize the openness. (I can try to craft some such thing--if there's any interest.)

#456 ::: Keith "Kilo" Watt ::: (view all by) ::: June 03, 2015, 01:32 PM:

David@455:

I understand what you're saying. I've been reading many of the other blogs, though, and the fact remains that there are many who will discount out-of-hand anything that they think is being pushed by Teresa or Patrick. I think that's completely unreasonable, but it's true nonetheless.

I'm not opposed to adding a link to the discussions here if people think that will help, though I'm not sure how much good trying to read 3500 messages would do.

Kilo

#457 ::: P J Evans ::: (view all by) ::: June 03, 2015, 01:49 PM:

456
More than 5MB, in the four threads starting with the first post by Schneier, and running to here. (I've been saving them local.)

#458 ::: David Dyer-Bennet ::: (view all by) ::: June 03, 2015, 02:06 PM:

If nothing else, letting people see the scope of the discussion can give them some appreciation of the amount of work involved.

I don't think many people would follow such a link and read all the messages. However, I think pointing out that the discussion was all public and is all archived might still be valuable.

ML is probably so high-profile on this issue that such a link wouldn't bring in notably more problems -- but that's also a consideration, and I'd certainly defer to any policy decision on that part from the owners here.

#459 ::: David Dyer-Bennet ::: (view all by) ::: June 03, 2015, 02:07 PM:

I'm skeptical that anybody who will dismiss out of hand anything with Nielsen Hayden cooties would be disarmed by a claim in the FAQ that they'd had almost no input, though. I kinda suspect those people are lost beyond recall, at least beyond recall by a sentence or two of text no matter how well-crafted.

#460 ::: David Dyer-Bennet ::: (view all by) ::: June 03, 2015, 02:17 PM:

junego@446: You've improved the structure a lot here I think. Having 3.A.1 announce that the selection process is in three phases, iterated, with the three phases as the three sub-sections there, and putting the termination condition in 3.A.2 works well I think.

In (1) you refer to doing the redivision "after the elimination phase below". I think you just mean that you loop back to here -- but you already said that. The way that bit is written now it sounds like you're doing the redivision twice, once there in (1) and once after (3). While, as a programmer, I do see that doing it twice would get the same results both times, and hence wouldn't really matter (in the results), I think it's confusing to people.

Something about the malleability of the ballots needs to be mentioned somewhere around here I think; otherwise it's seriously unclear.

#461 ::: Soon Lee ::: (view all by) ::: June 03, 2015, 02:52 PM:

[FAQ]

@junego, Keith & DD-B,

I too think it's a great idea if these discussions can be linked in a top-post at Making Light so we can point people to it easily. Maybe once we're all done these discussion threads can be locked & then linked?

[FAQ22]

I tweaked junego's Q22 (below) in response to D-DB@455; folding the salient points of the last paragraph into the second & I think "hosted" is a better descriptor. Thoughts?


Contrary to the idea that anyone is being excluded, if some hypothetical secret cabal were trying to maintain its dominance, they would not propose this system because E Pluribus Hugo works against both overt and covert attempts to control the nomination process. Although it's true that much of the discussion for this system was hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resisted bloc or slate nominating, treated all individual nominators equally and allowed a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.

#462 ::: Keith "Kilo" Watt ::: (view all by) ::: June 03, 2015, 03:13 PM:

[3.A.1]

Here's the current version of 3.A.1, taking into account the changes junego suggested (but less the changes that I thought were making things unnecessarily complex). I'm still of a mind that "in each category" is clearer at the beginning of the sentences, but I can live with it either way.

3.A.1: The finalist selection process shall be conducted in rounds consisting of three phases:

(1) Calculation Phase: First, the total number of nominations (the number of ballots on which each nominee appears) from all eligible ballots shall be tallied for each surviving nominee in each category. Next, a single “point” shall be assigned to each category on each nomination ballot. That point shall be divided equally among all of surviving nominees on that ballot in that category. Finally, all points from all nomination ballots shall be totaled for each nominee in that category. These two numbers, total nominations and total points, shall be used in the Selection and Elimination Phases.

(2) Selection Phase: The two nominees with the lowest point totals shall be selected for comparison in the Elimination Phase. (See 3.A.3 for ties.)

(3) Elimination Phase: Nominees chosen in the Selection Phase shall be compared, and the nominee with the fewest total nominations shall be eliminated and removed from all ballots for the Calculation Phase of all subsequent rounds. (See 3.A.3 for ties.)

#463 ::: Joshua Kronengold ::: (view all by) ::: June 03, 2015, 03:27 PM:

4.6.2: Looks solid from here. By defining terms in 3.1.1(1), the rest can be described very simply.

#464 ::: Keith "Kilo" Watt ::: (view all by) ::: June 03, 2015, 04:09 PM:

Here's a thought...

If we write 3.A.1 to say:

3.A.1: For each category, the finalist selection process shall be conducted in rounds consisting of three phases:

...could we then get away with not saying "in each category" in steps 1-3?

Kilo

#465 ::: Jameson Quinn ::: (view all by) ::: June 03, 2015, 04:12 PM:

What about moving the "for each category" mention up to 3.A.1, and removing it from (1)? Because if you're going to put it in (1), then technically speaking you should put it in (2) too.

#466 ::: Jameson Quinn ::: (view all by) ::: June 03, 2015, 04:12 PM:

Jinx!

#467 ::: Keith "Kilo" Watt ::: (view all by) ::: June 03, 2015, 04:13 PM:

Lol... I like that idea, Jameson. :)

Great minds?

Kilo

#468 ::: Cassy B. ::: (view all by) ::: June 03, 2015, 04:22 PM:

@464, 465; I like it. And it resolves that pesky "short story nominations dilute novel nominations" misunderstanding that I could see coming over the horizon.

#469 ::: junego ::: (view all by) ::: June 03, 2015, 04:26 PM:

Keith "Kilo" Watt @453
junego@446:

In the Calculation Phase description, I think we need the definition of total number of nominations in parenthesis, since this is the first time that term appears.

I disagree. The parenthetical phrase is redundent, it doesn't add any info that isn't in the sentence already. The term "nomination(s)" is defined by inference in almost every sub-section of 3.8 and other places in the constitution. We don't need to reinvent the wheel here.

Also, I think we do want "surviving", as not all works will still be there when we repeat this phase in the next round (I'm okay with another word that means the same thing).

I disagree.
"Surviving" a) is not defined anywhere; b) is confusing because you have to read the entire "rounds" section to figure out what is (maybe) meant (forshadowing without directional arrows to help); c) is confusing because (for those who don't speak computer program) when you do read the entire "rounds" section, the implication, as written, seems to be that the only number affected by a "survivor" is the number of nominations, which is the only number that's *not* affected; d) adding "surviving" in the "Next" and "Last" sentences to make the language consistent just makes more of the "foreshadowing" you object to and the confusion/inconsistency I dislike; e) We can, I think, just leave it out because the last sentence in the Elimination Phase will be adequate instruction - if our goal is to use the minimal verbiage as long as it's sufficient for clear instruction. (which matches the rest of the language in the constitution.)

I think "redivided as needed" is quite confusing, especially since you removed the reference to "surviving" -- why wouldn't I need to redistribute the points? (Yes, I know the answer is because some ballots haven't changed, but that's not obvious -- and there's no reason not to recalculate -everyone's- ballots, since that's what the code actually does anyway.)

I agree with dropping it.
I don't understand what 'forshadowing' is or why it's a problem but there's a good reason to not add my phrasing in order to be consistent with keeping the explanations and verbiage sufficient for instruction but to the minimum so as not to look too complex. That's the main motivation for not having the separate section to define terms and describe the process as in the previous version, aiui. I'm happy to support that goal.

Finally, I think "nominations total" is less clear than "total number of nominations". This may be in response to the fewer/lesser debate, but really, that one was unambiguous, so I think should be left alone.

Not sure what the problem is here. I was trying to standardize terminology and descriptions within the section. I thought the compromise about least/fewest was to use 'lowest' instead because that's grammatically correct no matter how you interpret the difference between adding points and adding nominations. I just used "lowest" plus 'total' as the noun with 'points' and 'nominations' as the two adjectives to differentiate; all to mean we're looking for the smallest number of points or the smallest number of nominations. Maybe you could just use "number of..." for both with "total" and "smallest"? As long as the terms are consistent throughout I'll be content.

Overall, I think there is too much uneeded foreshadowing in the way this phase is written here. It just comes out being much more complex than it needs to be. Once people have read all three phases, it's pretty clear what's going on and what the process is.

As I said, not sure what the problem with 'foreshadowing" is, but agree we should keep the verbal complexity to a minimum without sacrificing clarity.

BTW, wanted to thank you, if I haven't already, for your hard work and patience in herding all the cats, me included. 😋

#470 ::: Jameson Quinn ::: (view all by) ::: June 03, 2015, 04:30 PM:

@469: Instead of discussing the philosophy behind the wording, I think it works better if you just suggest wording. I think Keith's latest @462 is pretty good, and has benefitted a lot from your suggestions. But if you think that it still needs work, make suggestions.

#471 ::: junego ::: (view all by) ::: June 03, 2015, 04:42 PM:

David Dyer-Bennet @460

junego@446: You've improved the structure a lot here I think. Having 3.A.1 announce that the selection process is in three phases, iterated, with the three phases as the three sub-sections there, and putting the termination condition in 3.A.2 works well I think.

Kilo did the latest rewrite, my 446 is just tweaking. He should get the kudos. Retyping/formatting the proposal and FAQ to post here is *realllly* tedious and time consuming. Plus he's keeping track of all the suggestions and then compiling/editing each new iteration before the type/format stuff. Big round of applause.

#472 ::: David Dyer-Bennet ::: (view all by) ::: June 04, 2015, 06:52 AM:

Keith@462:

Still bothers me that the Calc phase requires tallying the total number of nominations each time, when it cannot change. People will go looking for ways it can change, and I'm afraid may find misinterpretations that let that number change, and justify it because mandating re-tallying each time clearly implies the possibility of change.

Saying in 3.A.1 that the rounds apply to each category separately is probably a good idea (and then, as has been said, we don't need to say it elsewhere).

If "surviving" confuses people in 3.A.1(1) (and it's true it's not formally defined; but it's not a term we're giving a special meaning to), how about changing to the even more generic "remaining"? To be absolutely pedantic I suppose we could say "not yet eliminated".

#473 ::: junego ::: (view all by) ::: June 04, 2015, 09:05 AM:

David Dyer-Bennet @472
Keith@462:

Still bothers me that the Calc phase requires tallying the total number of nominations each time, when it cannot change. People will go looking for ways it can change, and I'm afraid may find misinterpretations that let that number change, and justify it because mandating re-tallying each time clearly implies the possibility of change.
.............
If "surviving" confuses people in 3.A.1(1) (and it's true it's not formally defined; but it's not a term we're giving a special meaning to), how about changing to the even more generic "remaining"? To be absolutely pedantic I suppose we could say "not yet eliminated".

It's not a deal breaker for me, it's just not needed. You're the one who pushed for writing this more similar to a code. I think that's a valid and acceptable format. Using that formatting idea, we should only have the extraneous verbiage necessary to translate computer-speak to human understanding.

As you point out above, leaving the instructions to include re-adding the nominations each round is nonsensical to a human, but explaining it would make the whole thing look more complex. So we will probably leave it, because it is possible to understand as written without too much effort. For the same reasoning, I think we should drop "surviving" or any other descriptive of what happens/changes after the first round if it's understandable after reading the whole process.

#474 ::: Jameson Quinn ::: (view all by) ::: June 04, 2015, 09:44 AM:

"uneliminated (as explained below)"?

#475 ::: junego ::: (view all by) ::: June 04, 2015, 10:13 AM:

Jameson Quinn @474
"uneliminated (as explained below)"?

a) Any explanatory wording here is unnecessary.
b) Kilo objects to "foreshadowing" which I interpretted as meaning - explaining early on, what becomes apparent later.

#476 ::: Jameson Quinn ::: (view all by) ::: June 04, 2015, 11:23 AM:

I don't really care what words we put, but I really do think that we should have at least some word(s) meaning "uneliminated", even if it becomes apparent later.

#477 ::: junego ::: (view all by) ::: June 04, 2015, 11:54 AM:

[PROPOSAL]
Jameson Quinn @474
"uneliminated (as explained below)"?
***
a) Any explanatory wording here is unnecessary.
b) Kilo objects to "foreshadowing" which I interpretted as meaning - explaining early on, what becomes apparent later.

If everyone else thinks there needs to be some explanation/descriptor at this point and I'm the only objector, how about:

"remaining nominees (see Elimination Phase)" Except just leave this "remaining" and the parenthetical off of the "First..." sentence, which would respond to David Dyer-Bennet's point @472 about the potential confusion of implying that the total nominations number needs to be re-added each round.

#478 ::: junego ::: (view all by) ::: June 04, 2015, 12:05 PM:

Jameson Quinn @476
I don't really care what words we put, but I really do think that we should have at least some word(s) meaning "uneliminated", even if it becomes apparent later.

Then how about "still eligible nominees (see Elimination Phase)" or even "eligible nominees" because that would encompass all nominees at all points in the process and not carry the potential confusion of whether or not total number of nominations changes/needs recalculation?

#479 ::: junego ::: (view all by) ::: June 04, 2015, 12:14 PM:

junego @478

"...and not carry the potential confusion of whether or not total number of nominations changes/needs recalculation?"

Oops, nope, this doesn't resolve that issue.

#480 ::: Keith "Kilo" Watt ::: (view all by) ::: June 04, 2015, 12:28 PM:

David@473:

I think either "remaining" or "not yet eliminated" could work. How about this:

3.A.1: For each category, the finalist selection process shall be conducted as elimination rounds consisting of three phases:

Then using "remaining" should be unambiguous, since they know it's an elimination-type process.

Kilo

#481 ::: Jameson Quinn ::: (view all by) ::: June 04, 2015, 12:33 PM:

I think @480 is a good solution.

#482 ::: David Dyer-Bennet ::: (view all by) ::: June 04, 2015, 02:08 PM:

Keith's @480 works for me.

#483 ::: junego ::: (view all by) ::: June 04, 2015, 04:51 PM:

Yeah, Kilo, @480 is ok with me, too.

#484 ::: Soon Lee ::: (view all by) ::: June 04, 2015, 05:18 PM:

+1 for @480

#485 ::: Annie Y ::: (view all by) ::: June 04, 2015, 05:48 PM:

I like 480 as well.

#486 ::: David Harmon ::: (view all by) ::: June 04, 2015, 06:26 PM:

480 is fine by me, but I mostly wanted to say that this whole thing is really shaping up nicely!

#487 ::: Keith "Kilo" Watt ::: (view all by) ::: June 05, 2015, 02:46 PM:

All:

Okay, seeing as how comments and suggestions seem to have died down, I'm going to post a candidate final draft of the proposal and FAQs. I think in the main, we all pretty much agree with what's here, so we're really looking for show stoppers at this point. Let us know your thoughts!

Thanks,
Kilo

#488 ::: Keith "Kilo" Watt ::: (view all by) ::: June 05, 2015, 02:47 PM:

Short Title: E Pluribus Hugo (Out of the Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations), section 3.9 (Notification and Acceptance), and section 3.11 (Tallying of Votes) as follows:

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the process described in section 3.A.

Insert new section 3.A after Section 3.8 as follows:
Section 3.A Finalist Selection Process

3.A.1: For each category, the finalist selection process shall be conducted as elimination rounds consisting of three phases:

(1) Calculation Phase: First, the total number of nominations (the number of ballots on which each nominee appears) from all eligible ballots shall be tallied for each remaining nominee. Next, a single “point” shall be assigned to each nomination ballot. That point shall be divided equally among all remaining nominees on that ballot. Finally, all points from all nomination ballots shall be totaled for each nominee in that category. These two numbers, total nominations and total points, shall be used in the Selection and Elimination Phases.

(2) Selection Phase: The two nominees with the lowest point totals shall be selected for comparison in the Elimination Phase. (See 3.A.3 for ties.)

(3) Elimination Phase: Nominees chosen in the Selection Phase shall be compared, and the nominee with the fewest total nominations shall be eliminated and removed from all ballots for the Calculation Phase of all subsequent rounds. (See 3.A.3 for ties.)

3.A.2: The phases described in 3.A.1 are repeated in order for each category until the number of finalists specified in 3.8.1 remain. If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then instead no nominees will be eliminated this round, and all remaining nominees shall appear on the final ballot, extending it if necessary.

3.A.3: Ties shall be handled as described below:
(1) During the Selection Phase, if two or more nominations are tied for the lowest point total, all such nominees shall be selected for the Elimination Phase.
(2) During the Selection Phase, if one nominee has the lowest point total and two or more nominees are tied for the second-lowest point total, then all such nominees shall be selected for the Elimination Phase.
(3) During the Elimination Phase, if two or more nominees are tied for the fewest number of nominations, the nominee with the lowest point total at that round shall be eliminated.
(4) During the Elimination Phase, if two or more nominees are tied for both fewest number of nominations and lowest point total, then all such nominees tied at that round shall be eliminated.

3.A.4: After the initial Award ballot is generated, if any finalist(s) are removed for any reason, the nomination elimination system shall be rerun as though the removed finalist(s) had never been nominees. None of the remaining original finalists who have been notified shall be removed as a result of this rerun. The new finalist(s) shall be merged with the original finalists, extending the final ballot if necessary.

Section 3.9: Notification and Acceptance.
3.9.1 Worldcon Committees shall use reasonable efforts to notify the nominees finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each nominee finalist shall be asked at that time to either accept or decline the nomination. If the nominee a finalist(s) declines the nomination, that nominee finalist(s) shall not appear on the final ballot. The procedure for replacement of such finalist(s) is defined in subsection 3.A.4.

Section 3.11: Tallying of Votes.
3.11.4: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. During the same period, the results of the last ten rounds of the finalist selection process for each category (or all the rounds if there are fewer than ten) shall also be published.

Submitted by: Keith “Kilo” Watt, Jameson Quinn, Tammy Coxen, Yoana Yotova, Joshua Kronengold, Christopher Battey, David Gallaher, Adam Tilghman, Sara Watt, CJ Cabourne, Steven Halter, P J Evans, David Goldfarb, Seth Gordon, Ginger Tansey, Steve Wright, Catherine Faber, Andy Holloway, Duncan J. Macdonald, Claudia Beach, Derry Earnshaw, Jason Skiles, Soon Lee, David Harmon, Lydia Nickerson, Abigail Sutherland, Lee Billings, Oskari Rantala, Seth Breidbart, Chris Suslowicz, Fragano Ledgister, Lori Coulson, Jeffry Herman, Mark Shier, Buddha Buck, Lenore Jean Jones, David Langford, Christopher Hatton

Commentary: “E Pluribus Hugo” – the name of this proposal is also its goal: recognizing the many opinions within fandom as to what nominees might be worthy of the Hugo. It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members: They still list the nominees, unranked, they feel are Hugo-worthy (up to the maximum permitted) in any categories they choose, just as they have in the past. In the past, we have counted the number of nominations each nominee received, and the top five nominees were put on the final ballot. However, because SF fandom typically nominates a variety of different nominees, it was easy for an organized slate to make it so that no other nominees made the final ballot.

Using this system, fandom isn't penalized for nominating a wide variety of nominees. If you nominate something that ends up not having a chance to make the final ballot, then your remaining choices automatically get more of your support instead of just being wasted. In other words, you can safely nominate anything you feel is Hugo-worthy. If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot. In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which finalists should be voted on to be the Hugo winner.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no group of members -- of any sort, minority or majority -- can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Award.

#489 ::: Keith "Kilo" Watt ::: (view all by) ::: June 05, 2015, 02:48 PM:

FAQ’s:

1. Can you explain the system in plain language?
First, and most importantly, E Pluribus Hugo makes absolutely no changes in how members nominate (no ranking or complex strategies are required). The new system changes only the way nominations are tallied in order to create a more robust measure of which nominees have the broadest and deepest support among WSFS members. The final Hugo voting system, which actually chooses the winner, is completely unchanged.

The new system narrows down Worldcon members' nominations by knocking out the least supported works in repeated elimination rounds until only five (under current rules) finalists remain. Here are the basic steps, as applied to one category:

a. All the nominations for all nominees (that is, the number of ballots on which each nominee apears) are added up, just like in the current system. This number, called "total nominations" in the proposal, is one of the criteria used to determine the final nominees with the broadest support.

b. Next, one point is divided equally among all the nominees in a category on each member's ballot. If there are two nominees on a ballot in a category, each gets 1/2 point; if there are three nominees, each gets 1/3 point, and so on.

c. All the points for each nominee from all nomination ballots are then added together. This number, called "point total" in the proposal, is the second of the two criteria used to determine the final nominees with the broadest support.

d. The nominees with the lowest point totals are compared to each other. The nominee with the fewest number of nominations is eliminated, because it is the nominee with the least support.

e. After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed among their remaining nominees. So, if there were four nominees getting 1/4 point each, there are now three nominees getting 1/3 point each, and so on. All points are re-added for all nominees. As less-supported nominees are eliminated, the total points for remaining nominees increase.

This process is repeated until the designated number of finalists (five in the curent version of the constitution) remain.


2. E Pluribus Hugo?
The designers felt this name accurately reflects what the Hugo nomination system should be trying to achieve. Fandom is the Many. Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in science fictino and fantasy. This is a defining characteristic of fandom, and this system embraces it. Out of all of these many tastes in nominees, this system allows a single set of nominations to slowly emerge as the most popular candidates for the Hugo Award: “Out of the Many, a Hugo.”


3. But surely that’s not what the system is formally called?
Okay, sure. Formally, in the context of election theory, this system would be called “single divisible vote with least popular elimination” or SDV-LPE for short. E Pluribus Hugo is both cooler and geekier, as everything is better with a little Latin.


4. Why are there Selection and Elimination Phases? Couldn’t you simplify the system by doing it all in one step per elimination?
The Selection Phase looks for nominees with the lowest point total, and the Elimination Phase eliminates the one with the fewest nominations. Looking at the two different measures of popularity adds robustness to the system. With non-slate nominations, as occurred with most of the Hugo Awards to date, the nominees with the fewest points will tend to be the same ones as those with the fewest number of nominations. But while it might be possible for strategic blocs to manipulate one of these two measurements (as occurred in 2015), it would be very difficult to manipulate both in the same direction. A strategy that focuses on nominations for the Elimination Phase (such as bloc voting) will be stymied by the Selection Phase; a strategy that focuses on points for the Selection Phase (such as nominating only one nominee) will turn out to be counterproductive in the Elimination Phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.


5. Why re-run the system if someone declines a nomination? Why not just take the nominee that was eliminated in the next to last round?
We can do that, and in fact have an amendment ready if that’s what the people decide they want to do. Our system has only one major difference between it and taking the “6th place” nominee. Under our proposal, if a slate nominee makes the final ballot and that nomination is declined, that nomination will tend to be replaced by another nominee from the slate. In taking the “6th place” nominee, the declined nomination will tend to be replaced by a non-slate nominee. Either way works, but since our goal is to not prevent any section of fandom from being represented on the final ballot, the method of handling declined nominations outlined in the proposal seemed more in line with that philosophy.


6. Shouldn’t I just nominate one work if I want to give it the best chance to win?
The system was crafted specifically to make most strategies irrelevant. For example, suppose you wanted to give your full support behind one nominee that you really feel strongly about. You can do that, particularly if there isn’t anything else you feel is Hugo-worthy. On the other hand, if you nominate four other things that you think might be Hugo-worthy, you aren't hurting your favorite. This is because if your other nominees get eliminated, then your full support will go towards the remaining nominee, just as if you had never nominated the others. If enough people agree with your other four, some of them might make it, too. If enough people don't agree with your favorite, there's nothing you can do to get it on the final ballot, because you will never have more than one point. In general, the best strategy is simple: nominate as many nominees as you feel are worthy.


7. Surely strategy can make some difference, can't it?
It's impossible to create a voting system in which strategy never makes a difference. But in this system, there's no good way to know ahead of time that a strategy is safe and effective. Here's a specific example:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. Suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all could submit nomination ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its nominations compared in the Elimination Phase (you have no way to increase the number of nominations that Castle receives if it is selected for the Elimination Phase). If you push it up to 5th place, it can still be eliminated by Tentacles if Castle has fewer nominations. If you could somehow push it all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters so precisely before the election. Maybe Rocket isn’t so popular after all and needs your nominations to get over the top. Maybe lots of people will assume Rocket is safe and cause it to end up losing as a result of leaving it off their ballots. Maybe Tentacles has a broad enough base of support that your nominations for Castle aren’t enough to eliminate it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire, and only one, highly-specific way for it to work.

Nominating what you think is Hugo-worthy really is your best strategy.


8. Isn't it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws. The objective is to choose a system whose flaws are not in an area of concern to the electorate.


9. What are E Pluribus Hugo’s flaws?
In rare cases, it is possible that eliminating both members of a tie could change the final ballot slightly from what it would be if the tie were broken. The change is usually in the least popular of the finalists, and requires the two nominees in question to be very close in popularity. Hundreds of simulations were run using real and created data sets to help decide which tie-breaking methods would maintain our goal of supporting the wide range of opinions within fandom. It turned out that all the simulations showed that there was almost no difference in outcomes, no matter how we broke ties. We have chosen, therefore, to break ties in a manner consistent with section 6.4 of the Worldcon constitution. There are a number of additional tie breakers that could be used if it were deemed necessary in the future; however, simulations that have been run by the designers of this system show that it really shouldn’t be required.


10. What are E Pluribus Hugo’s benefits?
Simply put, it reduces the power of bloc voting without eliminating the chance that nominees that do appear on slates will make it to the final ballot. Conversely, it makes it very difficult for slates to prevent non-slate nominees from appearing on the ballot.


11. How does this system eliminate slate or bloc voting?
It doesn't, exactly, nor should a nominee be automatically eliminated just because it appears on a slate. On the other hand, any slate which nominates a full set of five nominees will find that each of its nominations only count 1/5 as much. With “non-slate” nominating, some of your nominees will be slowly eliminated, so your remaining nominees get more and more of your support. Since slate nominees tend to live or die together, they tend to eliminate each other until, in general, only one slate nominee remains. With a large enough support behind the slate (five times as much), the slate may still sweep a category; however, if that many voters support the slate, they arguably deserve to win, and no fair and unbiased system of nomination will prevent that. The answer in that case is, simply, to increase the general pool of voters. Regardless, with E Pluribus Hugo, slates will never receive a disproportionate share of the final ballot, as occurred in the 2015 Hugos.


12. Couldn't supporters of slates just recommend a single nominee for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate has enough supporters that is certainly a viable possibility – it’s also completely fair. It does not force all other nominees off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn't a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a nominee deserves a Hugo nomination, then it should, in fact, be represented on the final ballot.


13. What if there are multiple slates (slate wars, “parties”, etc.)?
As with a single slate, the nominees on each slate will tend to eliminate each other until only the most popular one remains. The end result is that even multiple slates are unable to sweep the nominations.


14. What happens if a broadly popular nominee is nominated by a group of unrelated people?
If it is broadly popular, the system will still select that nominee for the final ballot.


15. What happens if a broadly popular nominee also appears on a slate?
Even if it is on a slate, if the nominee garners support from individuals then the system will select that nominee for the final ballot. In general, slates neither help nor hurt any given nominee.


16. What happens when there are a lot of nominees with no obvious favorites and nomination slates are introduced?
Even in this case, we were unable to find a simulation in which no non-slate nominees appeared on the final ballot. Slate nominees did receive a larger proportion of nomination slots than they did otherwise, however, again, this could be considered a fair and valid result. If there was no general favorite, then members really had no collective preference.


17. How do the results of this system compare to the results under the current nomination system?
Statistical tests showed that this system and the current system overlap in about (conservatively) 4.5 out 5 candidates. In the absence of slates, it generally gives identical results to the current nomination system.


18. I think we should just increase the number of nomination slots on the final ballot to a larger number (for example, 6), and decrease the number of slots a member can nominate to to a smaller number (for example, 4). Wouldn't that be simpler and easier?
Unfortunately, this simply means that the largest slate will receive four of the nominations and the next largest will receive the remaining two. It doesn't solve the problem of slates forcing nominees off the ballot that had a chance to win the final election. In general, we want fandom to nominate as many nominees as they feel are Hugo-worthy, since under E Pluribus Hugo there is no strategic reason not to do so. For this reason, most of the designers of this system would prefer that members not be limited to nominating only four nominees. Keep in mind, however, that E Pluribus Hugo will work with this (or most any other) change as well, so one does not preclude the other.


19. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
This could work. The problem is that then you would have a small group of people serving as literal gatekeepers to the Hugo Awards. In spite of the word on the Internet, this has never been the case in the past. Establishing it would prove those groups who believed it existed in the past would then be correct. Ultimately, human judgement is fallible. The fairness of a committee’s decisions would forever be subject to opinion, and the prestige of the Hugo Awards would forever be tarnished.


20. I think we should use [insert other mathematical voting system].
We considered essentially every applicable type of voting system currently in the literature, guided by two experts in the field. It should be kept in mind, however, that the goals and requirements for choosing a set of representatives in a political situation are different from those for choosing a set of Hugo finalists. Some of these systems do, in fact, have positive properties that speak for them. None of them were as simple or as intuitive as E Pluribus Hugo, yet E Pluribus Hugo meets all of the stated goals for a Hugo nomination system.


21. Won’t E Pluribus Hugo be complicated to code and implement?
One of our non-experts coded a full simulator for the system in a matter of days. A full web-based app would not be much more difficult to handle.


22. Isn't this just an attempt by one group to keep out another?
No. E Pluribus Hugo is designed to fix the flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the award is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.

It is true that much of the discussion for this system occurred on the “Making Light” discussion board, owned by Teresa and Patrick Nielsen Hayden, however they had no real input in the discussions of the system. Those of us who worked on the system were very clear that our goal was not to keep anyone off the Hugo ballot, and that any system which specifically targets any type of nominee is inherently wrong and unfair. Our goal was to create a system resistant to slate or bloc nominating, a system that allows a consensus final ballot to emerge from all the nominations submitted.

#490 ::: David Langford ::: (view all by) ::: June 05, 2015, 03:23 PM:

#489 It is true that much of the discussion for this system occurred on the “Making Light” discussion board, owned by Teresa and Patrick Nielsen Hayden, however they had no real input in the discussions of the system.

Very minor point, but maybe a bit comma-splicey? I'd make it "... Hayden; however, they had no real input into ..."

#491 ::: Xopher Halftongue ::: (view all by) ::: June 05, 2015, 03:54 PM:

I agree with Dave. Semi and a comma is needed there.

#492 ::: Xopher Halftongue ::: (view all by) ::: June 05, 2015, 03:54 PM:

Internal server, you have erred.

#493 ::: Xopher Halftongue ::: (view all by) ::: June 05, 2015, 03:58 PM:

I meant to say "Beautiful work! And thanks to everyone who worked on it!"

Should we try to have E Pluribus Hugo parties at WorldCon, like bid parties? Should sympathetic parties close down early the night before the WSFS meeting so more of us will make it there? Buttons? Ribbons? Copies of the proposal?

(I think buttons or ribbons should say "Ask me about E Pluribus Hugo.")

#494 ::: Seth Gordon ::: (view all by) ::: June 05, 2015, 05:15 PM:

Typo in FAQ#2: “fictino”

#495 ::: Seth Gordon ::: (view all by) ::: June 05, 2015, 05:24 PM:

And FAQ#7 and FAQ#8 should be merged; they are basically the same question, with #7 providing a long answer and #8 providing a short answer.

Other than that, great work!

I have no particular commentary regarding the design of a T-shirt, but I want one.

#496 ::: Annie Y ::: (view all by) ::: June 05, 2015, 05:27 PM:

FAQ1: clause a: typo: "apears" should be "appears"

FAQ1 - the last sentence should end with "remains" and not remain (number of finalists remains). Same sentence - typo: "curent" should be "current"

#497 ::: Annie Y ::: (view all by) ::: June 05, 2015, 05:38 PM:

FAQ 16.

What exactly we are trying to say here?

If I understand this correctly, won't it be clearer to replace "no non-slate" with "only slate" in the first sentence? And if this is not what it means - it is unclear :)

#498 ::: Annie Y ::: (view all by) ::: June 05, 2015, 05:45 PM:

And today is the day of me posting too many messages in a row:)

FAQ 17:
"4.5 out 5 candidates" should be "4.5 out of 5 candidates"

FAQ 18: Extra "to" in the question text "can nominate to to a smaller number" should be "can nominate to a smaller number"

FAQ 18:
I really do not like the flow of this sentence:

It doesn't solve the problem of slates forcing nominees off the ballot that had a chance to win the final election

Maybe use commas somewhere there or change the word order (word order change proposal below):

It doesn't solve the problem of slates forcing nominees that had a chance to win the final election off the ballot

#499 ::: Soon Lee ::: (view all by) ::: June 05, 2015, 06:09 PM:

[FAQ22]

No takers for my suggested rewrite of junego's FAQ22 @461? Specifically, using "hosted" as a better descriptor of how important having a venue like Making Light has been, and also to emphasize the openness of these discussions.


"Contrary to the idea that anyone is being excluded, if some hypothetical secret cabal were trying to maintain its dominance, they would not propose this system because E Pluribus Hugo works against both overt and covert attempts to control the nomination process. Although it's true that much of the discussion for this system was hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resisted bloc or slate nominating, treated all individual nominators equally and allowed a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals."

#500 ::: P J Evans ::: (view all by) ::: June 05, 2015, 06:27 PM:

499
[FAQ22]
You're correct about that. Not that the juvenile canines won't whine and growl a lot anyway....

#501 ::: David Dyer-Bennet ::: (view all by) ::: June 05, 2015, 06:51 PM:

We can't make people committed to opposition on all fronts change their minds or behave better, no, not even by writing better.

It's still important to be as clear and accurate as we can. It makes it easier to influence people with some openness to their minds. It makes it more obvious that people are not engaging with what we've actually said, for example.

And I like Soon Lee's rewrite of FAQ 22 better than the current version. I'd like it even better if it included something like "The full discussion there is publicly visible; go see for yourself who thought of this, what the arguments were against that, and why people chose the other thing."

(We don't really know, no matter how confident we may feel in some cases, which people arguing with us are listening, and which are just arguing.)

#502 ::: David Dyer-Bennet ::: (view all by) ::: June 05, 2015, 06:56 PM:

Annie@497: [FAQ 16] is attempting to address the case where, setting aside slate nominators, there's a broad range of preferences in nominations, with no really strong leading contenders.

Your proposal does I think clarify it slightly. Both are a bit clumsy, in different ways, and I think yours is slightly better (which means there may be a third way that really fixes it; though I don't see it). But "no non-slate" is a classic confusion producer, yeah.

#503 ::: junego ::: (view all by) ::: June 05, 2015, 10:46 PM:

Soon Lee @499
[FAQ22]

No takers for my suggested rewrite of junego's FAQ22 @461? Specifically, using "hosted" as a better descriptor of how important having a venue like Making Light has been, and also to emphasize the openness of these discussions.
.....
"Contrary to the idea that anyone is being excluded, if some hypothetical secret cabal were trying to maintain its dominance, they would not propose this system because E Pluribus Hugo works against both overt and covert attempts to control the nomination process. Although it's true that much of the discussion for this system was hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resisted bloc or slate nominating, treated all individual nominators equally and allowed a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals."

Yes, I think this is an improvement on mine. But include some part of what's already there, specifically:

E Pluribus Hugo is designed to fix the flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the award is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations.

I also agree we should ask if we can link to the full discussion here.

My other suggestions are:

1) Move Q22 to Q1 and add "Why are you proposing this change? Isn't this..." It seems more logical to start with why we think this is necessary.

2) Move Q1 to Q2, then regroup remaining FAQs so that similar subjects are grouped together. All the "how does this work?" type together, all the "what happens if?" together, all the "why not do something else?" together, etc.


Hey, we're almost there, folks! GO US!! Especially all who've done the heavy lifting in this and earlier discussions.

#504 ::: junego ::: (view all by) ::: June 05, 2015, 10:54 PM:

junego @503

"regroup remaining FAQs so that similar subjects are grouped together. All the "how does this work?" type together, all the "what happens if?" together, all the "why not do something else?" together, etc."

MY BAD, Kilo has regrouped things, I just didn't read carefully before responding. Apologies.

#505 ::: Keith "Kilo" Watt ::: (view all by) ::: June 05, 2015, 11:01 PM:

junego@503:

I'd prefer to leave 22 for last, since we don't really want that to be the focus of the proposal. I think the "how does it work" type questions are far more important, since they are what will convince people to vote for the proposal.

I did re-arrange the FAQs into what seemed to me to be a logical order, but there are probably other ways to do that. What order would you prefer?

Soon Lee@499:
I'm okay with this as well. There have been some other additions to it proposed; is there a full version someone wants to propose?

Seth@495:
They aren't quite the same, and 8 is kind of a lead-in to 9. I'm not sure I'm in favor of combining 7 and 8, but it may be that we don't need 8 at all. What do people think?

Kilo

#506 ::: Keith "Kilo" Watt ::: (view all by) ::: June 05, 2015, 11:05 PM:

junego@504: No worries!


All: By the way, I've updated the draft with all of the typos indicated to date. I also changed the reference in 3.A.4 to "nomination elimination system" to be "finalist selection process" since that's the new title of section 3.A (the missing colon in the section title has been inserted as well).

Kilo

#507 ::: junego ::: (view all by) ::: June 05, 2015, 11:19 PM:

Annie Y @497

[FAQ/Q16]

How about this?

16. What happens when there are a lot of nominees with no obvious favorites and nomination slates are introduced?
In these cases slate nominees did receive a larger proportion of final ballot slots, but in multiple simulation runs with these conditions they never completely dominated the category. If there are no collective preferences who get a high percentage of nominations and a group of fans do have strong favorites, it's not unfair that some of their picks become finalists.

#508 ::: P J Evans ::: (view all by) ::: June 05, 2015, 11:30 PM:

507
NO. That's getting political. Any talk about 'fairness' in connection with minority/majority selections is getting too damned close to the juvenile canines and their desires.
Most popular, fine - that's what it is.

#509 ::: David Dyer-Bennet ::: (view all by) ::: June 06, 2015, 01:08 AM:

PJ@508: the question of what is fair is core to this whole endeavor. We can't not address it; we are proposing changing a first-past-the-post type nomination system to something else to produce a result that seems to us more fair under a very wide range of simulated conditions (which we think covers real-world possibilities reasonably well). That is the entire point of this exercise.

#510 ::: junego ::: (view all by) ::: June 06, 2015, 01:22 AM:

[PROPOSAL/COMMENTARY]

This is a proposed departure from the Commentary format we now have. I was thinking of editing this section a bit and Looked over the other proposals for the Business Meeting at Sasquan for reference. The Commentary for almost all of them are very concise and very spare. The only one that's more complex is the "Best Saga" proposal and I don't see a FAQ for them (or anyone else).

I ended up redoing our Commentary to try to make it more of a bare outline.The fuller explanations are in the FAQs which we're officially submitting as part of the proposal, right? We don't really gain anything by duplicating all those words here. If I'm incorrect about the FAQs, then ignore this. 😏


Commentary: “E Pluribus Hugo” – the name of this proposal is also its goal: recognizing the many opinions within fandom regarding what nominees are worthy of the Hugo.
......
The goal of this amendment is to promote more diversity in nominations and prevent a coordinated minority from dominating one or more categories.
.....
There will be no change in the nomination process from the perspective of WSFS members. The proposed change is in the way nomination ballots are tallied.
.....
Finalists will be determined not by "first past the post", as is done now, but by elimination of nominees with the least support, from the lowest to the highest using two criteria: 1) the current total nominations and 2) a number generated by the number of nominees on each ballot in a category. This system has some similarities to the current final voting system in that as the nominees on a ballot are eliminated, more weight is given to the remaining nominees. In this way the candidates with the broadest support of both criteria are selected as finalists.
.....
This system resists coordinated nominating because strategies that give an advantage with one criteria tend to create a disadvantage with the other criteria. (No system can completely resist the effects of organized, coordinated voting.)
....
It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no group of members -- of any sort, minority or majority -- can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Award.

#511 ::: Xopher Halftongue ::: (view all by) ::: June 06, 2015, 01:31 AM:

Seth 494: I thought a fictino was a very small ficton.

#512 ::: junego ::: (view all by) ::: June 06, 2015, 01:35 AM:

P J Evans 508
507
NO. That's getting political. Any talk about 'fairness' in connection with minority/majority selections is getting too damned close to the juvenile canines and their desires.
Most popular, fine - that's what it is.

Can you suggest another sentence that would work better to explain that if the majority doesn't really have a strong consensus about the all the nominees for the short list and a sub-group of members DO have a strong consensus that the system allows them to have a bit more influence and that outcome is not necessarily a bad thing?

#513 ::: Keith "Kilo" Watt ::: (view all by) ::: June 06, 2015, 02:08 AM:

Junego@510:

I'm having trouble finding the link in the previous thread, but someone pointed to some proposed amendments from the London Worldcon that were much more involved in commentary than ours. That was when Kevin encouraged us to have a detailed "legislative history" for the proposal. So I think we may be okay there.

Really, I see the commentary as our first chance to sell the proposal, so that was the mode I was writing it in. My thoughts were, "If they read nothing but this commentary, what would I say?" The details of the system are in the FAQ, of course. But while there is some duplication, I think that's more because my goals for the commentary and for the FAQ were different, and there needed to be some overlap as I was trying to achieve those separate goals.

Kilo

#514 ::: Soon Lee ::: (view all by) ::: June 06, 2015, 03:51 AM:

junego @510 and Keith @513:

I prefer the way we have it with both the proposal with commentary *and* the FAQs. The proposal uses formal language as it should be, while the FAQ section is more conversational (easier to understand?) and allows us to unpack our intent, with examples as needed. Some duplication is a good thing IMO so long as the meaning is consistent (and I think it is), as it can help reinforce the message.

#515 ::: Soon Lee ::: (view all by) ::: June 06, 2015, 04:33 AM:

[FAQ22]

22. Isn't this just an attempt by one group to keep out another?

No. E Pluribus Hugo is designed to fix the flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the award is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.

Although it is true that the discussions for this system was hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resisted bloc or slate nominating, treated all individual nominators equally and allowed a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.

If you wish to see how the sausage was made, we invite you to view the full & extensive discussions that led to E Pluribus Hugo at the following link: [INSERT LINK HERE]

Given that Teresa & Patrick have provided this space for us to conduct these discussions, I'm fairly certain they would be agreeable to having them linked to the proposal. Teresa or Patrick, may we link the discussions here to the E Pluribus Hugo proposal?

#516 ::: Soon Lee ::: (view all by) ::: June 06, 2015, 05:13 AM:

[PROPOSAL]

3.A.4:
After the initial Award ballot is generated, if any finalist(s) are removed for any reason, the nomination elimination system shall be rerun as though the removed finalist(s) had never been nominees. -> nominee(s)

[FAQ]

16. What happens when there are a lot of nominees with no obvious favorites and nomination slates are introduced?

[CHANGE TO] Simulations of this scenario showed that slate nominees did receive a larger proportion of nomination slots than they did otherwise. However, this could be considered a fair and valid result: If there was no general favorite, then members really had no collective preference. Even in this scenario, simulations showed that non-slate nominees were not completely shut out of the final ballot.


18. I think we should just increase the number of nomination slots on the final ballot to a larger number (for example, 6), and decrease the number of slots a member can nominate to to a smaller number (for example, 4). Wouldn't that be simpler and easier?

It doesn't solve the problem of slates forcing nominees off the ballot that had a chance to win the final election. -> It doesn't solve the problem of slates depriving non-slate nominees of the opportunity to compete in the final award vote.

#517 ::: JJ ::: (view all by) ::: June 06, 2015, 06:08 AM:
22. Isn't this just an attempt by one group to keep out another?

I am still strongly objecting to wording that seems to make a tacit admission that there are two groups involved. There aren't.

#518 ::: Soon Lee ::: (view all by) ::: June 06, 2015, 06:26 AM:

JJ #517:

I agree with you that there aren't two groups involved. For me, it's about addressing those who would ask *that* question directly, and I think the answer lays it out plainly that it's not about any "group".

Do you have a different wording of the question that is more satisfactory?

(I have & am still resisting the urge to snark about "...conducted 100% in the open, democratically, using a democratic process". But I digress...)

#519 ::: JJ ::: (view all by) ::: June 06, 2015, 08:45 AM:
21. Isn't this change just designed to keep certain people out of the Hugo nomination process?
#520 ::: Soon Lee ::: (view all by) ::: June 06, 2015, 09:15 AM:

JJ #519:

I like it.

#521 ::: Jameson Quinn ::: (view all by) ::: June 06, 2015, 09:46 AM:

I think that the question should be in the terms it would actually be phrased, even if that involves "groups". If we want to confront the idea that there are groups, we should do that in the answer.

To me, it seems clear that
(1) There was at least one coherent puppy faction (the rabids).
(2) There is not, in fact, a coherent anti-puppy faction. Yes, there is plenty of anti-puppy sentiment, and there is a developing consensus about what to do about it. But that's just a bunch of people exercising their independent judgement; it's not a faction; and there are various objective criteria on which to make that distinction.
(3) There is no point in insisting on (2) in the FAQ. People who don't believe (2) are wrong; but that should be irrelevant to whether they support this proposal, because even if (2) were false, this proposal would fix the problem, not make it worse.
(4) Therefore, while I understand JJ's objection to an FAQ that seems to cede (2) as false, and I'd be open to wording that was clearer about any anti-puppy faction being a hypothetical construct, I think that any good version of FAQ#22 is going to "seem to make a tacit admission that there are two groups involved" at some level.

#522 ::: Jameson Quinn ::: (view all by) ::: June 06, 2015, 09:49 AM:

In other words, if we're talking about the wording of the question itself, I think the wording in @517 is actually better than @519. I am open to proposals to fix the response, adding words like "hypothetical", as long as we don't devote too much space to this issue, because it's really beside the point of the answer.

#523 ::: junego ::: (view all by) ::: June 06, 2015, 11:03 AM:

Keith "Kilo" Watt @513
Junego@510:

I'm having trouble finding the link in the previous thread, but someone pointed to some proposed amendments from the London Worldcon that were much more involved in commentary than ours. That was when Kevin encouraged us to have a detailed "legislative history" for the proposal. So I think we may be okay there.
...
Really, I see the commentary as our first chance to sell the proposal, so that was the mode I was writing it in. My thoughts were, "If they read nothing but this commentary, what would I say?" The details of the system are in the FAQ, of course. But while there is some duplication, I think that's more because my goals for the commentary and for the FAQ were different, and there needed to be some overlap as I was trying to achieve those separate goals.

Ok, we're good, just do that 'ignore' thing I mentioned. ;^]

I was going to ask first if this was the correct format, but I'd already been messing around with how to reword it. Just decided to include my brain farts with the question.

What still isn't clear is where are the FAQs? Are they part of the permanent record or not? I couldn't find any links or mention.

#524 ::: Keith "Kilo" Watt ::: (view all by) ::: June 06, 2015, 11:15 AM:

junego@523:

We'll submit the proposal and FAQs as one document. I broke them apart here just to make it easier to discuss.

Kilo

#525 ::: junego ::: (view all by) ::: June 06, 2015, 11:26 AM:

#515 ::: Soon Lee ::: (view all by) ::: June 06, 2015, 04:33 AM:
[FAQ22]

With the change of question wording by JJ @519 added and a few slight edits (I like using synonyms 😋). I think this is a winning Q22.
***************************************
22. Isnt't this change just designed to keep certain people out of the Hugo nomination process?
No. E Pluribus Hugo is designed to fix the flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw [limitation] allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw [deficiency] that must be fixed if the integrity of the award is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.
...
Although it is true that the discussions for this system was were hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resisted bloc or slate nominating, treated all individual nominators equally and allowed a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.
...
If you wish to see how the sausage was made, we invite you to view the full & extensive discussions that led to E Pluribus Hugo at the following link: [INSERT LINK HERE]
************************
Given that Teresa & Patrick have provided this space for us to conduct these discussions, I'm fairly certain they would be agreeable to having them linked to the proposal. Teresa or Patrick, may we link the discussions here to the E Pluribus Hugo proposal?

#526 ::: junego ::: (view all by) ::: June 06, 2015, 11:44 AM:

Keith "Kilo" Watt @524
junego@523:

We'll submit the proposal and FAQs as one document. I broke them apart here just to make it easier to discuss.

So the FAQ will be posted/printed right below the proposal on the New Business page? I just can't find where any such documents are included with other proposals for Sasquan (if they exist).

#527 ::: Keith "Kilo" Watt ::: (view all by) ::: June 06, 2015, 11:48 AM:

junego@526:

That was the impression Kevin gave me. Basically whatever we submit will be posted on the Business Meeting site.

Kilo

#528 ::: junego ::: (view all by) ::: June 06, 2015, 02:18 PM:

Keith "Kilo" Watt @527
junego@526:

"That was the impression Kevin gave me. Basically whatever we submit will be posted on the Business Meeting site."

Thx for the info.

#529 ::: Jameson Quinn ::: (view all by) ::: June 06, 2015, 03:16 PM:

I prefer the prior wording of #22, but I have no strong objection to @525. Let's go with it. I think this is just about ready.

#530 ::: Keith "Kilo" Watt ::: (view all by) ::: June 06, 2015, 03:58 PM:

junego@525:

I don't have any strong objection to this wording, either, if everyone else is okay with it. The only thing I'd suggest is that I would keep the word "flaw", since that really gives the sense that something is broken and needs to be fixed. In my mind, a deficiency or limitation is something you can work around if you need to.

Kilo

#531 ::: Soon Lee ::: (view all by) ::: June 06, 2015, 04:04 PM:

junego#525:

[FAQ22]

I don't like the use of synonyms for "flaw". (Sorry) I think consistent use of the same word aids clarity. But I can live with it.

...treated all individual nominators equally [INSERT COMMA?] and allowed a consensus...

But yes, I think we're done too.

#532 ::: junego ::: (view all by) ::: June 06, 2015, 05:00 PM:

[PROPOSAL/COMMENTARY]
Here's what I started out trying to accomplish last night before I noodled out about the format and the FAQ.

I hadn't really looked the commentary over carefully before now. When I did I was really puzzled by the first sentence. What it says when you break it down is:

"the goal of our proposal is to acknowledge diverse opinions within fandom."

I think this ended up being something of an incomplete thought. So, in the interests of clarifying what the proposal's goals are and based on what people have said about what should be in the Commentary (more formal language than the FAQ, enough information for a positive reaction if this is all a person reads), I've edited the original as suggestions for discussion of final language.

First is my edit below the line. After my edit is another line. The current language is copied after that line (since it gets harder to scroll back to the last draft the longer the comments get.)

*********************************************
Commentary: “E Pluribus Hugo” – the name of this proposal is also one of its goals: recognizing the many opinions within fandom regarding which nominees might be worthy of a Hugo. Its other goals are promoting diversity in nominations while preventing outcomes that decrease diversity and treating all those who nominate equally. It accomplishes these goals by merely changing the way nominations are tallied.

It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members. They still list their nominees, unranked, up to the maximum permitted, in any categories they choose, just as they have in the past. However, because SF fandom typically picks a large variety of diverse nominees, it has always been vulnerable to a small, organized group concentrating their choices and preventing most or all other nominees from making the final ballot.

In the proposed system, finalists will be determined, not by first-past-the-post, as is done now and is the main cause of vulnerability, but by elimination of nominees with the least support, from the bottom up, using two criteria: 1) the current total nominations and 2) "points" generated by the number of nominees on each ballot in a category. This system has some similarities to the current final voting system in that as the nominees on a ballot are eliminated, more support is given to the remaining nominees. In this way, by eliminating the candidates with the least support in each round, fans' nominations slowly converge on a consensus of which finalists should be on the final Hugo ballot. Using this system, fandom isn't penalized for nominating a wide variety of candidates by the possibility of slate or bloc nominating.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no group of members -- of any sort, minority or majority -- can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Award.

***********************************************
[ORIGINAL]
Commentary: “E Pluribus Hugo” – the name of this proposal is also its goal: recognizing the many opinions within fandom as to what nominees might be worthy of the Hugo. It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members: They still list the nominees, unranked, they feel are Hugo-worthy (up to the maximum permitted) in any categories they choose, just as they have in the past. In the past, we have counted the number of nominations each nominee received, and the top five nominees were put on the final ballot. However, because SF fandom typically nominates a variety of different nominees, it was easy for an organized slate to make it so that no other nominees made the final ballot.

Using this system, fandom isn't penalized for nominating a wide variety of nominees. If you nominate something that ends up not having a chance to make the final ballot, then your remaining choices automatically get more of your support instead of just being wasted. In other words, you can safely nominate anything you feel is Hugo-worthy. If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot. In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which finalists should be voted on to be the Hugo winner.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no group of members -- of any sort, minority or majority -- can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Award.

#533 ::: junego ::: (view all by) ::: June 06, 2015, 05:06 PM:

#531 ::: Soon Lee ::: (view all by) ::: June 06, 2015, 04:04 PM:
junego#525:

[FAQ22]

I don't like the use of synonyms for "flaw". (Sorry) I think consistent use of the same word aids clarity. But I can live with it.
........
...treated all individual nominators equally [INSERT COMMA?] and allowed a consensus...
.......
But yes, I think we're done too.

Yes, on the comma, I think.

{GRIN} I'm obsessive about using the same word in a technical document, but almost anarchistic when in more casual usage. Tastes differ, I can live with having all the same word. If you want it changed back, I'm good.

#534 ::: David Harmon ::: (view all by) ::: June 06, 2015, 05:08 PM:

Another vote here for calling a flaw a flaw.

If you must "change it up", "limitation" is wrong on the sense of the word, and "deficiency" is both vague and oblique. "Weakness" might work for the first instance, and "vulnerability" or "problem" for the second replacement.

#535 ::: David Harmon ::: (view all by) ::: June 06, 2015, 05:22 PM:

junego #533: I like most of your new text, but the sentence "Using this system, fandom isn't penalized..." is awkward, and it loses some of the information from the original. How about: "Under this system, fandom's wide-ranging tastes no longer leave it vulnerable to slate or bloc voting. Fans can nominate what they like without worrying about strategic voting -- either their own or someone else's."?

#536 ::: David Harmon ::: (view all by) ::: June 06, 2015, 05:23 PM:

And my #535 should have been pointed at #532, natch.

#537 ::: Keith "Kilo" Watt ::: (view all by) ::: June 06, 2015, 06:50 PM:

junego@532:

I actually prefer the original wording, but was there a particular part of the original commentary that was bothering you? I had a number of goals for that section, though of course whether those are worthy goals is certainly open to debate. Still, if you can point out what parts you didn't like, we might be able to mesh the two ideas a bit better.

Basically, what I was trying to do for the commentary was come up with a rationale that sells the system based on an argument that most people could agree with. The salient concepts I was trying to get across are:

1) Fandom's opinions on what is good SF are many and varied.
2) The original nomination system has a flaw that allowed the Puppies to exploit that variety (if there weren't variety, then slates wouldn't work); fandom was effectively penalized for having that variety.
3) E Pluribus Hugo is designed so that fandom is -not- penalized for variety. Threfore, you can safely nominate anything you feel is worthy, without having to feel your vote might be wasted.
4) We explicitly are not trying to keep anyone off the ballot; we just want to keep any group from preventing -others- to be on the ballot.


You'll notice that the entire system is summed up in two sentences for the tl;dr crowd:

"If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot."


BTW, one other thing that was discussed in the previous thread is that we might want to avoid using the word "diversity" in the commentary, as this was one of the "fighting words" that the Puppy supporters would likely have an involuntary reaction to. I agree it's a much more concise word for what we're trying to say, though.


Kilo

#538 ::: Pfusand ::: (view all by) ::: June 06, 2015, 06:53 PM:

Junego@367

"Criteria" is plural; you want to use "criterion" in your 1.

#539 ::: Keith "Kilo" Watt ::: (view all by) ::: June 06, 2015, 07:13 PM:

pfusand@538:

I think it says "one of the criteria", which would be correct (since there it is plural). There could be another spot I'm missing, though. (#489 is the latest FAQ, I believe.)

K

#540 ::: Soon Lee hailing TNH and PNH ::: (view all by) ::: June 06, 2015, 09:21 PM:

Teresa & Patrick,

May we link to the the discussions here at Making Light in the "E Pluribus Hugo" Proposal & FAQ?

The current iteration of the Proposal & FAQ in full are upthread and when done, we plan to submit both to the WSFS Business Meeting. The reference & link to the Making Light discussions will be at the end of FAQ22, the latest version of which I have included below:

[FAQ22]

22. Isnt't this change just designed to keep certain people out of the Hugo nomination process?
No. E Pluribus Hugo is designed to fix the flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the award is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.

Although it is true that the discussions for this system were hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resisted bloc or slate nominating, treated all individual nominators equally, and allowed a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.
If you wish to see how the sausage was made, we invite you to view the full & extensive discussions that led to E Pluribus Hugo at the following link: [INSERT LINK HERE]
#541 ::: David Goldfarb ::: (view all by) ::: June 07, 2015, 01:10 AM:

BTW, has anyone talked to the Hugo Award administrators about getting this year's nomination data? (Anonymized, of course.) I'm really curious to know how the ballot would have looked with "E Pluribus" in place, and I'm sure I'm not alone in that.

#542 ::: junego ::: (view all by) ::: June 07, 2015, 06:25 AM:

Keith "Kilo" Watt @537
junego@532:

I actually prefer the original wording, but was there a particular part of the original commentary that was bothering you? I had a number of goals for that section, though of course whether those are worthy goals is certainly open to debate. Still, if you can point out what parts you didn't like, we might be able to mesh the two ideas a bit better.

I'll outline what I think the goals of the section should be and we can see where our ideas overlap or not, ok?

Basically, what I was trying to do for the commentary was come up with a rationale that sells the system based on an argument that most people could agree with. The salient concepts I was trying to get across are:
...
1) Fandom's opinions on what is good SF are many and varied.
2) The original nomination system has a flaw that allowed the Puppies to exploit that variety (if there weren't variety, then slates wouldn't work); fandom was effectively penalized for having that variety.
3) E Pluribus Hugo is designed so that fandom is -not- penalized for variety. Threfore, you can safely nominate anything you feel is worthy, without having to feel your vote might be wasted.
4) We explicitly are not trying to keep anyone off the ballot; we just want to keep any group from preventing -others- to be on the ballot.

My outline/goals are:

1) Clearly state what our goals are. This should be a selling point.
2) Clearly state how the proposal accomplishes those goals. This should be a selling point.
3) Outline, very briefly, how the system works.
4) As part of that outline emphasize that the change will be transparent to the membership. This is a big selling point and includes the idea that you can nominate exactly like you always have. (Yes, keep the variety)
5) You won't have to strategize or worry about slates or what others may do because the new tallying system prevents anyone else from nullifying your choices. Another, related, big selling point. (Yes, keep the variety).
6) The proposal is neutral and doesn't favor either the majority or a minority; it doesn't keep anyone from participating. A selling point.
7) This section should outline and sell our whole 'argument' for those who don't read the FAQs. Else the FAQs should be the Commentary.


You'll notice that the entire system is summed up in two sentences for the tl;dr crowd:
...
"If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot."


Sorry, unless I'm misreading the intended meaning, I don't see the above wording as THE quintessential reason anyone should agree with changing the Constitution.

In the current system, the majority of people don't get any (or maybe 1-2 in the whole ballot) of their nominations on the ballot. My understanding is that E Pluribus may only increase your odds very slightly, if at all. It's not a selling point.

This whole section, Commentary, is the tl;dr. But I would say the biggest selling point is and should be that no one will have to change their nominating habits to reduce the power of slates.

BTW, one other thing that was discussed in the previous thread is that we might want to avoid using the word "diversity" in the commentary, as this was one of the "fighting words" that the Puppy supporters would likely have an involuntary reaction to. I agree it's a much more concise word for what we're trying to say, though.

If I used that dread word anywhere (and I don't remember right now) it can, of course, be synonimized. ;^]

#543 ::: P J Evans ::: (view all by) ::: June 07, 2015, 01:10 PM:

542
In the current system, the majority of people don't get any (or maybe 1-2 in the whole ballot) of their nominations on the ballot. My understanding is that E Pluribus may only increase your odds very slightly, if at all. It's not a selling point.

You keep misunderstanding the way the system is set up. THIS IS SELECTING BY POPULARITY ONLY. It isn't designed or intended for a minority to have a guarantee of any slot. It isn't a designated-slot system at all. [/cluebat]

#544 ::: junego ::: (view all by) ::: June 07, 2015, 03:24 PM:

P J Evans @543
542 junego said:
"In the current system, the majority of people don't get any (or maybe 1-2 in the whole ballot) of their nominations on the ballot. My understanding is that E Pluribus may only increase your odds very slightly, if at all. It's not a selling point."

You keep misunderstanding the way the system is set up. THIS IS SELECTING BY POPULARITY ONLY. It isn't designed or intended for a minority to have a guarantee of any slot. It isn't a designated-slot system at all. [/cluebat]

Can you explain what your response has to do with the part of my post you're quoting? Or *any* part of my post?

Are you denying that many or most people don't get any of their nominations onto the final ballot in the current system?

Are you denying that the proposed system will not change that fact at all or only very slightly?

Did you misread my post, maybe?

#545 ::: junego ::: (view all by) ::: June 07, 2015, 05:38 PM:

David Harmon @535
[PROPOSAL/COMMENTARY]

junego #533: I like most of your new text, but the sentence "Using this system, fandom isn't penalized..." is awkward, and it loses some of the information from the original. How about: "Under this system, fandom's wide-ranging tastes no longer leave it vulnerable to slate or bloc voting. Fans can nominate what they like without worrying about strategic voting -- either their own or someone else's."?

Hi, David. I thought I answered this yesterday, but don't see it in the thread.

I agree that your change makes it clearer and explains the point really well. It gets my vote.

#546 ::: David Harmon ::: (view all by) ::: June 07, 2015, 08:27 PM:

junego #544 Are you denying that many or most people don't get any of their nominations onto the final ballot in the current system?

Pardon? I could possibly believe that for some individual categories, but that gets stretchier for the less competitive categories, and really dubious for the entire ballot. I'd want to see numbers.

#547 ::: P J Evans ::: (view all by) ::: June 07, 2015, 09:15 PM:

544
Y3es. Because before this year, that was how it worked. This year we had a slate - well, one and a half or two, depending on who you believe - that subverted the process and locked out the majority.

Which you should be aware of, since you've been here for most of the discussion. (Also, since you have a set of the nominations for 1984, you can see how it actually is done.)

#548 ::: junego ::: (view all by) ::: June 08, 2015, 12:06 AM:

David Harmon @546
junego #544 Are you denying that many or most people don't get any of their nominations onto the final ballot in the current system?
....
Pardon? I could possibly believe that for some individual categories, but that gets stretchier for the less competitive categories, and really dubious for the entire ballot. I'd want to see numbers.

Am I mistaken? I certainly could be misremembering the numbers discussed in different threads. I thought it was reported that at least a substantial plurality of people (if not the majority) don't get any of their nominees on the final ballot.

I know I've read anecdata from people on different blogs and discussion boards saying that most, if not all, of their noms every year don't make the ballot or that it's only one or two every few years that succeed. From the way it's been discussed, I had assumed this was general information. I'm a new member and have no personal experience in the matter.

But I did do some quick, back-of-the-envelope checks on the written story categories from what Hugo data is available and it seemed to support the claims. If there are an average of 3 nominations per ballot per category (number mentioned in these discussions) and you look at what percentage of those nominations are represented by the finalists in a category in a year, it tends to run in the 10%-30% range of all (estimated) nominations.

Taken all together I have been operating on this assumption being basically correct.

On your point that this wouldn't be true in some of the less competitive categories (graphic story, editor, fanzine, etc) - a whole lot fewer people nominate anything in those categories. So overall their somewhat higher hit rate may not counter the much higher numbers that miss in novel, short story, etc. OTOH I haven't heard people talking about those categories and I'm not sure if anyone was using those when crunching numbers earlier, you may have a point.


#549 ::: Keith "Kilo" Watt ::: (view all by) ::: June 08, 2015, 12:06 AM:

I think we may be talking past each other here a bit. I'm not totally sure -I- understand exactly what everyone is saying...


I think the main difference in approaches for the commentary is that I'm not looking to -fully- explain the system here. I just want to sell it as a Good Thing. Fundamentally, the problem this year really was that the wide diversity of works that are typically nominated means that there is no concentration of votes, even with a large electorate. Slates take advantage of that fact. Even though they have fewer people nominating, their votes are concentrated, so they easily swamp the other voters. Currently, the only way to counter that is with another slate -- which means the wide diversity of nominations will be gone. This diversity of opinion is something we value in fandom -- it's why the opposition to slates (and slate wars) is so strong. E Pluribus Hugo at its core means that this diversity of opinion no longer works against the electorate. They can, as you say, continue to nominate as they always have without the fear that unless they bloc vote their opinions have no chance of making the final ballot. That's a very strong selling point indeed -- it's the whole reason we're doing this and one reason we rejected a number of other systems early on.

This issue of not being penalized for nominating a variety of works doesn't appear anywhere else in the proposal or in the FAQ. It's not really a "question" that anyone would naturally think to ask, so really doesn't make sense in the FAQ. On the other hand, suppose you have a member at the business meeting who doesn't want to read the legalese and doesn't really want to follow the intracies of the system itself. This idea of "hey, you can now safely nominate a wide variety works without worrying that you're helping a slate keep your nominations off the ballot" is a good way to convince that member to support the proposal. While we could take any number of approaches for the commentary (and there's not a thing wrong with what you suggested), I think what you propose works better in the FAQ (in fact, it's essentially the first question in the FAQ). It seems to me that the argument I made in the original commentary is a good use of that space -- it will help convince those who don't -want- to know anything about the inner workings of the system, but want to know why the system is a Good Thing.

Kilo

#550 ::: junego ::: (view all by) ::: June 08, 2015, 12:13 AM:

P J Evans @547

544
Y3es. Because before this year, that was how it worked. This year we had a slate - well, one and a half or two, depending on who you believe - that subverted the process and locked out the majority.
...
Which you should be aware of, since you've been here for most of the discussion. (Also, since you have a set of the nominations for 1984, you can see how it actually is done.)

I have no idea what you're on about. Nothing you've said in this or your previous post has anything to do with what I've posted. You seem to be having a conversation with someone else. I'll bow-out unless and until you actually respond to something I've said.

#551 ::: P J Evans ::: (view all by) ::: June 08, 2015, 02:22 AM:

550
I respond to what you say (with its implications). You've been not quite getting this from the time you first posted.

It's always been about preventing organized groups from stuffing the ballot box and keeping others from having their say, as happened this year, in spades.

#552 ::: David Harmon ::: (view all by) ::: June 08, 2015, 05:54 AM:

junego #548: If there are an average of 3 nominations per ballot per category (number mentioned in these discussions) and you look at what percentage of those nominations are represented by the finalists in a category in a year, it tends to run in the 10%-30% range of all (estimated) nominations.

Except that's not the number I was challenging you on. 10-30% of the nominations means more than that per category, which means yet more per ballot.

Unless you're claiming that's the proportion of ballots nominating anyone who turns out to be one of the finalists in any category? For that number, I could believe under 50%, but 10% seems pretty dubious.

#553 ::: Jameson Quinn ::: (view all by) ::: June 08, 2015, 08:44 AM:

@551: It would be easier to understand your point if you'd make directly, rather than belittling people.

@552: I believe the number may be under 50% in "smaller" categories, but is probably well above 50% in Best Novel at least. And I agree that 10% is way too low. This is my general understanding and familiarity with the data speaking; I have not closely investigated this specific question with the data.

#554 ::: Jameson Quinn ::: (view all by) ::: June 08, 2015, 10:09 AM:

This is very close to done; the current wrangling seems to be about relatively minor wording changes (insofar as I even understand what's being debated now).

I'd like to also have a proposal to lift the 5-per-category voting restriction. So, I have a few questions to go with that:

1) Would people here be interested in helping me draft that?
2) Ditto, signing on?
3) What should I call the proposal? Would calling it "E pluribissimus Hugo" be OK? It's clearly related, but I wouldn't want to harm the chances of the main proposal in any way.

#555 ::: Kevin Standlee ::: (view all by) ::: June 08, 2015, 11:27 AM:

Jameson Quinn @554:

I'd like to also have a proposal to lift the 5-per-category voting restriction.
That's a modification of one of the two parameters in the "4/6" proposal, and thus properly should be discussed during the Preliminary Business Meeting. I'm strongly encouraging any discussion of either the number of nominations per member per category or the number of finalists per category to be grouped into the discussion of the PBM. It's much easier to handle these things if we can keep all related discussions in the same bucket, It think.

Thus, if you want to increase the number of nominations per member per category, propose it as an amendment to 4/6 at the PBM.

In fact, I'm expecting lots of different possibilities there, to the extent that it's likely to be a case of using the system of "filling blanks" with all of the different possibilities that individuals want, with a multi-way argument about why any given one is better than another (or why no change at all is preferable).

#556 ::: junego ::: (view all by) ::: June 08, 2015, 11:48 AM:

David Harmon @552
junego #548: If there are an average of 3 nominations per ballot per category (number mentioned in these discussions) and you look at what percentage of those nominations are represented by the finalists in a category in a year, it tends to run in the 10%-30% range of all (estimated) nominations.
...
Except that's not the number I was challenging you on. 10-30% of the nominations means more than that per category, which means yet more per ballot.
...
Unless you're claiming that's the proportion of ballots nominating anyone who turns out to be one of the finalists in any category? For that number, I could believe under 50%, but 10% seems pretty dubious.

Well, crud. I left out a step in my rough calculations when I did them a coupla weeks ago, so the claim that the majority do *not* get *any* of their nominees on the ballot is incorrect. (There is no direct measure of what percentage of total ballots in each category had at least 1 successful nominee, so it's all estimates to get a rough idea.)

I just spent a good while recalculating different categories for different years and in the big ones - novel, movie - it appeared to always be more than 50% of the ballots had at least 1 success. In a few categories there were occasional instances where obviously less than 50% of ballots had a success. In most other categories it appeared to probably be more than 50% most of the time with some instances of uncertainty.

So, overall, the majority of nominators probably get at least 1 or more of their nominees on the final ballot.

I had gotten this idea into my head from what other people said and the original check I did on the claim was sloppy.

Thank you for challenging me and making me recheck my assumptions.

I retract the places where I said that *most* people don't get a nominee on the ballot.

#557 ::: Keith "Kilo" Watt ::: (view all by) ::: June 08, 2015, 11:50 AM:

All:

I think Jameson is right that there is very little left to do here, but I'd like to wrap it up and get it submitted. Kevin has said that we can make changes in response to community feedback after we submit, but that submitting earlier rather than later is a good idea.

So, it seems to me that the only real remaining issue is the commentary, but I confess that I don't fully understand the issues that are being debated here, so I don't know where to go on a re-write (or if there should be a re-write at all). Could everyone weigh in with their thoughts? For reference, junego's 532 contains both the original text as well as a proposed re-write. 537 and 549 have my thoughts on what this section should be used for; 542 has juengo's thoughts (though appartently there is some contention here).

Let's iron out this last bit and send the proposal out into the wild.

Thanks much,
Kilo

#558 ::: Jameson Quinn ::: (view all by) ::: June 08, 2015, 11:56 AM:

@555: You're the authority here, so I really want to do it in a way you think is best. But on the other hand, I think that it would be better to have separate votes on 4/6 and on my "no limits" proposal. Yes, they are modifying the same parameters. But there is a clear constituency for 4/6, and I think they deserve to get a clean up-or-down vote on their proposal as it stands; whereas my "no limits" idea is, properly speaking, piggybacking on the EPH proposal. (But I don't want to make it an amendment to EPH, because EPH should get a clean up-or-down vote before considering extras.)

So, would it be OK if the "no limits" proposal were separate from 4/6, but included language saying "if this gets fewer yes votes than 4/6, this proposal is null and void; and vice versa"?

Again, I want to do this in a way that you like. But honestly I'd rather stand aside than try to do something that would at best prevent a clean up-or-down vote on 4/6. I hope my proposal would have more support than 4/6, but I don't want to do anything that could cause collateral damage of any kind, even to proposals I myself don't actually support.

#559 ::: Duncan J Macdonald ::: (view all by) ::: June 08, 2015, 12:18 PM:

[COMMENTARY]
Keith "Kilo" Watt @ 557: Could everyone weigh in with their thoughts?

I see no need to re-write the original. While I agree with juengo @542's list of goals, I don't see how they don't mesh with the current wording.

My recommendation is to stay with the current wording.

#560 ::: Duncan J Macdonald ::: (view all by) ::: June 08, 2015, 12:37 PM:

Jameson Quinn @554:

I'd like to also have a proposal to lift the 5-per-category voting restriction. So, I have a few questions to go with that:

1) Would people here be interested in helping me draft that?
2) Ditto, signing on?
3) What should I call the proposal? Would calling it "E pluribissimus Hugo" be OK? It's clearly related, but I wouldn't want to harm the chances of the main proposal in any way.

I'd love to see a discussion of that as well. Searching back through your posts I've found only two other references to lifting the 5-per-category limit for nominations, and neither of them have any meat behind them.

In one, you explicitly tie lifting the 5-per-category to your 'Quasi long-list' idea. Since I am already on record as disliking [insert stronger word here] any of the long-list proposals, I'm not sanguine on liking removing the 5 nomination limit.

So, in answer to your bullet points, 1) No, but I'd like to discuss the idea. 2) No, not with out convincing that it is/would be a good idea, 3) How about "Nominate Everything and Let God Sort 'Em Out?"

#561 ::: Kevin Standlee ::: (view all by) ::: June 08, 2015, 01:08 PM:

Jameson Quinn @558:

So, would it be OK if the "no limits" proposal were separate from 4/6, but included language saying "if this gets fewer yes votes than 4/6, this proposal is null and void; and vice versa"?
If you insist on doing so, there's nothing technically wrong with doing so. Heck, there's nothing technically wrong with every single 4/6 variation being considered completely separately, other than doing so would cause parliamentary chaos. The Chairman can't prevent member of an organization from tangling themselves in procedural knots, just try to warn them that there are rocks ahead.

You are also of course effectively requiring counted voted on both proposals (which aren't required if an uncounted show of hands is obvious one way or another). That's not a problem; I just have to remember to do counts in order to administer the provision of your proposal, assuming it clears the PBM.

#562 ::: Annie Y ::: (view all by) ::: June 08, 2015, 01:49 PM:

[COMMENTARY]
Keith "Kilo" Watt @ 557

I like the current wording...

#563 ::: Cally Soukup ::: (view all by) ::: June 08, 2015, 02:18 PM:

Jameson Quinn: I'm not fond of the idea of making a superlonglist. Five nominees per category seem like a reasonable number to me; after all, we are looking for what WSFS members consider to be the BEST SFF they've read that was published in a qualifying year, not ALL the SFF they've read.

#564 ::: David Harmon ::: (view all by) ::: June 08, 2015, 02:35 PM:

junego #556: And thank you for a classy walkback! Just to be clear, you're saying that finalist representation for each category mostly covered 50% of ballots?

Hmm, now I wonder how much it correlates between categories. Random overlap would suggest a very small fraction¹ of ballots with no finalist in any category. But how well does that match the data?

¹ on the order of (1/2)^n, where n is the number of categories

#565 ::: Jameson Quinn ::: (view all by) ::: June 08, 2015, 03:05 PM:

@563: That's not what I meant at all. I meant, you can nominate as many as you want; there would still be 5 (or occasionally 6) finalists.

#566 ::: Jameson Quinn ::: (view all by) ::: June 08, 2015, 03:07 PM:

@565: By "can", I mean "should be able to, if my extra proposal were to pass".

#567 ::: Brad from Sunnyvale ::: (view all by) ::: June 08, 2015, 03:11 PM:

Hi. I've been away on the road for several weeks, so my comments are late, but.

a) I do generally prefer this to some of the other proposals from the first thread, but

b) I disagree with the claims it is less subject to strategic voting.

To wit: Consider the simplified case where you strongly support a lesser known great work, and also a widely known great work you feel is very likely to be nominated without you. You might consider two nomination ballots:

1) Both "Obscura" and "Shoo-in" -- which is your true opinion
2) Only "Obscura"

If you use ballot #1 you grant 0.5 points to both works in any tally, and if Obscura is eliminated you give 1 full point to Shoo-In.

If you use ballot #2 you give one full point to Obscura, which means twice as much support -- it is as though you got the chance to cast two ballots in the old system. You feel comfortable that Shoo-In does not "need" your support, though in rare cases you may misjudge, or others may do this, and it might fail to be nominated.

While this failure would be rare, the real issue is you have a strong incentive to not nominate your true heart.

It gets worse in the full-5 case. There, if you list 4 better known works you support, as well as Obscura, the likely result is you give only 0.2 points of support to Obscura when it gets down to the wire, when you could give in theory give it five times as much if you hide your opinion about the other 4 works -- and they are all top contenders.

This is actually, to my mind, one of the strongest incentives for strategic voting that you can present -- you can, in theory, quintuple your support. Indeed, a writer with a dedicated fan base or even large group of friends could, by convincing that base to nominate her work and only her work, without much effort. Fans are sometimes rabid fans.

Now it's not as strong as it appears, because as elimination takes place, the other ballots who named only one contender also contribute a full point to it. But you can certainly increase your chances by promoting this strategy.

#568 ::: Cally Soukup ::: (view all by) ::: June 08, 2015, 05:01 PM:

If you can't narrow down your reading to only five works you think were the best of the best, how ever are you going to vote for the one winner on the final ballot? Allowing infinite nominations per person invites a simple "this is what I read last year" with no thought behind it.

#569 ::: Jameson Quinn ::: (view all by) ::: June 08, 2015, 05:08 PM:

@567: Considered in a purely self-interested, "rational" framework, whether strategy is worthwhile comes down to an expected value calculation. You're weighing two possibilities:

A. The chance your strategy pays off, and your extra vote was key to ensuring Obscura becomes a finalist.

B. The chance your strategy backfires, and the lack of your vote was what made a difference in Shoo-in not becoming a finalist.

In order for (A) to happen, Obscura has to get all the way from 6th place or below in points, up to 4th place or above, a fraction of a point at a time. You have to believe that other Obscura fans will also be strategic, while the fans of your hated 5th place rival work ("Darkness") will not be. You also have to hope that a significant number of Darkness fans like some more-popular work such as Shoo-in. But most of all, you have to be pretty accurate in your assessment that it's at 6th place without strategy.

In order for a backfire (B) to happen, all that has to be true is that various people agree that Shoo-in is likely to win, and thus strategically not vote for it, just as you have.

Fundamentally, it's easier to know what the winner or second place will be, than it is to know exactly which work will be in 6th place. Thus, the risks of (B) will always be more knowable in advance than the hope of (A).

Now, if you actually care significantly more about (A) than about (B), the expected value of strategy could still be positive. But that means that you actually prefer Obscura over Shoo-in. In that case, we're talking about a strategy that is at least semi-honest, and is arguably nothing but maximum honesty. I don't see that as any kind of a problem.

#570 ::: P J Evans ::: (view all by) ::: June 08, 2015, 05:20 PM:

569
No more than 5 names into the bin, please. This isn't about strategy or about ranking in order of preference.

I think we've done this argument already, and the longlist side lost. Why is it coming up again?

#571 ::: Jameson Quinn ::: (view all by) ::: June 08, 2015, 05:21 PM:

@568: In this system, there is no reason that voting for 5 works is a more appropriate limit than 4 or 3, and in many ways it's less appropriate. In fact, you can make the same argument you just made, and just replace "five" with "three", and it's just as valid. But if you take that logic to its absurd extreme, that would mean each voter could only nominate one work; and there are many, many reasons why that would make an absolutely horrible voting system, including the fact that it would make it so a moderately organized voting bloc could probably dominate a category, and the fact that most votes would be wasted.

If there were some voter who wanted to vote for everything they'd read in the last year, it is likely that they would end up voting for more than one finalist, and thus having their vote not count for much in the end. It's also not unlikely that they would have voted for both 5th and 6th place, so that they had no say in which of those two was eliminated. I agree that would be a pretty stupid way to vote, but in the end, they'd only really be hurting themselves.

#572 ::: Jameson Quinn ::: (view all by) ::: June 08, 2015, 05:24 PM:

@570: This is entirely separate from the debate we referred to as "longlist". There's really nothing at all in common. What makes you think there's any similarity? I don't even know how to respond, because I don't understand why you would say that.

#573 ::: Jameson Quinn ::: (view all by) ::: June 08, 2015, 05:28 PM:

@570: limiting people to 5 choices increases strategic incentives, including (for people who decide to be strategic) more thought about preference orders. Lifting that limit makes an unstrategic electorate more resistant to slate voting. These are undeniable mathematical facts; I'd be happy to explain them, but first I'd like to understand where you are coming from that your intuition points the opposite way.

#574 ::: Jameson Quinn ::: (view all by) ::: June 08, 2015, 05:32 PM:

Oh, wait. I think I see how we got our wires crossed. You thought that @569 had to do with @554, @565, etc? It doesn't; it was a response to @567, and it relates to the system as it is currently proposed. Separate issues.

If I've understood correctly now, then I apologize for being confused about @570.

#575 ::: junego ::: (view all by) ::: June 08, 2015, 07:45 PM:

[PROPOSAL/COMMENTARY]

Since I seem to be outvoted, I will withdraw my suggestions for a revision of the Commentary. If that's all you need to know, stop here. Below the line are my pro forma objections to the present Commentary and one red flag.

**************************************

1) I object, somewhat strenuously, that the goal mentioned for E Pluribus Hugo is to acknowledge the variety of opinion within fandom. That's the only goal at the beginning of our proposal. I don't see how this starts a rational presentation for why anyone should support this amendment.

2) This would have been the most logical place to set out the problem this amendment is trying to solve, what the goals are and how the amendment meets the goals and solves the problem. As it's now written some of these aspects are there somewhere, but are not clearly stated.

3) I disagree that there is one main point that is the 'killer' argument that will convince people to support amending the constitution. Different people will be swayed by different ideas. Instead of spending most of the Commentary trying to sell this one idea, it makes more sense to advance, briefly, all main "selling points" and rely on the FAQs to give details.

4) Last, but important (and only occured to me this morning) is that we've not addressed any arguments to the people at the BizMeet who don't want to change anything. This may be a significant group.

#576 ::: Brad from Sunnyvale ::: (view all by) ::: June 08, 2015, 09:14 PM:

Quite simply, Jameson #596, that is what being a fan of an obscure work is all about. People love to promote things that they love that they feel are neglected. There is not nearly so much personal satisfaction in declaring your love for already popular works as there is in promoting your more obscure favourites.

This even happens in today's existing approval system, but it's much less useful to do there, and so has less of an effect. Here you can make a bigger difference, and give your obscure favourite a much better chance.

It is always the case that it is rare for one ballot to make the difference -- and indeed many people don't bother to participate in many elections because of that fact. But if you do believe you can make a difference, you are often interested in maximizing it.

Now, I personally believe the Hugo nominations should be a measurement of fannish opinion, not a vote or a contest. As such I believe we want people to provide the most honest ballot they can, stating their true opinion. I oppose systems which would make me want to lie and thus presumably make others want to lie.

This is not idle. Looking at the 2012 ballot, I believe I nominated Embassytown, Leviathan Wakes, Quantum Thief, and Ready Player One that year (I don't recall the other choice) but my clear favourite was The Quantum Thief. I believe in this system had I nominated only it, for example, it would have made the ballot -- perhaps at the expense of one of my other choices, Leviathan Wakes. It is harder to say what I would have done without the knowledge of hindsight. (In fact, in hindsight, leaving LW off my ballot would have had them both get on in a tie under the existing system, but you obviously don't plan for that.)

In this case, though, dropping Embassytown and LW from my ballot would have significantly increased the support I gave to QT or RP1 and I would be happy with that, particularly in the case of Embassytown which is by an author well loved by Hugo voters, when the others were by newcomers.

If this amendment passes, I suspect I will falsely nominate moderately often from now on, at least if I nominate for selfish reasons, and that is saddening.

#577 ::: Jameson Quinn ::: (view all by) ::: June 08, 2015, 11:09 PM:

@576: so you're saying that, due to cognitive biases, people care more about an less-known work, even when they judge its quality to be slightly less than a better-known work? And that voting based on what they care about, rather than on what they judge as being high quality, they would be being dishonest? And that this would lead to poorer outcomes for the award and its prestige?

You may be right on the first claim. The second is just a matter of definition. But on the third, i think you're really overestimating the size of the impact as well as being pessimistic about its direction. Chances are that a work that's perceived as overwhelmingly popular, will be, despite some strategic voters neglecting it. And if strategy is enough to bring a niche work to the fore, it probaably legitimately deserves the extra attention.

I doubt that most people will be as strategic as you fear; I doubt that if they did it would have much impact; and I doubt that if it did the impact would be particularly negative.

#578 ::: David Dyer-Bennet ::: (view all by) ::: June 08, 2015, 11:31 PM:

Brad@567: Yes, that's the obvious line of reasoning leading to strategic voting.

The thing is -- the chances that you actually know the things you need to know to make that argument true are extremely, I would say vanishingly, small. Jameson Quinn has laid some of that out in responses to you.

AND, if you miscalculate, you're as likely to do harm as good (relative to your goals, I mean).

There are situations where strategic voting wins (there are in any system, most especially simple first-past-the-post), BUT in this system, you can't tell in advance whether you're in one of those situations or not. Assuming you're not, and hence just voting your actual preferences, is the best strategy you can pick based on information available at the time you have to decide.

#579 ::: David Dyer-Bennet ::: (view all by) ::: June 08, 2015, 11:36 PM:

I think perhaps discussing a proposal for unlimited rather than 5 nominations would be more constructive elsewhere (another post here, a post on some other blog, whatever). Admittedly we're close to finished with E Pluribus Hugo development, but we're not quite, and mixing our final work with this new discussion is going to lead to errors, things missed, and people wandering off I'm afraid.

(I think I might support the unlimited nominations proposal, not that it matters since I won't be at Sasquan. So I hope it gets well-discussed and perhaps proposed and even passed. But I don't want it to damage the other ongoing work here.)

#580 ::: Soon Lee ::: (view all by) ::: June 09, 2015, 02:05 AM:

[PROPOSAL/COMMENTARY]

I agree with Duncan J Macdonald @559 that the current wording is fine.

junego #575:
I do think that the current version, between the commentary & FAQs do cover your first three points. As for your 4th point, I don't know how much more we can do: if given the evidence of this year's ballot, the will of the Business Meeting is to vote for no change, then so be it. We have done what we can.

Maybe I'm suffering from overfamiliarity with the current text, so I'm wondering if there are fresh eyes we can get to take a look at the current draft?

Jameson Quinn #554:
While I don't have a strong opinion about it, I think that any proposal other than the "E pluribus Hugo" should be discussed somewhere else. I'd like for us to stay focused on getting this done, then submitted so the general membership get a chance to go over it properly. I'm also worried we might end up in a Pareto Principle situation where the last 20% of the work takes 80% of the effort.

Shorter me: let's get this done first.

#581 ::: Tim Illingworth ::: (view all by) ::: June 09, 2015, 11:08 AM:

Minor nitpickery: in Section 3.A I'd be inclined to drop the "(See 3.A.3 for ties.)" in 3.A.1 and split 3.A.3 into two.

3.A.3: Ties During the Selection Phase:
(1) If two or more nominations are tied for the lowest point total, all such nominees shall be selected for the Elimination Phase.
(2) If one nominee has the lowest point total and two or more nominees are tied for the second-lowest point total, then all such nominees shall be selected for the Elimination Phase.

3.A.4: Ties During the Elimination Phase:
(1) If two or more nominees are tied for the fewest number of nominations, the nominee with the lowest point total at that round shall be eliminated.
(2) If two or more nominees are tied for both fewest number of nominations and lowest point total, then all such nominees tied at that round shall be eliminated.

This looks cleaner to me, though it doesn't really affect the wording much.

#582 ::: David Dyer-Bennet ::: (view all by) ::: June 09, 2015, 12:08 PM:

I really like having the forward reference "(See 3.A.3 for ties)". I don't like text that puts me into a confused state deliberately, and those two sections not giving rules for resolving ties would bother me extremely. That sort of thing makes me stop and waste time trying to figure out how it makes sense; and of course it doesn't, until you get to the backward-reaching instructions later on. Having a reference to the rules for ties makes me much happier.

#583 ::: Keith "Kilo" Watt ::: (view all by) ::: June 09, 2015, 01:26 PM:

Tim@581:

I can see what you're saying here, but my only issue with it is with the "weighting" of the subsections, for want of a better word. The way it's written currently, we have "phases", "repeat", "ties", and "removals". We can point to a single section for each topic. Breaking ties into two sections sort of upsets that balance. It's purely an aesthetic thing, so it's not a major issue for me. It just seems to me that it's easier to point to a single section with something like "Ties? Those are handled in 3.A.3," rather than having to point to two different places. That's just an opinion, though.

I'm planning to post the next candidate final draft this afternoon, so please let me know of any other issues we should discuss. I think we're very close...

Thanks,
Kilo

#584 ::: rcade ::: (view all by) ::: June 09, 2015, 06:37 PM:

[SIGNING] and possibly [ATTENDING], though I won't be able to confirm that for a while.

Thank you for tackling this, Kilo. I look forward to the consideration of this proposal.

Rogers Cadenhead

#585 ::: Keith "Kilo" Watt ::: (view all by) ::: June 09, 2015, 07:23 PM:

All:

Okay, as promised, here's the next candidate final draft. Are we done?

Kilo

#586 ::: Keith "Kilo" Watt ::: (view all by) ::: June 09, 2015, 07:24 PM:

Short Title: E Pluribus Hugo (Out of the Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations), section 3.9 (Notification and Acceptance), and section 3.11 (Tallying of Votes) as follows:

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the process described in section 3.A.

Insert new section 3.A after Section 3.8 as follows:
Section 3.A: Finalist Selection Process

3.A.1: For each category, the finalist selection process shall be conducted as elimination rounds consisting of three phases:

(1) Calculation Phase: First, the total number of nominations (the number of ballots on which each nominee appears) from all eligible ballots shall be tallied for each remaining nominee. Next, a single “point” shall be assigned to each nomination ballot. That point shall be divided equally among all remaining nominees on that ballot. Finally, all points from all nomination ballots shall be totaled for each nominee in that category. These two numbers, total nominations and total points, shall be used in the Selection and Elimination Phases.
(2) Selection Phase: The two nominees with the lowest point totals shall be selected for comparison in the Elimination Phase. (See 3.A.3 for ties.)
(3) Elimination Phase: Nominees chosen in the Selection Phase shall be compared, and the nominee with the fewest total nominations shall be eliminated and removed from all ballots for the Calculation Phase of all subsequent rounds. (See 3.A.3 for ties.)

3.A.2: The phases described in 3.A.1 are repeated in order for each category until the number of finalists specified in 3.8.1 remain. If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then instead no nominees will be eliminated this round, and all remaining nominees shall appear on the final ballot, extending it if necessary.

3.A.3: Ties shall be handled as described below:
(1) During the Selection Phase, if two or more nominations are tied for the lowest point total, all such nominees shall be selected for the Elimination Phase.
(2) During the Selection Phase, if one nominee has the lowest point total and two or more nominees are tied for the second-lowest point total, then all such nominees shall be selected for the Elimination Phase.
(3) During the Elimination Phase, if two or more nominees are tied for the fewest number of nominations, the nominee with the lowest point total at that round shall be eliminated.
(4) During the Elimination Phase, if two or more nominees are tied for both fewest number of nominations and lowest point total, then all such nominees tied at that round shall be eliminated.

3.A.4: After the initial Award ballot is generated, if any finalist(s) are removed for any reason, the finalist selection process shall be rerun as though the removed finalist(s) had never been nominee(s). None of the remaining original finalists who have been notified shall be removed as a result of this rerun. The new finalist(s) shall be merged with the original finalists, extending the final ballot if necessary.

Section 3.9: Notification and Acceptance.
3.9.1 Worldcon Committees shall use reasonable efforts to notify the nominees finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each nominee finalist shall be asked at that time to either accept or decline the nomination. If the nominee a finalist(s) declines the nomination, that nominee finalist(s) shall not appear on the final ballot. The procedure for replacement of such finalist(s) is defined in subsection 3.A.4.

Section 3.11: Tallying of Votes.
3.11.4: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. During the same period, the results of the last ten rounds of the finalist selection process for each category (or all the rounds if there are fewer than ten) shall also be published.

Submitted by: Keith “Kilo” Watt, Jameson Quinn, Tammy Coxen, Yoana Yotova, Joshua Kronengold, Christopher Battey, David Gallaher, Adam Tilghman, Sara Watt, CJ Cabourne, Steven Halter, P J Evans, David Goldfarb, Seth Gordon, Ginger Tansey, Steve Wright, Catherine Faber, Andy Holloway, Duncan J. Macdonald, Claudia Beach, Derry Earnshaw, Jason Skiles, Soon Lee, David Harmon, Lydia Nickerson, Abigail Sutherland, Lee Billings, Oskari Rantala, Seth Breidbart, Chris Suslowicz, Fragano Ledgister, Lori Coulson, Jeffry Herman, Mark Shier, Buddha Buck, Lenore Jean Jones, David Langford, Christopher Hatton, Rogers Cadenhead

Commentary: “E Pluribus Hugo” – the name of this proposal is also its goal: recognizing the many opinions within fandom as to what nominees might be worthy of the Hugo. It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members: They still list the nominees, unranked, they feel are Hugo-worthy (up to the maximum permitted) in any categories they choose, just as they have in the past. In the past, we have counted the number of nominations each nominee received, and the top five nominees were put on the final ballot. However, because SF fandom typically nominates a variety of different nominees, it was easy for an organized slate to make it so that no other nominees made the final ballot.

Using this system, fandom isn't penalized for nominating a wide variety of nominees. If you nominate something that ends up not having a chance to make the final ballot, then your remaining choices automatically get more of your support instead of just being wasted. In other words, you can safely nominate anything you feel is Hugo-worthy. If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot. In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which finalists should be voted on to be the Hugo winner.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no group of members -- of any sort, minority or majority -- can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Award.

#587 ::: Keith "Kilo" Watt ::: (view all by) ::: June 09, 2015, 07:24 PM:

FAQ’s:

1. Can you explain the system in plain language?
First, and most importantly, E Pluribus Hugo makes absolutely no changes in how members nominate (no ranking or complex strategies are required). The new system changes only the way nominations are tallied in order to create a more robust measure of which nominees have the broadest and deepest support among WSFS members. The final Hugo voting system, which actually chooses the winner, is completely unchanged.

The new system narrows down Worldcon members' nominations by knocking out the least supported works in repeated elimination rounds until only five (under current rules) finalists remain. Here are the basic steps, as applied to one category:

a. All the nominations for all nominees (that is, the number of ballots on which each nominee appears) are added up, just like in the current system. This number, called "total nominations" in the proposal, is one of the criteria used to determine the final nominees with the broadest support.

b. Next, one point is divided equally among all the nominees in a category on each member's ballot. If there are two nominees on a ballot in a category, each gets 1/2 point; if there are three nominees, each gets 1/3 point, and so on.

c. All the points for each nominee from all nomination ballots are then added together. This number, called "point total" in the proposal, is the second of the two criteria used to determine the final nominees with the broadest support.

d. The nominees with the lowest point totals are compared to each other. The nominee with the fewest number of nominations is eliminated, because it is the nominee with the least support.

e. After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed among their remaining nominees. So, if there were four nominees getting 1/4 point each, there are now three nominees getting 1/3 point each, and so on. All points are re-added for all nominees. As less-supported nominees are eliminated, the total points for remaining nominees increase.

This process is repeated until the designated number of finalists (five in the current version of the constitution) remains.


2. E Pluribus Hugo?
The designers felt this name accurately reflects what the Hugo nomination system should be trying to achieve. Fandom is the Many. Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in science fiction and fantasy. This is a defining characteristic of fandom, and this system embraces it. Out of all of these many tastes in nominees, this system allows a single set of nominations to slowly emerge as the most popular candidates for the Hugo Award: “Out of the Many, a Hugo.”


3. But surely that’s not what the system is formally called?
Okay, sure. Formally, in the context of election theory, this system would be called “single divisible vote with least popular elimination” or SDV-LPE for short. E Pluribus Hugo is both cooler and geekier, as everything is better with a little Latin.


4. Why are there Selection and Elimination Phases? Couldn’t you simplify the system by doing it all in one step per elimination?
The Selection Phase looks for nominees with the lowest point total, and the Elimination Phase eliminates the one with the fewest nominations. Looking at the two different measures of popularity adds robustness to the system. With non-slate nominations, as occurred with most of the Hugo Awards to date, the nominees with the fewest points will tend to be the same ones as those with the fewest number of nominations. But while it might be possible for strategic blocs to manipulate one of these two measurements (as occurred in 2015), it would be very difficult to manipulate both in the same direction. A strategy that focuses on nominations for the Elimination Phase (such as bloc voting) will be stymied by the Selection Phase; a strategy that focuses on points for the Selection Phase (such as nominating only one nominee) will turn out to be counterproductive in the Elimination Phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.


5. Why re-run the system if someone declines a nomination? Why not just take the nominee that was eliminated in the next to last round?
We can do that, and in fact have an amendment ready if that’s what the people decide they want to do. Our system has only one major difference between it and taking the “6th place” nominee. Under our proposal, if a slate nominee makes the final ballot and that nomination is declined, that nomination will tend to be replaced by another nominee from the slate. In taking the “6th place” nominee, the declined nomination will tend to be replaced by a non-slate nominee. Either way works, but since our goal is to not prevent any section of fandom from being represented on the final ballot, the method of handling declined nominations outlined in the proposal seemed more in line with that philosophy.


6. Shouldn’t I just nominate one work if I want to give it the best chance to win?
The system was crafted specifically to make most strategies irrelevant. For example, suppose you wanted to give your full support behind one nominee that you really feel strongly about. You can do that, particularly if there isn’t anything else you feel is Hugo-worthy. On the other hand, if you nominate four other things that you think might be Hugo-worthy, you aren't hurting your favorite. This is because if your other nominees get eliminated, then your full support will go towards the remaining nominee, just as if you had never nominated the others. If enough people agree with your other four, some of them might make it, too. If enough people don't agree with your favorite, there's nothing you can do to get it on the final ballot, because you will never have more than one point. In general, the best strategy is simple: nominate as many nominees as you feel are worthy.


7. Surely strategy can make some difference, can't it?
It's impossible to create a voting system in which strategy never makes a difference. But in this system, there's no good way to know ahead of time that a strategy is safe and effective. Here's a specific example:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. Suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all could submit nomination ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its nominations compared in the Elimination Phase (you have no way to increase the number of nominations that Castle receives if it is selected for the Elimination Phase). If you push it up to 5th place, it can still be eliminated by Tentacles if Castle has fewer nominations. If you could somehow push it all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters so precisely before the election. Maybe Rocket isn’t so popular after all and needs your nominations to get over the top. Maybe lots of people will assume Rocket is safe and cause it to end up losing as a result of leaving it off their ballots. Maybe Tentacles has a broad enough base of support that your nominations for Castle aren’t enough to eliminate it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire, and only one, highly-specific way for it to work.

Nominating what you think is Hugo-worthy really is your best strategy.


8. Isn't it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws. The objective is to choose a system whose flaws are not in an area of concern to the electorate.


9. What are E Pluribus Hugo’s flaws?
In rare cases, it is possible that eliminating both members of a tie could change the final ballot slightly from what it would be if the tie were broken. The change is usually in the least popular of the finalists, and requires the two nominees in question to be very close in popularity. Hundreds of simulations were run using real and created data sets to help decide which tie-breaking methods would maintain our goal of supporting the wide range of opinions within fandom. It turned out that all the simulations showed that there was almost no difference in outcomes, no matter how we broke ties. We have chosen, therefore, to break ties in a manner consistent with section 6.4 of the Worldcon constitution. There are a number of additional tie breakers that could be used if it were deemed necessary in the future; however, simulations that have been run by the designers of this system show that it really shouldn’t be required.


10. What are E Pluribus Hugo’s benefits?
Simply put, it reduces the power of bloc voting without eliminating the chance that nominees that do appear on slates will make it to the final ballot. Conversely, it makes it very difficult for slates to prevent non-slate nominees from appearing on the ballot.


11. How does this system eliminate slate or bloc voting?
It doesn't, exactly, nor should a nominee be automatically eliminated just because it appears on a slate. On the other hand, any slate which nominates a full set of five nominees will find that each of its nominations only count 1/5 as much. With “non-slate” nominating, some of your nominees will be slowly eliminated, so your remaining nominees get more and more of your support. Since slate nominees tend to live or die together, they tend to eliminate each other until, in general, only one slate nominee remains. With a large enough support behind the slate (five times as much), the slate may still sweep a category; however, if that many voters support the slate, they arguably deserve to win, and no fair and unbiased system of nomination will prevent that. The answer in that case is, simply, to increase the general pool of voters. Regardless, with E Pluribus Hugo, slates will never receive a disproportionate share of the final ballot, as occurred in the 2015 Hugos.


12. Couldn't supporters of slates just recommend a single nominee for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate has enough supporters that is certainly a viable possibility – it’s also completely fair. It does not force all other nominees off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn't a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a nominee deserves a Hugo nomination, then it should, in fact, be represented on the final ballot.


13. What if there are multiple slates (slate wars, “parties”, etc.)?
As with a single slate, the nominees on each slate will tend to eliminate each other until only the most popular one remains. The end result is that even multiple slates are unable to sweep the nominations.


14. What happens if a broadly popular nominee is nominated by a group of unrelated people?
If it is broadly popular, the system will still select that nominee for the final ballot.


15. What happens if a broadly popular nominee also appears on a slate?
Even if it is on a slate, if the nominee garners support from individuals then the system will select that nominee for the final ballot. In general, slates neither help nor hurt any given nominee.


16. What happens when there are a lot of nominees with no obvious favorites and nomination slates are introduced?
Simulations of this scenario showed that slate nominees did receive a larger proportion of nomination slots than they did otherwise. However, this could be considered a fair and valid result: If there was no general favorite, then members really had no collective preference. Even in this scenario, simulations showed that non-slate nominees were not completely shut out of the final ballot.


17. How do the results of this system compare to the results under the current nomination system?
Statistical tests showed that this system and the current system overlap in about (conservatively) 4.5 out of 5 candidates. In the absence of slates, it generally gives identical results to the current nomination system.


18. I think we should just increase the number of nomination slots on the final ballot to a larger number (for example, 6), and decrease the number of slots a member can nominate to a smaller number (for example, 4). Wouldn't that be simpler and easier?
Unfortunately, this simply means that the largest slate will receive four of the nominations and the next largest will receive the remaining two. It doesn't solve the problem of slates depriving non-slate nominees of the opportunity to compete in the final award vote. In general, we want fandom to nominate as many nominees as they feel are Hugo-worthy, since under E Pluribus Hugo there is no strategic reason not to do so. For this reason, most of the designers of this system would prefer that members not be limited to nominating only four nominees. Keep in mind, however, that E Pluribus Hugo will work with this (or most any other) change as well, so one does not preclude the other.


19. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
This could work. The problem is that then you would have a small group of people serving as literal gatekeepers to the Hugo Awards. In spite of the word on the Internet, this has never been the case in the past. Establishing it would prove those groups who believed it existed in the past would then be correct. Ultimately, human judgement is fallible. The fairness of a committee’s decisions would forever be subject to opinion, and the prestige of the Hugo Awards would forever be tarnished.


20. I think we should use [insert other mathematical voting system].
We considered essentially every applicable type of voting system currently in the literature, guided by two experts in the field. It should be kept in mind, however, that the goals and requirements for choosing a set of representatives in a political situation are different from those for choosing a set of Hugo finalists. Some of these systems do, in fact, have positive properties that speak for them. None of them were as simple or as intuitive as E Pluribus Hugo, yet E Pluribus Hugo meets all of the stated goals for a Hugo nomination system.


21. Won’t E Pluribus Hugo be complicated to code and implement?
One of our non-experts coded a full simulator for the system in a matter of days. A full web-based app would not be much more difficult to handle.


22. Isnt't this change just designed to keep certain people out of the Hugo nomination process?
No. E Pluribus Hugo is designed to fix the flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the award is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.

Although it is true that the discussions for this system were hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resisted bloc or slate nominating, treated all individual nominators equally, and allowed a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.

#588 ::: Cally Soukup ::: (view all by) ::: June 09, 2015, 07:27 PM:

Typo in 22: "Isnt't"

#589 ::: David Dyer-Bennet ::: (view all by) ::: June 09, 2015, 09:34 PM:

Keith@586:

3.A.2 "...then instead no nominees will be eliminated this round..." I feel like it sounds better as "that" round instead. I can't make any headway into technical arguments for this change. Unless there's a technical reason it's right or there's a strong consensus, don't worry about it.

I just noticed that the tie-breaking rules are specific to only one phase each. We could actually eliminate 3.A.3 entirely, and put (1) and (2) directly into 3.A.1(2) (replacing the parenthetical), and ditto for (3) and (4) into 3.A.1(3).

That gives a Selection Phase like this:

(2) Selection Phase: The two nominees with the lowest point totals shall be selected for comparison in the Elimination Phase. If two or more nominations are tied for the lowest point total, all such nominees shall be selected for the Elimination Phase. If one nominee has the lowest point total and two or more nominees are tied for the second-lowest point total, then all such nominees shall be selected for the Elimination Phase.

(And correspondingly for the Elimination phase).

3.9.1: The first sentence specifies that notification should be made to suitable successors in the cased that the original creator of the work is unavailable. In the second and third sentences, it then says just "finalist", but I believe what it really means is "whoever the first sentence told you to notify". The use of different phrases in the sentences can easily be argued to mean that they intend to refer to different people, so I think it's important to somehow unify it. (The whole thing is complicated by the fact that finalists in many categories are works not people, so the instruction to notify a "finalist" is incoherent to begin with; that's existing 3.9 text and beyond the scope of our proposal to fix though.)

So, how about:

3.9.1: Worldcon Committees shall use reasonable efforts to notify the finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each person notified shall be asked at that time to either accept or decline the nomination. If the person notified declines the nomination, that finalist(s) shall not appear on the final ballot. The procedure for replacement of such finalist(s) is defined in subsection 3.A.4.


#590 ::: David Dyer-Bennet ::: (view all by) ::: June 09, 2015, 10:01 PM:

Keith@587: FAQ

1.e. Instead of say "All points are re-added for all nominees", I think it's be better to use the same term used in the actual proposal and say "All points are re-calculated for all nominees."

4. is a bit repetitive and a bit political. How about condensing it down to:

4. Why are there Selection and Elimination Phases? Couldn’t you simplify the system by doing it all in one step per elimination?
The Selection Phase looks for nominees with the lowest point total, and the Elimination Phase eliminates the one with the fewest nominations. Looking at the two different measures of popularity adds robustness to the system. There are strategies to manipulate one of these two measurements, but it is very difficult to manipulate both in the same direction. A strategy that focuses on nominations for the Elimination Phase (such as bloc voting) will be stymied by the Selection Phase; a strategy that focuses on points for the Selection Phase (such as nominating only one nominee) will turn out to be counterproductive in the Elimination Phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.

5. Remove reference to having an amendment handy? We still have it handy, and can pull it out if necessary. (Um, do we have it handy? We haven't reviewed that text that I recall.) As written, 5 kind of sounds like a call to anti-puppy activists (of which surely some will be at the Business Meeting) to change how the withdrawn work is replaced. But if we think what's in the proposal is best, then we shouldn't be promoting an alternative.

Can 8. be collapsed into 7? The first sentence of 7 could just be replaced by 8, perhaps with minor tweaks.

12. Not "be represented on the final ballot", but simply "appear on the final ballot". A nominee doesn't "represent" itself!

17. That first sentence is confusing. In fact it's flat-out wrong as I understand it. What I think the tests showed is that our system overlaps historical results in around 4.5 out of 5 cases.

18. "largest slate" and "next largest" -- the "slate" is the list of works. The number of works on the list is not what's at issue in this question. What's meant is "most popular slate" and "next most popular".

Wait, didn't we discuss this already? Did I lose and not remember it? If so, please ignore this; I don't intend to resurrect a settled discussion, but I thought we decided to fix this.

19. "In spite of the word on the Internet" How about saying instead "In spite of some claims on the Internet"? "The word on the Internet" sounds like it's the established opinion, and it isn't; it's merely being argued by one faction.

#591 ::: Soon Lee ::: (view all by) ::: June 09, 2015, 10:31 PM:

[FAQ]

For Q6 the sentence is a fragment, and for Q22 I much preferred it when we used the present tense; it's more active & more in line with the rest of the FAQ.

[FAQ6] For example, suppose you wanted to give your full support behind one nominee that you really feel strongly about. You can do that, particularly if there isn’t anything else you feel is Hugo-worthy.

-> For example, if you wanted to support a single nominee you feel strongly about, and there isn't anything else you think is Hugo-worthy, you can.

[FAQ22]...our goals were to create a system that resisted bloc or slate nominating, treated all individual nominators equally, and allowed a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.

-> ...our goals were to create a system that resists bloc or slate nominating, treats all individual nominators equally, and allows a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.

#592 ::: Keith "Kilo" Watt ::: (view all by) ::: June 09, 2015, 10:51 PM:

David@589/590:

Most of what you suggest sounds fine; just a couple of quick comments...

We actually originally had the tie-breakers in with the phases, but after much discussion moved them into a separate section. I'd rather not backtrack on that one at this point.

The FAQ suggestions look fine to me overall. For number 5, I think we do need to have the amendment ready, but we had agreed to work on that after finishing the main proposal. Whether to mention it or not is a valid question, but there was enough debate on the issue that most felt the alternative should be mentioned. I think it's a given that someone at the business meeting is going to suggest it, so we may as well go ahead and address it. I do think it's worth pointing out what the functional difference between the two options is, though.

I'm unsure about combining 7 and 8, but I think what you suggest can work. Let me play with that a bit, but I think we can do it.

Thanks for the feedback,
Kilo

#593 ::: Keith "Kilo" Watt ::: (view all by) ::: June 09, 2015, 10:54 PM:

David@599:

Oh, I forgot to mention that for 3.9, we were asked by the admins to use the term "finalist", since that's what the rest of the constitution is going to be switching to. I personally don't have a problem with how you've phrased it, but given the request that was made of us, I think we should leave that as-is.

K

#594 ::: David Harmon ::: (view all by) ::: June 09, 2015, 11:08 PM:

David Dyer-Bennet %589, #590: Not an argument against any particular of your changes, but I'd like to note that we don't necessarily want to compress the FAQ into its smallest representation. ;-) Seriously, for explanatory writing some redundancy is probably needed.

More specific cases:
The tie-breaking stuff collects several special cases, sparing the digressions from the middle of defining the main mechanics.

5: The phrasing could be smoother, but we do want to note (a) that as far as we're concerned, the difference is negotiable, and fits within our concept of the proposal, and (b) why we picked the option we did.

My suggestion:

We considered that as an option, including analyzing and simulating the effects on the vote either way. We discovered only one major difference in the results: When taking the “6th place” nominee, the declined nomination will tend to be replaced by a non-slate nominee. With our system of re-running the vote, if a slate nominee makes the final ballot and its nomination is declined, that nomination will tend to be replaced by another nominee from the slate. Either way works, but as noted above, our goal is to avoid preventing any section of fandom from being represented on the final ballot. The results of re-running the calculation entirely, seem more in line with that philosophy.

Combining 8/7: Agree, and arranged as it is (with the tiny 8 after the lengthy 7, it looks odd. My suggestion:

7. Isn't it true that any voting system can be gamed (or strategized, etc.)? How can this system be gamed?
Yes, there is a theorem which proves that all voting systems must have inherent flaws -- that is, it's impossible to create a voting system in which strategy never makes a difference. The objective therefore is to choose a system whose flaws are not an issue to the election at hand. In this system, "strategic voting" is technically possible, but risky: There's no good way to know ahead of time whether a given strategy will help, or backfire. Here's a specific example:

then continue with the rest of 7..

12: Heh. Good spot! I suspect "represented" popped up in the service of "sounding legalistic and offical".

18: Agreed.

19: Also, "the word on the Internet" is pretty chatty.
----
Another point:
16: "However, this could be considered a fair and valid result" I suggest deweaseling this to "We consider this a fair and valid result".

#595 ::: David Dyer-Bennet ::: (view all by) ::: June 09, 2015, 11:37 PM:

Keith@593: Huh? I'm not going back to nominee there, which is what we are avoiding; that's not the issue or the point of my @589 3.9.1 changes.

Not sure I'm understanding what you are objecting to, though, so I'll leave it at that for now; can you clarify?

#596 ::: David Dyer-Bennet ::: (view all by) ::: June 09, 2015, 11:41 PM:

Soon Lee@591:

Agree about using present tense.

And now that I'm looking at it (both original and your changes) I'm bothered by "give your whole support behind", which is embedded in both versions. Could be just "give your whole support to", for example.

(One puts ones support behind something, or gives ones support to something, as I understand it.)

#597 ::: Keith "Kilo" Watt ::: (view all by) ::: June 09, 2015, 11:44 PM:

DavidDB@595:

It's not just that we were supposed to avoid "nominee"; they want us to actually use "finalist".

K

#598 ::: David Dyer-Bennet ::: (view all by) ::: June 09, 2015, 11:46 PM:

Keith@592: Well, this version of the tallying rules is both fewer sections and fewer words, arguably simpler, and that's something we're pursuing. When we pulled the tie rules out I thought they were being shared between multiple uses; but they're not now. However, I won't pursue that further in the absence of significant encouragement.

Is somebody keeping a worklist somewhere? Because I'd rather not end up going with an amendment that hasn't been looked at carefully when the time comes.

#599 ::: Keith "Kilo" Watt ::: (view all by) ::: June 09, 2015, 11:50 PM:

DavidDB@598:

Is somebody keeping a worklist somewhere?

Oh yes. You should see all my notes... :)

Kilo

#600 ::: David Dyer-Bennet ::: (view all by) ::: June 09, 2015, 11:52 PM:

Keith@597: I do not believe that my changes there violate that request at all. The coming change is "nominee" changing to "finalist", and I strictly kept to that.

The problem is, it's vital to make it clear that the next two sentences refer to the finalist (person) or his guardian, heirs, or assigns, as specified in the first sentence.

Using one formulation in the first sentence and a different one in the second two is a clear sign that they mean different things. We either need to repeat the language used in the first sentence, or refer back to it. Given its length, I chose to refer back to it (a special case of different language that does mean the same thing).

#601 ::: David Dyer-Bennet ::: (view all by) ::: June 09, 2015, 11:55 PM:

David Harmon@594: I do agree that ultimate compression is certainly not the goal, especially with the FAQ. But I haven't gone anywhere near ultimate compression; trust me on this :-) .

Your 7 suggestion is the same as mine.

Strongly agree with you on 16.

#602 ::: David Harmon ::: (view all by) ::: June 09, 2015, 11:56 PM:

Also, FAQ22: Suggest changing "fix the flaw" to "fix a flaw". It's not necessarily the only one.

#603 ::: Keith "Kilo" Watt ::: (view all by) ::: June 10, 2015, 12:15 AM:

DavidDB@600:

Okay, I think maybe your last parenthetical comment about works vs. people was throwing me off. They don't want us to make a distinction between works and people, they want us to just say finalist (or nominee before they are finalists). But I think maybe that's not what you're talking about. I was struggling to see how finalist could refer to anyone else, but is it this part that you think is causing the trouble:

"or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians"

Is your issue, then, that "finalist" doesn't refer to these people? I think it's clear that it does, but I also don't object to the change if this is the issue. It's not clear from your original post what you're referring to, though, so I'm not sure.

Kilo

#604 ::: David Dyer-Bennet ::: (view all by) ::: June 10, 2015, 12:38 AM:

keith@603: My issue is that using one phrase in the first sentence, and a different phrase in the second two, can very reasonably be read as making a careful distinction. Since that's not what we mean, I want to change the second two to refer back to the first one (the same effect could be achieved by just duplicating the phrase used there, but it's very long, so I took the other approach) to make it clear that they're referring to the same thing.

(Especially given that in this particular sentence, it refers to "finalist, or <special cases>"; if you later refer to just "finalist" you're clearly excluding the special cases.)

My suggested wording, back when, takes care of that I think. I'm not trying to fix the nominee vs. person problem, that's out of scope.

#605 ::: Keith "Kilo" Watt ::: (view all by) ::: June 10, 2015, 11:16 AM:

All:

All discussions to date have been incorporated in the proposal. I'll post another version late this afternoon (UT-7/MST), so please continue to look over the current version and post your comments. Since we're getting close, I will try to post the latest version each day.

Thanks!
Kilo

#606 ::: Brad from Sunnyvale ::: (view all by) ::: June 10, 2015, 12:27 PM:

As to the question of why people love to promote their "discoveries" rather than what is popular, it does not matter so much as to why -- what matters is if the voting system gives them an incentive to not speak the truth on their ballot, and how much incentive it gives.

With the rules themselves not up for debate in this thread, however, let me state that I think the explanation portion is in accurate when it says there is no reason not to vote for a broad set of candidates. There is a reason to leave off candidates you consider shoo-ins from your ballot, it increases the support you give the other candidates. It does so much more so than the old basic approval did, in theory up to five times more at the maximum, though fortunately rarely that.

Whether you judge it is likely that many fans would exploit this or not, it is still true. As such the language should not state otherwise. The FAQ goes into more detail, which is good, but I find some fault for two reasons. One is that I believe it is easier to predict in advance, thanks to the internet and other sources that are readily available such as the Locus poll and online polls. Secondly, the presence of slates makes this worse, as voters might be inclined (in all systems, but even more in this one) to not nominate works they have learned are on some slate, but to reduce support for them when it is seen that they are getting unfair support, and because if the slate is going to get them on, you can now put your support elsewhere.

The second reason is more meta, but still valid -- all that matters is that people BELIEVE the strategy might work. It is not necessary that it actually be that successful. The fact that this system allows, in theory, you to offer 5 times as much support in the count for something as one can in the old system will lead people to optimism.

A simple case -- say you are the author or publisher or family member or close associate of a work (and no other work.) Your strong incentive is to nominate only your work, even though that might be unsporting. Why? Because imagine you miss the ballot by 0.5 points because you did otherwise. It's been known to happen even in the old system, and here there was something you could have done about it. The fear of this, unlikely as it might be, will push you to both not express your true opinion and to be unsporting when you otherwise would have been.


I should also admit my bias here -- I don't believe this system will work, and thus it is not my choice. I think it will still allow 2 slate candidates, displacing genuine ones, and I believe the slate attackers will search for and find other strategies to control the system. However, I do want this proposal to be accurately described if it is to be put forward.

#607 ::: David Dyer-Bennet ::: (view all by) ::: June 10, 2015, 01:51 PM:

Brad@606: if you have a couple of hundred slate voters, any basically democratic system will allow them to get a nominee they agree on onto the final ballot. The total number of nominators, plus the large number of works published each year, make this unavoidable. Whether slate voters or not, a few hundred people who agree on one work will get that work onto the ballot.

There are proposals to try to change that; but, basically, such proposals are anti-democratic. When a couple of hundred people think a work belongs on the final ballot, then it DOES; I don't want that fixed. (Though I'd like there to be social consequences for trying to promote a slate, which I think is damaging to the awards.)

But you're just mathematically wrong about the other things. Not nominating your secondary choices will DECREASE the chances of things you support making the final ballot. While a simplistic analysis makes it look like you might do better just nominating your strong first choice, that's almost never true -- and you can't tell, at the time you have to decide what to nominate, whether this is one of the times when it's true or not.

Various people have run many hundreds of runs, with anonymized real data, slightly modified real data, and completely synthetic data sets, trying to find cases that cause trouble. They're essentially nonexistent in large data sets (easy enough to create with 6 ballots, but that's not what we deal with in the real world), and they cannot be recognized with the information available to a voter at the nominating deadline.

#608 ::: Annie Y ::: (view all by) ::: June 10, 2015, 02:45 PM:

Brad from Sunnyvale @ 606

But we are not even trying to stop the slate from placing a proportional number of nominees on the ballot. The problem is when they replace all other nominees as it happened this year.

Genuine according to whom? According to the slate voters, theirs are the genuine ones and we acknowledge that. So they will get their positions proportionately on the number of voters they have (or close enough). If 1000 people nominate and 300 of them vote with a slate, they are still 30% of the nominators. Trying to exclude them just because they went for a slate is exactly as bad as what happened this year when everyone else was locked out from the ballot. A system that attempts to exclude nominators is a a bad idea - not in the least because defining a slate is impossible (if 100 people nominate only Ancillary Mercy next year, is this a slate or are these people just liking only one novel?). Let's not go down that rabbit hole again.

As for the not adding a second work - how does this differ from the current system when adding a second book on your ballot may be the vote that allows that second work to kick your more prefer work off the ballot? If you want to vote strategically, the things do not change much between the current system and the proposed one. In both cases you cannot judge what your actions will lead to - and adding an additional work may end up harming the ones you like more... Does not stop people from nominating now; shouldn't do it in any system. You cannot be effectively strategical if you do not have all the data - and you cannot have it unless if you know what everyone else is voting.

#609 ::: Jameson Quinn ::: (view all by) ::: June 10, 2015, 03:49 PM:

@606: I think your main mistake is here:

it increases the support you give the other candidates. It does so much more so than the old basic approval did, in theory up to five times more at the maximum

For non-slate candidates, nominations are more important than points. Half a point more, or even (implausibly, but possibly) 0.8 of a point more, does not do anything at all in normal cases to protect a work from elimination based on nominations. Thus:

Because imagine you miss the ballot by 0.5 points because you did otherwise. It's been known to happen even in the old system, and here there was something you could have done about it.

If your work happens to be in 6th place, the only thing that can happen if you give it an extra half point is that it moves up to 5th place; runs against the true 5th-place work that you helped it "displace"; and is eliminated anyway based on nominations, exactly as it would have been if it had stayed in 6th place. In order to actually make a difference, you'd have to get a large number of your friends to also be strategic, so that it would go up to 4th place.

(Furthermore, though this is clearly a secondary point, the fact we're talking about half points here does in fact make this kind of strategy half as likely to move you even just uselessly from 6th to 5th place as it would have been in the old system. And yet another thing: your scenarios seem to be suggesting that the strategy is semi-honest, in which case it is actually good for the expected average quality of the finalists unless there are slates. So the only way this could possibly make things worse is if it totally fails to discourage slate voters, AND some voters are irrationally over-strategic, BUT only one subset of voters, AND a bunch of other things are in perfect balance SO that a slightly inferior, lesser known work ends up beating a slightly superior, better known one. Which... you know, it seems that Terry Pratchett thought that kind of result would actually be good, and you wouldn't disagree with a dead guy, WOULD YOU?)

#610 ::: Chris Battey ::: (view all by) ::: June 10, 2015, 05:41 PM:

Brad @ 606 et al:

We acknowledge that with otherwise typical nominating patterns, a block of 20% of the nominators could take two out of five slots. That was always going to happen, in part because there will always be nominators whose choices are all eliminated early due to a lack of popularity. That's still far better than what we get out of today's system.

And the scenarios you mention about your favorite nominee barely missing out because you also nominated a more popular work also happens in today's system. If I have my favorite nominee A and my also-liked-it, more popular nominee B on my ballot, and overall B beat A for fifth place by 100 nominations to 99, then leaving B off your nominating ballot would have allowed A to tie B for fifth with 99, and also get onto the ballot. But if C, which you hate, was sitting in fifth place at 100 votes as well, then all leaving B off your ballot did was deny it a place on the final ballot; A was never making it on anyway.

No system is immune to strategy, but we can at least try to limit the effectiveness of those strategies. We don't consider this kind of single-ballot strategizing to be a significant weakness of the system, because it requires far more exact knowledge about the rest of the ballots than you're going to have, and its effectiveness is limited to a very small, specific case (while there are other cases where it can backfire). In this case, blind balloting plus a large nominating population is pretty much all we need to render this strategy unlikely to have the desired impact.

[SHIRTS/RIBBONS]

I haven't forgotten about this topic, but decided it would be better to table it until after the proposal language was locked. We'll probably want a future thread focused on the ancillary logistics of the proposal, including shirts/ribbons/buttons, coordination at con (both for the BM itself and also other meetups as desired), promotion activities (reaching out to other fan commmunities online), etc.

#611 ::: Brad from Sunnyvale ::: (view all by) ::: June 10, 2015, 07:01 PM:

David @607 -- Since this thread is to discuss the proposal and not others, I will have to point you to other threads where I have advanced that there are systems which fix slate voting completely, and they are not only democratic, but they are in fact the primary means used by almost all democratic societies and systems to solve problems like this! This algorithmic approach is not un-democratic, but it is atypical.

However, if my math is wrong, you need to point me to the actual counter-math. I do have a degree in mathematics. I understand what the simulations show and it's good that they were done, but as simulations they do not factor in dynamic human minds attempting to either find strategies to increase their influence (or in the case of slates, to corrupt the system entirely.)

Point out the flaw with this math: If I have candidates A and B, and I correctly consider A to be a shoo-in, then the ballot AB allocates 0.5 points to both A and B until such time as B is eliminated. However, the ballot "B" allocates 1 point to B at all times unless it is eliminated, and so if there is a round where 0.5 points or less makes the difference on being eliminated, the B ballot is better for my choice than the AB ballot. If 10 fans of B follow this approach (not by collusion, but out of fandom) then in fact their choice helps B a lot in any close situation. Of course, any and all ballot strategies (including the non-strategy of an honest list) affect the result only in close situations -- to resolve the rare potential close situation is the only reason it ever matters to cast a ballot. The B ballot makes twice as much difference in that situation than the AB does.

Of course, going B instead of AB hurts A's chances, but since this hypothetical voter believes there is not going to be a close situation for A, they don't believe their strategy will hurt it. They are probably right.


However, I welcome an explanation of how that math is wrong.

Annie #608 -- I do disagree with this. I believe we want to measure true fan opinion, and so if people colluding can multiply their influence to place something on the ballot that was not the result you would get from true fan opinion, I think that is a flaw and I would fix it if I could, and I believe we can.

Times must have changed. I remember people getting deeply upset at the suggestion that scientologists colluded to get one LRH work on the ballot, so I am pretty sure that, at least in the past, many people felt the same way and didn't think that "let attackers get 1 or 2 victories as long as they don't get all 5" was acceptable.

#609 -- Let's consider a ballot where I can choose between "AD" and "D". Say I choose AD

A (1st) 150 points
...
B (4th) 105.5 points
C (5th) 100 points - N nominations
D (6th) 99.8 points - N nominations
E (7th) 99.8 points - N nominations

In this case, both D and E go into the elimination round, and both go, leaving the final ballot. However, in the situation where I vote just for D, it now has 100.3 points, and it is C and E that will go into the elimination round, and both go, leaving the final ballot -- but this time with D on it.

It's a close situation -- but again, all voting is only about making the difference in the close situation.

The difference here from plain approval is the strategy there is much weaker, and involves not listing C or E on your ballot because you will compete with them. Here you are encouraged both to not list C or E, but also A.

Now, as I said, I think this system is superior to the ones we discussed in the thread last month and it reduces this strategy, so as algorithmic approaches, it is indeed one of the better ones. It is false to say it does not increase the utility of strategic voting, however.

And alas, it I don't believe it attains the goal I want (no getting *any* works on the ballot through what I consider cheating) and it is unable to respond to a dynamic adversary. When you have a truly adversarial situation, you must have an OODA loop, not a fixed algorithm. At least that's what the people who battle adversaries for a living all say.

#612 ::: Annie Y ::: (view all by) ::: June 10, 2015, 07:22 PM:

Brad,

And how are you defining a slate? Let's be realistic instead of idealistic. There is no way to punish collusion because there is no real way to prove it short of starting a witch hunt every time when things look dicey. When slates runners realize that their slated works cannot win (people will send them under no award for the most part), they will give up. But as much as I do not like slates, you cannot exclude people just because they voted a slate; this is the same as excluding anyone that is not voting the slate.

I am not sure that I want to ask you how you think to fix it in this thread -- we are trying to finalize wording on a proposal here. But if you have an idea, go ahead and propose a change based on that.

I guess we will need to agree to disagree for now. :)

#613 ::: David Dyer-Bennet ::: (view all by) ::: June 10, 2015, 07:24 PM:

Brad@611: Yes, you can construct cases. No, you cannot have enough information to know what case you're actually functioning in in a real nomination round. Attempting to do so is nearly certain to backfire; to represent your interests less well than making simple, straightforward, nominations.

My degree is in math too, and more to the point several of the participants here have done a lot of work with voting systems (I've only done some amateur study of them over the last few decades).

And electoral systems have different goals as well as different threat scenarios from our award nominating situation.

Yes, abandoning democratic nominations would be a more reliable way of getting only the right works nominated -- so long as people who agree with you about what the right works are are in charge of the non-democratic nomination mechanism. Which is why we prefer algorithmic methods -- we want a democratic system that selects a range of nominees reasonably representative of the preferences of the nominators (which then feeds in to a voting system that pretty much maximizes happiness among the voters).

There is absolutely zero possibility that you could get the WSFS Business Meeting to approve a system that put a review panel into the Hugo process. Everybody is too aware of what a huge change that is, and how capturing that panel becomes the goal of all people willing to cheat.

#614 ::: Jameson Quinn ::: (view all by) ::: June 10, 2015, 08:04 PM:

@611: If the only concrete case you can make for strategy involves a simultaneous 3-way tie or near-tie in two different metrics (one of which is essentially continuous), and you're trying to show something significant with that, then you may have a degree in math, but I think you brush up on your measure theory.

#615 ::: Jameson Quinn ::: (view all by) ::: June 10, 2015, 08:15 PM:

...The probability of such an event is o(n⁻²), that is, it is asymptotically dominated by a constant divided by the number of voters squared. For 1000 voters, that is essentially zero.

#616 ::: Jameson Quinn ::: (view all by) ::: June 10, 2015, 08:24 PM:

...the upper-bound constant does involve the square of the number of viable candidates, but since this is reliably less than 30 for any reasonable definition of "viable", the probability for n=1000 is still less than 0.1%.

#617 ::: Jameson Quinn ::: (view all by) ::: June 10, 2015, 08:27 PM:

...How did I get n²? It's one factor of sqrt(n) for each 2-way tie, and another factor of the same for adding a third participant in each tie. That's the fourth power of the square root, which is the square.

OK, that's enough of that.

#618 ::: P J Evans ::: (view all by) ::: June 10, 2015, 08:54 PM:

611
I would oppose something like that. It's not what we want to do, it's so very much not where we want to go.

#619 ::: Jameson Quinn ::: (view all by) ::: June 10, 2015, 09:02 PM:

@611:

It is false to say it does not increase the utility of strategic voting, however.

Compared to the current situation? Where strategic voters just managed to dominate most categories using a simple, robust strategy? I honestly can't imagine how you could even say that with a straight face, unless you don't understand that "strategy" includes slates.

#620 ::: Cally Soukup ::: (view all by) ::: June 10, 2015, 11:03 PM:

Brad, are you calling for literal gatekeepers? Because NO. No no no no no. Also, no.

#621 ::: Keith "Kilo" Watt ::: (view all by) ::: June 11, 2015, 12:00 AM:

All: Latest version!

Get those comments and typo-crushing in! :)

Kilo

#622 ::: Keith "Kilo" Watt ::: (view all by) ::: June 11, 2015, 12:01 AM:

Short Title: E Pluribus Hugo (Out of the Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations), section 3.9 (Notification and Acceptance), and section 3.11 (Tallying of Votes) as follows:

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the process described in section 3.A.

Insert new section 3.A after Section 3.8 as follows:
Section 3.A: Finalist Selection Process

3.A.1: For each category, the finalist selection process shall be conducted as elimination rounds consisting of three phases:

(1) Calculation Phase: First, the total number of nominations (the number of ballots on which each nominee appears) from all eligible ballots shall be tallied for each remaining nominee. Next, a single “point” shall be assigned to each nomination ballot. That point shall be divided equally among all remaining nominees on that ballot. Finally, all points from all nomination ballots shall be totaled for each nominee in that category. These two numbers, total nominations and total points, shall be used in the Selection and Elimination Phases.
(2) Selection Phase: The two nominees with the lowest point totals shall be selected for comparison in the Elimination Phase. (See 3.A.3 for ties.)
(3) Elimination Phase: Nominees chosen in the Selection Phase shall be compared, and the nominee with the fewest total nominations shall be eliminated and removed from all ballots for the Calculation Phase of all subsequent rounds. (See 3.A.3 for ties.)

3.A.2: The phases described in 3.A.1 are repeated in order for each category until the number of finalists specified in 3.8.1 remain. If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then instead no nominees will be eliminated this round, and all remaining nominees shall appear on the final ballot, extending it if necessary.

3.A.3: Ties shall be handled as described below:
(1) During the Selection Phase, if two or more nominations are tied for the lowest point total, all such nominees shall be selected for the Elimination Phase.
(2) During the Selection Phase, if one nominee has the lowest point total and two or more nominees are tied for the second-lowest point total, then all such nominees shall be selected for the Elimination Phase.
(3) During the Elimination Phase, if two or more nominees are tied for the fewest number of nominations, the nominee with the lowest point total at that round shall be eliminated.
(4) During the Elimination Phase, if two or more nominees are tied for both fewest number of nominations and lowest point total, then all such nominees tied at that round shall be eliminated.

3.A.4: After the initial Award ballot is generated, if any finalist(s) are removed for any reason, the finalist selection process shall be rerun as though the removed finalist(s) had never been nominee(s). None of the remaining original finalists who have been notified shall be removed as a result of this rerun. The new finalist(s) shall be merged with the original finalists, extending the final ballot if necessary.

Section 3.9: Notification and Acceptance.
3.9.1 Worldcon Committees shall use reasonable efforts to notify the nominees finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each nominee person notified shall be asked at that time to either accept or decline the nomination. If the nominee the person notified declines the nomination, that nominee finalist(s) shall not appear on the final ballot. The procedure for replacement of such finalist(s) is defined in subsection 3.A.4.

Section 3.11: Tallying of Votes.
3.11.4: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. During the same period, the results of the last ten rounds of the finalist selection process for each category (or all the rounds if there are fewer than ten) shall also be published.

Submitted by: Keith “Kilo” Watt, Jameson Quinn, Tammy Coxen, Yoana Yotova, Joshua Kronengold, Christopher Battey, David Gallaher, Adam Tilghman, Sara Watt, CJ Cabourne, Steven Halter, P J Evans, David Goldfarb, Seth Gordon, Ginger Tansey, Steve Wright, Catherine Faber, Andy Holloway, Duncan J. Macdonald, Claudia Beach, Derry Earnshaw, Jason Skiles, Soon Lee, David Harmon, Lydia Nickerson, Abigail Sutherland, Lee Billings, Oskari Rantala, Seth Breidbart, Chris Suslowicz, Fragano Ledgister, Lori Coulson, Jeffry Herman, Mark Shier, Buddha Buck, Lenore Jean Jones, David Langford, Christopher Hatton, Rogers Cadenhead

Commentary: “E Pluribus Hugo” – the name of this proposal is also its goal: recognizing the many opinions within fandom as to what nominees might be worthy of the Hugo. It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members: They still list the nominees, unranked, they feel are Hugo-worthy (up to the maximum permitted) in any categories they choose, just as they have in the past. In the past, we have counted the number of nominations each nominee received, and the top five nominees were put on the final ballot. However, because SF fandom typically nominates a variety of different nominees, it was easy for an organized slate to make it so that no other nominees made the final ballot.

Using this system, fandom isn't penalized for nominating a wide variety of nominees. If you nominate something that ends up not having a chance to make the final ballot, then your remaining choices automatically get more of your support instead of just being wasted. In other words, you can safely nominate anything you feel is Hugo-worthy. If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot. In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which finalists should be voted on to be the Hugo winner.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no group of members -- of any sort, minority or majority -- can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Award.

#623 ::: Keith "Kilo" Watt ::: (view all by) ::: June 11, 2015, 12:02 AM:

FAQ’s:

1. Can you explain the system in plain language?
First, and most importantly, E Pluribus Hugo makes absolutely no changes in how members nominate (no ranking or complex strategies are required). The new system changes only the way nominations are tallied in order to create a more robust measure of which nominees have the broadest and deepest support among WSFS members. The final Hugo voting system, which actually chooses the winner, is completely unchanged.

The new system narrows down Worldcon members' nominations by knocking out the least supported works in repeated elimination rounds until only five (under current rules) finalists remain. Here are the basic steps, as applied to one category:

a. All the nominations for all nominees (that is, the number of ballots on which each nominee appears) are added up, just like in the current system. This number, called "total nominations" in the proposal, is one of the criteria used to determine the final nominees with the broadest support.

b. Next, one point is divided equally among all the nominees in a category on each member's ballot. If there are two nominees on a ballot in a category, each gets 1/2 point; if there are three nominees, each gets 1/3 point, and so on.

c. All the points for each nominee from all nomination ballots are then added together. This number, called "point total" in the proposal, is the second of the two criteria used to determine the final nominees with the broadest support.

d. The nominees with the lowest point totals are compared to each other. The nominee with the fewest number of nominations is eliminated, because it is the nominee with the least support.

e. After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed among their remaining nominees. So, if there were four nominees getting 1/4 point each, there are now three nominees getting 1/3 point each, and so on. All points are re-calculated for all nominees. As less-supported nominees are eliminated, the total points for remaining nominees increase.

This process is repeated until the designated number of finalists (five in the current version of the constitution) remains.


2. E Pluribus Hugo?
The designers felt this name accurately reflects what the Hugo nomination system should be trying to achieve. Fandom is the Many. Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in science fiction and fantasy. This is a defining characteristic of fandom, and this system embraces it. Out of all of these many tastes in nominees, this system allows a single set of nominations to slowly emerge as the most popular candidates for the Hugo Award: “Out of the Many, a Hugo.”


3. But surely that’s not what the system is formally called?
Okay, sure. Formally, in the context of election theory, this system would be called “single divisible vote with least popular elimination” or SDV-LPE for short. E Pluribus Hugo is both cooler and geekier, as everything is better with a little Latin.


4. Why are there Selection and Elimination Phases? Couldn’t you simplify the system by doing it all in one step per elimination?
The Selection Phase looks for nominees with the lowest point total, and the Elimination Phase eliminates the one with the fewest nominations. Looking at the two different measures of popularity adds robustness to the system. There are strategies to manipulate each of these two measurements, but it is very difficult to manipulate both in the same direction. A strategy that focuses on nominations for the Elimination Phase (such as bloc voting) will be stymied by the Selection Phase; a strategy that focuses on points for the Selection Phase (such as nominating only one nominee) will turn out to be counterproductive in the Elimination Phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.


5. Why re-run the system if someone declines a nomination? Why not just take the nominee that was eliminated in the next to last round?
We considered that as an option, including analyzing and simulating the effects on the vote either way. We discovered only one major difference in the results: When taking the “6th place” nominee, the declined nomination will tend to be replaced by a non-slate nominee. With our system of re-running the vote, if a slate nominee makes the final ballot and its nomination is declined, that nomination will tend to be replaced by another nominee from the slate. Either way works, but as noted above, our goal is to avoid preventing any section of fandom from being represented on the final ballot. The results of re-running the calculation entirely, seem more in line with that philosophy.

6. Shouldn’t I just nominate one work if I want to give it the best chance to win?
The system was crafted specifically to make most strategies irrelevant. For example, if you wanted to support a single nominee you feel strongly about, and there isn't anything else you feel is Hugo-worthy, you can do that. On the other hand, if you nominate four other things that you feel might be Hugo-worthy, you aren't hurting your favorite. This is because if your other nominees get eliminated, then your full support will go to the remaining nominee, just as if you had never nominated the others. If enough people agree with your other four, some of them might make it, too. If enough people don't agree with your favorite, there's nothing you can do to get it on the final ballot, because you will never have more than one point. In general, the best strategy is simple: nominate as many nominees as you feel are worthy.


7. Isn't it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws -- that is, it's impossible to create a voting system in which strategy never makes a difference. The objective therefore is to choose a system whose flaws are not an issue to the election at hand. In this system, "strategic voting" is technically possible, but extremely risky: There's no good way to know ahead of time whether a given strategy will help, or backfire. Here's a specific example:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. Suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all could submit nomination ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its nominations compared in the Elimination Phase (you have no way to increase the number of nominations that Castle receives if it is selected for the Elimination Phase). If you push it up to 5th place, it can still be eliminated by Tentacles if Castle has fewer nominations. If you could somehow push it all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters so precisely before the election. Maybe Rocket isn’t so popular after all and needs your nominations to get over the top. Maybe lots of people will assume Rocket is safe and cause it to end up losing as a result of leaving it off their ballots. Maybe Tentacles has a broad enough base of support that your nominations for Castle aren’t enough to eliminate it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire, and only one, highly-specific way for it to work.

Nominating what you think is Hugo-worthy really is your best strategy.


8. Isn't it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws. The objective is to choose a system whose flaws are not in an area of concern to the electorate.


9. What are E Pluribus Hugo’s flaws?
In rare cases, it is possible that eliminating both members of a tie could change the final ballot slightly from what it would be if the tie were broken. The change is usually in the least popular of the finalists, and requires the two nominees in question to be very close in popularity. Hundreds of simulations were run using real and created data sets to help decide which tie-breaking methods would maintain our goal of supporting the wide range of opinions within fandom. It turned out that all the simulations showed that there was almost no difference in outcomes, no matter how we broke ties. We have chosen, therefore, to break ties in a manner consistent with section 6.4 of the Worldcon constitution. There are a number of additional tie breakers that could be used if it were deemed necessary in the future; however, simulations that have been run by the designers of this system show that it really shouldn’t be required.


10. What are E Pluribus Hugo’s benefits?
Simply put, it reduces the power of bloc voting without eliminating the chance that nominees that do appear on slates will make it to the final ballot. Conversely, it makes it very difficult for slates to prevent non-slate nominees from appearing on the ballot.


11. How does this system eliminate slate or bloc voting?
It doesn't, exactly, nor should a nominee be automatically eliminated just because it appears on a slate. On the other hand, any slate which nominates a full set of five nominees will find that each of its nominations only count 1/5 as much. With “non-slate” nominating, some of your nominees will be slowly eliminated, so your remaining nominees get more and more of your support. Since slate nominees tend to live or die together, they tend to eliminate each other until, in general, only one slate nominee remains. With a large enough support behind the slate (five times as much), the slate may still sweep a category; however, if that many voters support the slate, they arguably deserve to win, and no fair and unbiased system of nomination will prevent that. The answer in that case is, simply, to increase the general pool of voters. Regardless, with E Pluribus Hugo, slates will never receive a disproportionate share of the final ballot, as occurred in the 2015 Hugos.


12. Couldn't supporters of slates just recommend a single nominee for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate has enough supporters that is certainly a viable possibility – it’s also completely fair. It does not force all other nominees off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn't a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a nominee deserves a Hugo nomination, then it should, in fact, appear on the final ballot.


13. What if there are multiple slates (slate wars, “parties”, etc.)?
As with a single slate, the nominees on each slate will tend to eliminate each other until only the most popular one remains. The end result is that even multiple slates are unable to sweep the nominations.


14. What happens if a broadly popular nominee is nominated by a group of unrelated people?
If it is broadly popular, the system will still select that nominee for the final ballot.


15. What happens if a broadly popular nominee also appears on a slate?
Even if it is on a slate, if the nominee garners support from individuals then the system will select that nominee for the final ballot. In general, slates neither help nor hurt any given nominee.


16. What happens when there are a lot of nominees with no obvious favorites and nomination slates are introduced?
Simulations of this scenario showed that slate nominees did receive a larger proportion of nomination slots than they did otherwise. This is a fair and valid result: If there was no general favorite, then members really had no collective preference. Even in this scenario, simulations showed that non-slate nominees were not completely shut out of the final ballot.


17. How do the results of this system compare to the results under the current nomination system?
Statistical tests showed that this system duplicates historical results under the present system in about (conservatively) 4.5 out of 5 candidates. In the absence of slates, it generally gives identical results to the current nomination system.


18. I think we should just increase the number of nomination slots on the final ballot to a larger number (for example, 6), and decrease the number of slots a member can nominate to a smaller number (for example, 4). Wouldn't that be simpler and easier?
Unfortunately, this simply means that the slate with the largest number of supporters will receive four of the nominations and the slate with the next largest number of supporters will receive the remaining two. It doesn't solve the problem of slates depriving non-slate nominees of the opportunity to compete in the final award vote. In general, we want fandom to nominate as many nominees as they feel are Hugo-worthy, since under E Pluribus Hugo there is no strategic reason not to do so. For this reason, most of the designers of this system would prefer that members not be limited to nominating only four nominees. Keep in mind, however, that E Pluribus Hugo will work with this (or most any other) change as well, so one does not preclude the other.


19. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
This could work. The problem is that then you would have a small group of people serving as literal gatekeepers to the Hugo Awards. In spite of some claims on the Internet, this has never been the case in the past. Establishing it would prove those groups who believed it existed in the past would then be correct. Ultimately, human judgement is fallible. The fairness of a committee’s decisions would forever be subject to opinion, and the prestige of the Hugo Awards would forever be tarnished.


20. I think we should use [insert other mathematical voting system].
We considered essentially every applicable type of voting system currently in the literature, guided by two experts in the field. It should be kept in mind, however, that the goals and requirements for choosing a set of representatives in a political situation are different from those for choosing a set of Hugo finalists. Some of these systems do, in fact, have positive properties that speak for them. None of them were as simple or as intuitive as E Pluribus Hugo, yet E Pluribus Hugo meets all of the stated goals for a Hugo nomination system.


21. Won’t E Pluribus Hugo be complicated to code and implement?
One of our non-experts coded a full simulator for the system in a matter of days. A full web-based app would not be much more difficult to handle.


22. Isn't this change just designed to keep certain people out of the Hugo nomination process?
No. E Pluribus Hugo is designed to fix a flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the award is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.

Although it is true that the discussions for this system were hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resists bloc or slate nominating, treats all individual nominators equally, and allows a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.

#624 ::: David Dyer-Bennet ::: (view all by) ::: June 11, 2015, 01:49 AM:

Well, setting aside arguments we've had and I lost, this all looks good to me. I'll try to make a careful pass through at the typo level rather than the concept level, I guess tomorrow morning.

#625 ::: Annie Y ::: (view all by) ::: June 11, 2015, 02:05 AM:

FAQ 7 and 8 are repeated - with somewhat different answers. We should either combine or have different texts for the questions.

#626 ::: Keith "Kilo" Watt ::: (view all by) ::: June 11, 2015, 02:16 AM:

Annie@625:

Oops, totally forgot to delete 8 and renumber after I combined 7 and 8. I'll fix that now.

K

#627 ::: Keith "Kilo" Watt ::: (view all by) ::: June 11, 2015, 02:19 AM:

Updated FAQ, deleting #8 and renumbering...

FAQ’s:

1. Can you explain the system in plain language?
First, and most importantly, E Pluribus Hugo makes absolutely no changes in how members nominate (no ranking or complex strategies are required). The new system changes only the way nominations are tallied in order to create a more robust measure of which nominees have the broadest and deepest support among WSFS members. The final Hugo voting system, which actually chooses the winner, is completely unchanged.

The new system narrows down Worldcon members' nominations by knocking out the least supported works in repeated elimination rounds until only five (under current rules) finalists remain. Here are the basic steps, as applied to one category:

a. All the nominations for all nominees (that is, the number of ballots on which each nominee appears) are added up, just like in the current system. This number, called "total nominations" in the proposal, is one of the criteria used to determine the final nominees with the broadest support.

b. Next, one point is divided equally among all the nominees in a category on each member's ballot. If there are two nominees on a ballot in a category, each gets 1/2 point; if there are three nominees, each gets 1/3 point, and so on.

c. All the points for each nominee from all nomination ballots are then added together. This number, called "point total" in the proposal, is the second of the two criteria used to determine the final nominees with the broadest support.

d. The nominees with the lowest point totals are compared to each other. The nominee with the fewest number of nominations is eliminated, because it is the nominee with the least support.

e. After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed among their remaining nominees. So, if there were four nominees getting 1/4 point each, there are now three nominees getting 1/3 point each, and so on. All points are re-calculated for all nominees. As less-supported nominees are eliminated, the total points for remaining nominees increase.

This process is repeated until the designated number of finalists (five in the current version of the constitution) remains.


2. E Pluribus Hugo?
The designers felt this name accurately reflects what the Hugo nomination system should be trying to achieve. Fandom is the Many. Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in science fiction and fantasy. This is a defining characteristic of fandom, and this system embraces it. Out of all of these many tastes in nominees, this system allows a single set of nominations to slowly emerge as the most popular candidates for the Hugo Award: “Out of the Many, a Hugo.”


3. But surely that’s not what the system is formally called?
Okay, sure. Formally, in the context of election theory, this system would be called “single divisible vote with least popular elimination” or SDV-LPE for short. E Pluribus Hugo is both cooler and geekier, as everything is better with a little Latin.


4. Why are there Selection and Elimination Phases? Couldn’t you simplify the system by doing it all in one step per elimination?
The Selection Phase looks for nominees with the lowest point total, and the Elimination Phase eliminates the one with the fewest nominations. Looking at the two different measures of popularity adds robustness to the system. There are strategies to manipulate each of these two measurements, but it is very difficult to manipulate both in the same direction. A strategy that focuses on nominations for the Elimination Phase (such as bloc voting) will be stymied by the Selection Phase; a strategy that focuses on points for the Selection Phase (such as nominating only one nominee) will turn out to be counterproductive in the Elimination Phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.


5. Why re-run the system if someone declines a nomination? Why not just take the nominee that was eliminated in the next to last round?
We considered that as an option, including analyzing and simulating the effects on the vote either way. We discovered only one major difference in the results: When taking the “6th place” nominee, the declined nomination will tend to be replaced by a non-slate nominee. With our system of re-running the vote, if a slate nominee makes the final ballot and its nomination is declined, that nomination will tend to be replaced by another nominee from the slate. Either way works, but as noted above, our goal is to avoid preventing any section of fandom from being represented on the final ballot. The results of re-running the calculation entirely, seem more in line with that philosophy.

6. Shouldn’t I just nominate one work if I want to give it the best chance to win?
The system was crafted specifically to make most strategies irrelevant. For example, if you wanted to support a single nominee you feel strongly about, and there isn't anything else you feel is Hugo-worthy, you can do that. On the other hand, if you nominate four other things that you feel might be Hugo-worthy, you aren't hurting your favorite. This is because if your other nominees get eliminated, then your full support will go to the remaining nominee, just as if you had never nominated the others. If enough people agree with your other four, some of them might make it, too. If enough people don't agree with your favorite, there's nothing you can do to get it on the final ballot, because you will never have more than one point. In general, the best strategy is simple: nominate as many nominees as you feel are worthy.


7. Isn't it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws -- that is, it's impossible to create a voting system in which strategy never makes a difference. The objective therefore is to choose a system whose flaws are not an issue to the election at hand. In this system, "strategic voting" is technically possible, but extremely risky: There's no good way to know ahead of time whether a given strategy will help, or backfire. Here's a specific example:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. Suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all could submit nomination ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its nominations compared in the Elimination Phase (you have no way to increase the number of nominations that Castle receives if it is selected for the Elimination Phase). If you push it up to 5th place, it can still be eliminated by Tentacles if Castle has fewer nominations. If you could somehow push it all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters so precisely before the election. Maybe Rocket isn’t so popular after all and needs your nominations to get over the top. Maybe lots of people will assume Rocket is safe and cause it to end up losing as a result of leaving it off their ballots. Maybe Tentacles has a broad enough base of support that your nominations for Castle aren’t enough to eliminate it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire, and only one, highly-specific way for it to work.

Nominating what you think is Hugo-worthy really is your best strategy.


8. What are E Pluribus Hugo’s flaws?
In rare cases, it is possible that eliminating both members of a tie could change the final ballot slightly from what it would be if the tie were broken. The change is usually in the least popular of the finalists, and requires the two nominees in question to be very close in popularity. Hundreds of simulations were run using real and created data sets to help decide which tie-breaking methods would maintain our goal of supporting the wide range of opinions within fandom. It turned out that all the simulations showed that there was almost no difference in outcomes, no matter how we broke ties. We have chosen, therefore, to break ties in a manner consistent with section 6.4 of the Worldcon constitution. There are a number of additional tie breakers that could be used if it were deemed necessary in the future; however, simulations that have been run by the designers of this system show that it really shouldn’t be required.


9. What are E Pluribus Hugo’s benefits?
Simply put, it reduces the power of bloc voting without eliminating the chance that nominees that do appear on slates will make it to the final ballot. Conversely, it makes it very difficult for slates to prevent non-slate nominees from appearing on the ballot.


10. How does this system eliminate slate or bloc voting?
It doesn't, exactly, nor should a nominee be automatically eliminated just because it appears on a slate. On the other hand, any slate which nominates a full set of five nominees will find that each of its nominations only count 1/5 as much. With “non-slate” nominating, some of your nominees will be slowly eliminated, so your remaining nominees get more and more of your support. Since slate nominees tend to live or die together, they tend to eliminate each other until, in general, only one slate nominee remains. With a large enough support behind the slate (five times as much), the slate may still sweep a category; however, if that many voters support the slate, they arguably deserve to win, and no fair and unbiased system of nomination will prevent that. The answer in that case is, simply, to increase the general pool of voters. Regardless, with E Pluribus Hugo, slates will never receive a disproportionate share of the final ballot, as occurred in the 2015 Hugos.


11. Couldn't supporters of slates just recommend a single nominee for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate has enough supporters that is certainly a viable possibility – it’s also completely fair. It does not force all other nominees off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn't a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a nominee deserves a Hugo nomination, then it should, in fact, appear on the final ballot.


12. What if there are multiple slates (slate wars, “parties”, etc.)?
As with a single slate, the nominees on each slate will tend to eliminate each other until only the most popular one remains. The end result is that even multiple slates are unable to sweep the nominations.


13. What happens if a broadly popular nominee is nominated by a group of unrelated people?
If it is broadly popular, the system will still select that nominee for the final ballot.


14. What happens if a broadly popular nominee also appears on a slate?
Even if it is on a slate, if the nominee garners support from individuals then the system will select that nominee for the final ballot. In general, slates neither help nor hurt any given nominee.


15. What happens when there are a lot of nominees with no obvious favorites and nomination slates are introduced?
Simulations of this scenario showed that slate nominees did receive a larger proportion of nomination slots than they did otherwise. This is a fair and valid result: If there was no general favorite, then members really had no collective preference. Even in this scenario, simulations showed that non-slate nominees were not completely shut out of the final ballot.


16. How do the results of this system compare to the results under the current nomination system?
Statistical tests showed that this system duplicates historical results under the present system in about (conservatively) 4.5 out of 5 candidates. In the absence of slates, it generally gives identical results to the current nomination system.


17. I think we should just increase the number of nomination slots on the final ballot to a larger number (for example, 6), and decrease the number of slots a member can nominate to a smaller number (for example, 4). Wouldn't that be simpler and easier?
Unfortunately, this simply means that the slate with the largest number of supporters will receive four of the nominations and the slate with the next largest number of supporters will receive the remaining two. It doesn't solve the problem of slates depriving non-slate nominees of the opportunity to compete in the final award vote. In general, we want fandom to nominate as many nominees as they feel are Hugo-worthy, since under E Pluribus Hugo there is no strategic reason not to do so. For this reason, most of the designers of this system would prefer that members not be limited to nominating only four nominees. Keep in mind, however, that E Pluribus Hugo will work with this (or most any other) change as well, so one does not preclude the other.


18. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
This could work. The problem is that then you would have a small group of people serving as literal gatekeepers to the Hugo Awards. In spite of some claims on the Internet, this has never been the case in the past. Establishing it would prove those groups who believed it existed in the past would then be correct. Ultimately, human judgement is fallible. The fairness of a committee’s decisions would forever be subject to opinion, and the prestige of the Hugo Awards would forever be tarnished.


19. I think we should use [insert other mathematical voting system].
We considered essentially every applicable type of voting system currently in the literature, guided by two experts in the field. It should be kept in mind, however, that the goals and requirements for choosing a set of representatives in a political situation are different from those for choosing a set of Hugo finalists. Some of these systems do, in fact, have positive properties that speak for them. None of them were as simple or as intuitive as E Pluribus Hugo, yet E Pluribus Hugo meets all of the stated goals for a Hugo nomination system.


20. Won’t E Pluribus Hugo be complicated to code and implement?
One of our non-experts coded a full simulator for the system in a matter of days. A full web-based app would not be much more difficult to handle.


21. Isn't this change just designed to keep certain people out of the Hugo nomination process?
No. E Pluribus Hugo is designed to fix a flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the award is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.

Although it is true that the discussions for this system were hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resists bloc or slate nominating, treats all individual nominators equally, and allows a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.


#628 ::: Soon Lee ::: (view all by) ::: June 11, 2015, 02:43 AM:

[PROPOSAL/FAQ]

Looks good Kilo.

[FAQ]
[FAQ1]The final Hugo voting system, which actually chooses the winner, is completely unchanged. -> delete "completely"

c. All the points for each nominee from all nomination ballots are then added together. This number, called "point total" -> Did we settle on "point total" or "total points"? Because both are used...

#629 ::: Soon Lee ::: (view all by) ::: June 11, 2015, 02:56 AM:

[FAQ]

Q8 is redundant because this latest draft includes it in Q7, so should be deleted & subsequent questions renumbered.

#630 ::: Annie Y ::: (view all by) ::: June 11, 2015, 03:06 AM:

Soon Lee @629

Already done in 627 :)

#631 ::: Soon Lee ::: (view all by) ::: June 11, 2015, 03:17 AM:

Ninja'ed!

#632 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 06:48 AM:

3.A.3: Ties shall be handled as described below:
...
(3) During the Elimination Phase, if two or more nominees are tied for the fewest number of nominations, the nominee with the lowest point total at that round shall be eliminated, unless doing so would leave the number of finalists specified in 3.8.1. In that case—that is, if this is the last elimination round—no nominee shall be eliminated, and elimination rounds shall cease, leaving one more finalist than specified in 3.8.1
...


....
In other words, we should not be going out of our way to break the last tie; just allow both in. It would really suck to be the work that lost on a tiebreaker, far more than it would suck to very, very occasionally have one extra finalist.

#633 ::: David Harmon ::: (view all by) ::: June 11, 2015, 09:02 AM:

Oh, this is looking good. I can still see a few bits of polish to add:

our goal is to avoid preventing any section of fandom from being represented on the final ballot.

Rereading (my own words, yeah), this now rings like a double negative, despite not actually being one. (Maybe it's the V's.) Perhaps "...avoid shutting out any section..."?

#7 needs a bit more punctuation and tightening (added punctuation marked with asterisks):

By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its nominations compared in the Elimination Phase.* (You have no way to increase the number of nominations that Castle receives if it is selected for the Elimination Phase.) If you push it up to 5th place, it can still be eliminated by Tentacles,* if Castle has fewer nominations. If you could somehow push it all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters so precisely before the election. Maybe Rocket isn’t so popular after all,* and needs your nominations to get over the top. Maybe lots of too many people will assume Rocket is safe,* and it loses as a result of leaving because they left it off their ballots. Maybe Tentacles has a broad enough base of support,* that your nominations for Castle aren’t enough to eliminate it. Or Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire, and only one, highly-specific way for it to work.

And I see with 16->15, you redeweaseled it to go from opinion to assertion. Excellent!

#634 ::: Chris Battey ::: (view all by) ::: June 11, 2015, 01:21 PM:

Brad @ 611:

Effectively using strategy in a situation like that still requires unrealistically accurate knowledge of the final ballots. Suppose one single additional person nominated E - now D drops off the ballot anyway, and your ballot is dead, whether A ended up needing the support or not. And if you were wrong and it turned out that A needed the support (e.g. a large number of additional nominators that you don't agree with showed up to submit ballots - that happens sometimes), then your strategy is to blame if A ended up missing the ballot by one nomination, and D never had a chance.

If the propose system makes you want to vote strategically like that, it's your ballot. I don't think it's a good idea, though.

As far as keeping all works off the ballot that got there by "cheating" - looking only at the ballots, how do you distinguish a "cheating" slate from an uncoordinated group of voters that happen to closely align? The latter happens frequently, and most obviously in the Dramatic Presentation, Short Form category - consider a group of Whovians who all think the five best TV episodes of the year all came from Doctor Who, a group of GoT fans whose ballot in that category is all GoT episodes, and so on. Assuming a moderate amount of consensus on which episodes of each show were the best of the year (there are usually a few obvious standouts), the largest group of fans will disproportionately dominate the ballot. From 2011-2014, 13 out of 21 total DPSF nominees were Doctor Who - do you think 66% of nominators were putting DW episodes on the ballot? (Ironically, this is the one place where the Puppy slates actually did increase the diversity of the ballot for 2015.)

You can't detect a bad-faith slate from looking at the nominating ballots alone, and so you cannot build a voting system that tries to keep them off the ballot without an awful lot of collateral damage. And while I understand your impulse to treat slate voters as "adversaries" of the system, any person that purchases a supporting membership still has the right to submit a ballot, and your "battle" metaphor either falls down at that point or goes to some places we really don't want to be.

#635 ::: David Dyer-Bennet ::: (view all by) ::: June 11, 2015, 02:32 PM:

Jameson@632: That's a proposed change? It's kind of drastic. Yeah, it's sad when anything is eliminated off the ballot, but that's, as they say, our one job. The whole point of this exercise is to pare down the complete list of nominations to a list short enough to be considered by voters. We decided a long time ago that the systems that created huge ballots were not what we wanted.

Now, six instead of five isn't huge, and we do allow it in some cases; but equally, I don't see that last elimination as any worse than the previous umpteen, either.

#636 ::: Brad from Sunnyvale ::: (view all by) ::: June 11, 2015, 02:36 PM:

Chris @634 and others -- for a single ballot's strategic efforts to change the results is always going to be unlikely. That is not the source of the disagreement. The question is not this, but whether any given system offers a greater (though still small) chance of success using strategy in contrast to the cost of using strategy. The greater the ratio of strategic benefit to strategic risk, the more likely it is a user will vote in a strategic manner.

With baseline approval as we use today, the ratio of strategic benefit to risk is extremely small. In the proposed system it is still small, but larger. This is because in plain Approval, removing a candidate simply removes support for that candidate, reducing its chance of competing with your preferred choice. In the new proposal removing a candidate both reduces its chance of competing and also increases the support you give your choice.

Now, one thing we can't answer at present is, how many people will alter ballots because of this increase? We don't know, but it is something to consider.

As noted, no single ballot ever makes the difference in anything but a very close situation, and of course no strategy does either, but what is true is that when the benefit ratio of strategy increases, a voter can conclude, "In the rare event that a single ballot makes the difference, I can increase the chances it goes my way." That's the source of disagreement here -- I suspect people don't think about how rare it is for a single ballot to make the difference, because it is for that rare event that they are voting at all. Some will evaluate whether to be strategic, *presuming we are in that situation*.

Anyway, thus my call, apropos the topic of this thread, to reduce language stating that this voting system scores well on immunity to strategy. Approval scores better.

Now, as to whether fans would accept a special and presumably-never-used system of human judgment as the solution to Hugo problems, that's for a different thread, but I do believe that in fact that is what pretty much all democratic systems do for a good reason. Fans might choose to ignore it, of course. Being nerds as we are, an algorithmic solution has appeal, even though we can only change it after 2 years of work. But I don't think we're that smart, frankly.

Pretty much everybody else who battles a dynamic adversary with success starts with algorithms but has an equally dynamic use of humans in counter. That's true in battle, true in policing, true in law, true in computer security, true in business, true even in law where we value the rule of law greatly over the rule of men.

That's a pretty compelling trend. I am a huge fan of finding a way to do things with the rule of law instead of the rule of men, and it takes only overwhelming evidence to drive me the other way.

#637 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 03:07 PM:

@634: "With baseline approval as we use today, the ratio of strategic benefit to risk is extremely small."

Again, the only way I can imagine you could say this is if you don't understand the definition of "strategic". The puppies got an enormous strategic advantage at essentially no risk in the current system. That totally overwhelms any possible benefit/risk ratio under SDV-LPE. That point is entirely beyond debate.

Furthermore, I simply disagree with your philosophical point that a human system is better for this case than an algorithmic one. This is not a case of computer security, where the good guys are well-defined, and the task is simply to ensure they win. Different voting algorithms represent different philosophies of how to aggregate preferences; that is, of how to figure out who the good guys even are. The fact that no algorithm can possibly be 100% strategy-proof shows that unless you assume all voters are honest, no philosophy of social choice can be entirely coherent. Passing the buck to a set of human gatekeepers wouldn't resolve that philosophical problem; it would just make it vaguer and more contentious.

You might have further arguments to make in this philosophical sense. But given that the chances that such arguments will lead to a change in this proposal are essentially null, I don't think the discussion is likely to be productive.

Earlier in this process, I found your contributions thought-provoking and useful. But now it seems to me that you are just tiresomely insisting that you are right and at least half a dozen other people here — with collectively more mathematical credentials than you have — are all missing something(s) basic.

#638 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 03:09 PM:

Oops; my @637 was a response to Brad's @636, not to @634. Sorry.

#639 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 03:17 PM:

@635: Yes, my @632 is a proposal. I've actually tried to make substantively this same proposal twice before, but apparently I didn't explain what I meant well enough until now.

You say it's "drastic", but to me it seems pretty minor. It's just a small change to a tiebreaker, which would only be a factor in a minority of elections. It's not the only way we could end up with 6 finalists. And yes, the last elimination — the one I'm modifying the rules for — is qualitatively different from all the preceding eliminations. For any other elimination round, if we look at Alpha and Beta and eliminate Beta, the odds are very high that Alpha will be eliminated in the next round. The only common exception to that rule would be if the fandom for Alpha and Beta had a greater-than-usual overlap; for instance, if they were members of a common slate. But in the last elimination round, the choice between Alpha and Beta is inevitably significant. So in earlier rounds, you can't avoid breaking ties somehow; in the last round, I think it's better to err on the side of not breaking them.

#640 ::: P J Evans ::: (view all by) ::: June 11, 2015, 03:26 PM:

632
The current rules, AIUI, say a minimum of four nominees and a maximum of seven, including No Award, on the final ballot. I see no reason why that should be changed.

#641 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 03:36 PM:

@640: I'm not proposing changing that. What makes you think I am? I'm proposing that if in the last round the two nominees are tied on nominations, we should have 6 nominees instead of breaking the tie using points and having 5.

#642 ::: Cassy B. ::: (view all by) ::: June 11, 2015, 03:45 PM:

Jameson Quinn @641, isn't that already the default? If it's tied for fifth, both candidates get on the list? Or am I missing something?

#643 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 03:49 PM:

@642: Currently, if they're tied and we can't break the tie, both get in. I'm saying that they should both get in even if we could break the tie; that we shouldn't even try. It's a small difference, and it would only rarely come up; but for the people it would affect, I think my proposal is clearly the fairer way to do it.

#644 ::: David Dyer-Bennet ::: (view all by) ::: June 11, 2015, 03:51 PM:

P J Evans@640: In the copy of the constitution here, which may be a year out of date, 3.8.1 says all the tied nominations in a last-place tie are included on the final ballot (no limit of 7, or 6 excluding No Award).

3.8.5 says the top three nominations make the ballot even if they miss the 5% cutoff, so that's a minimum of four including No Award as you say.

(There's a technical possibility, with no real-world relevance, that can result in a ballot of fewer than four including No Award, but it's not real, not relevant, not worth considering. But I can't bring myself to write "the rules require a minimum of four finalists including NO Award" because that's not actually true.)

It's good that the rules don't have a cutoff at 7, because our tallying procedure would need to specify how to deal with that; what we specify right now does NOT actually cut off at any given number, though high numbers are less and less likely with any real-world data set of significant size, to the point where I'm perfectly fine with a technically-unlimited finalist list.

(There's nothing preventing an n-way tie for last place in both nominations and points, and we specify that everything in a perfect tie for last place goes on the final ballot. That's also true of the current system, it includes all ties for last place on the ballot, and it's easier to have a tie with just one measure than with our two. Hence there's no technical limit on the size of the final ballot, but we have years of experience showing it doesn't actually produce large final ballots, so I'm not worried about it.)

#645 ::: P J Evans ::: (view all by) ::: June 11, 2015, 03:59 PM:

644
I checked the description at the Hugo site, which says with a two-way tie for fifth or three-way tie for fourth, they'd all go on the ballot. Doesn't say anything about more than that.

#646 ::: David Dyer-Bennet ::: (view all by) ::: June 11, 2015, 04:02 PM:

Jameson@634: I can't see a fairness issue there. Using the same elimination rule we use in every other elimination step is clearly the fair thing to do—or, if it isn't, then we should change the rule for all steps.

And of course we don't want to do that; the two different metrics for the two different steps are an important part of what makes this method robust.

Seriously, presenting it as a fairness issue, to my mind, calls into question the entire proposal. If you think it's more fair to use only half the tie-breaking rules on the last round, why do you think they're fair in earlier rounds?

#647 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 05:34 PM:

@646: 2 reasons:

Because in earlier rounds, whichever isn't eliminated this round will probably be eliminated next round, so a relatively arbitrary choice is OK.

Because in earlier rounds, we can't just give up and stop.

Note that I've tried to bring up this issue several times, but apparently didn't explain myself well enough. I did try to discuss it when we were still talking about tiebreakers. It's a relatively minor corner case, and it wouldn't have been the end of the world if we'd passed over it unnoticed; but now that we're talking about it, it seems to me that the thing I'm proposing should be uncontroversially the right thing to do. I notice that I'm getting more push-back than I expected, and I am not entirely sure why; my current theory is that I'm still doing a poor job of explaining how this could happen, so here's a scenario. After there are only

(Alpha, Beta, Gamma, over 110 points and 140 nominations)
Delta, 100 points, 130 nominations
Epsilon, 95.5 points, 125 nominations
Iota, 94 points, 125 nominations

Lowest two points are Epsilon and Iota. Whoops, they're tied on nominations. Normally, we'd look at points, and eliminate Iota. But I'm saying that the fact that three more Iota voters than Epsilon voters voted for Alpha or Beta or something shouldn't really make a difference; they are really tied.

So why do we break the tie if it's not the last round? Because we need to to keep going. Notice also that if this weren't the last round, after Iota, Epsilon would be the next to go; so it doesn't really matter how we break the tie. (The only time the tiebreaker really matters is if the two selected works are from the same slate; and in that case, it's fair to eliminate one. And the chances of two slate works being tied for 5th are pretty small.)

#648 ::: Chris Battey ::: (view all by) ::: June 11, 2015, 05:40 PM:

Jameson made a good point at 637 that I want to expand upon. The Puppy strategy (align your ballot with others to block out candidates you don't want to see reach the finalist stage) scales well to larger numbers of ballots. The more people (out of the minority who agree with the strategy's goal in the first place) that engage in the strategy, the more effective it becomes. It is also resilient to small changes in the rest of the ballot set - choose any 1% of the nominating ballots to randomly change, duplicate, or drop, and the strategy will be just as effective.

On the other hand, assuming there is a group of voters whose goal is "get both A and B onto the ballot" and A is believed to be widely popular, the drop-A strategy (both in the current approval system and in the proposed SDV-LPE system) does not scale well: the more voters that participate in it, the more likely it is that A will drop off the ballot as a result of the strategy. It is also not resilient at all to small changes in the ballot set (as demonstrated by how specifically we have to construct the ballot set to make the strategy work); minor changes can render the strategy pointless and major ones can make it backfire. I've mentioned this before, but I bring it up again specifically in contrast to the Puppies' strategy.

Even if we stipulate that the drop-the-more-popular strategy is more effective under SDV-LPE than under the current system, it's still a far less effective strategy at large scale than the Puppies' block-out-the-majority strategy, which SDV-LPE mostly neutralizes.

So the statement that SDV-LPE reduces overall strategic effectiveness is still accurate, even if there exist individual strategies that become more effective at small scales.

#649 ::: David Dyer-Bennet ::: (view all by) ::: June 11, 2015, 06:13 PM:

Jameson@647: I just don't get why the last round is so special.

And I continue to think that your argument goes against the whole basis of this tallying method. You talk about the tie-breaker as being "relatively arbitrary" -- but in fact the difference between points and nominations is the core of the value this proposal brings to the table.

#650 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 06:21 PM:

@649: The difference between points and nominations is, as you say, the core of the value of this proposal. But that's because using both adds robustness. In an ideal Hugo election, where there were no slates or blind genre factionalism, points and nominations would point in exactly the same direction, so using both would be no more or less than redundant. But when there is a slate, it can fool only one of the two measures (nominations but not points).

Currently, the Hugos keep a 6th finalists whenever it's tied with 5th on nominations. My proposal is just to keep doing that. It's still SDV-LPE in all regards. It's just a very slight adjustment to the tiebreaker; one which I suggested back when we were talking about that, but apparently without explaining myself well enough for anyone to understand.

#651 ::: Annie Y ::: (view all by) ::: June 11, 2015, 06:24 PM:

David Dyer-Bennet @ 649

Just wanted to say that I agree with you here. If we go down that road, we weaken our arguments for the whole thing. It may make it fairer from statistical or mathematical perspective (ok - if you squint a bit and treat that last tie breaker as very special) but not for this system... Let's not try to overcomplicate things just for the hell of complicating them.

#652 ::: P J Evans ::: (view all by) ::: June 11, 2015, 06:30 PM:

Let's not try to overcomplicate things just for the hell of complicating them.

Or for the sake of getting something closer to some mythical ideal system that we know we can't reach.
Ties are part of the Hugos. Let's not try to get rid of them completely.

#653 ::: David Harmon ::: (view all by) ::: June 11, 2015, 06:50 PM:

Jameson Quinn #647: You're getting pushback because we already argued the tie-breaker thing through. Your arguments were considered, but for whatever reasons, the group decided to go the other way. To try and restart that argument about the substance of the proposal, when everybody else is trying to polish the finished wording, is not being helpful. This goes triple for "Brad from Sunnyvale"'s attempt to rewind the discussion on strategic voting, which got thoroughly nailed down with analysis and simulations.

#654 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 07:33 PM:

Here's how I feel. I raised the idea twice previously, and nobody argued against it. I realize now that that was because I didn't explain it well, and nobody understood what I was talking about. But I actually thought it had made it into the text. So yes, I understand that other people feel that I'm coming out with this at the last minute or after. But it doesn't feel like that to me.

In terms of substance, I feel pretty confident that insofar as this issue actually occurred (which would probably be in a minority of elections), there would be more complaints about breaking the ties than about not breaking them. In that sense, I think my proposal is clearly the right thing to do. It is also closer to what is done currently; I can run bootstraps to get exact numbers, but I'm certain it would increase the average similarity between the outcomes of the current system and the new one.

If the consensus is that it is too much complexity, and my proposal is rejected, I can live with that. But it's basically a sentence and a half, and could probably be written even more succinctly if that were the goal.

#655 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 07:37 PM:

@652:

Ties are part of the Hugos. Let's not try to get rid of them completely.

I feel as if this is an argument in favor of my proposal, not against it.

#656 ::: Annie Y ::: (view all by) ::: June 11, 2015, 07:41 PM:

Jameson,

Is it not about how long the sentence will be if we add it. The issue is that it does not really make sense based on all we had discussed and agreed on (and yes - this exact case was also discussed somewhere along the line - singling out the last tie as special and handling differently).

#657 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 07:50 PM:

@656: Yes, we did discuss treating the last tie differently. And we agreed that we should in fact do so; that is, if the last two are tied on both points and nominations, we should keep both, instead of eliminating both as we do in other cases.

...

It's like programming. It's often worth it to add additional lines of code to handle edge cases, even if it's not strictly necessary (your code wouldn't actually crash without it).

We came up with the basic system; then we needed to add special cases for tiebreakers, because otherwise the system would have gotten stuck; and now I'm saying that in the cases where one specific tiebreaker isn't necessary to avoid getting stuck, we shouldn't use it, and instead should stick closer to the original, unadorned system. That is extra "code" complexity, but in terms of outcomes, it's arguably simpler, or at least closer to the "core" of SDV-LPE.

...

I'm going to stop arguing for this now, unless somebody else speaks up in its favor. As I said, I feel confident that it's the right thing to do — the thing that would cause fewer feelings of unfairness in the long run — but it's a minor issue and not worth holding up the proposal over, especially if (as seems so far) I am alone in my views.

#658 ::: Annie Y ::: (view all by) ::: June 11, 2015, 08:09 PM:

Jameson @650

Currently, the Hugos keep a 6th finalists whenever it's tied with 5th on nominations. My proposal is just to keep doing that.

Hold on. The process is already doing that. This is from the latest text:

3.A.2: ...If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then instead no nominees will be eliminated this round, and all remaining nominees shall appear on the final ballot, extending it if necessary.

Which means that for a tie at #5, we don't eliminate. Your explanation above sounded like something else at the start - which is why everyone reacted. If this is all you want, we already have it. No need for a change or amendment.

So are you good with this? Or do you want us to keep the ties at #6 (which really does not make sense anymore)

#659 ::: Keith "Kilo" Watt ::: (view all by) ::: June 11, 2015, 08:14 PM:

Annie@658:

I believe he's saying to suspend 3.A.3(3) if the tie is at 5th place.

Kilo

#660 ::: Soon Lee ::: (view all by) ::: June 11, 2015, 08:15 PM:

Piping up to agree with David Harmon, Annie Y, & PJ Evans.

We have had some long (very long) discussions that has got us here, very close to the finish line. Let's focus on finalizing the wording & getting the proposal submitted to the WSFS Business Metting.

#661 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 08:16 PM:

With the current proposal, if there is a two-way tie for #5 on nominations, we'd first attempt to break it by points. If it is also tied on points, then we'd keep both.

I am suggesting that if there is a is a two-way tie for #5 on nominations, we'd keep both.

If you put it that way, my proposal actually sounds simpler.

I don't know why it's so hard to explain this. Actually, I guess I do; it seems so simple to me, but I guess for most people who don't literally dream about voting theory, it's not simple. So of course my explanations fail.

Is it clear now? It really is a minor change, and I think it is in the spirit of the proposal as a whole, not some utterly new issue that I'm just randomly bringing in at the last minute.

#662 ::: P J Evans ::: (view all by) ::: June 11, 2015, 08:17 PM:

661
We already went through this, probably in the previous thread. NO.

#663 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 08:18 PM:

Note: @661 was an answer to the question of @658. I was unable to resist making arguments in favor while I was explaining, but I will still hold to @657 otherwise; that is, I'll drop this issue unless at least one other person speaks up in support of it.

#664 ::: Annie Y ::: (view all by) ::: June 11, 2015, 08:25 PM:

@659 Kilo,

Yeah, I just reread the whole thing and this is how I read it initially and then got confused from his sentence.

@661

So how come in a position under 5, we claim that points AND nominations matter but then when we are closer to the finish, we suddenly switch to the old system and ignore points? :)

I understand what you mean. My maths side actually kicked in (after reading again). I still think it is complicating with no good reason. Let's not undermine out own tie-breaker rule. It does not matter if you tie at #100 or #10 or #5.

A tie in this system is ONLY a full tie of the pair of (nominations, points). If the pair is different, it is resolved. We discussed similar things earlier, let's not keep reopening them.

Will there be an upset author who realizes they did not make it because of this? Yes - of course. The same way that the same author will be unhappy if someone with less nominations make it.

The backbone of this proposal is that JUST nominations are not enough - we introduce points to help even out the field and eliminate strategy vote. Let's not then contradict ourselves and say "well, yeah but when we get close to the goal, they actually are enough". It makes sense on the maths/programming side. It does not make sense when you explain this to non-specialists. It makes the whole system prone to a "so why bother with points?", Or "so why not switch to nominations only 2 positions earlier?". And even if I know you can answer those, you cannot answer them in an non-arbitrary for non mathematicians way.

Does this make sense?

#665 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 08:31 PM:

@662: In the current discussion, there have been arguments against this proposal, including the fact that it's too late to be bringing it up. When I brought it up previously, if I recall, there was one person who misunderstood and argued against something I hadn't proposed, and I responded explaining what I had proposed, and then it fell out of the thread. Go ahead and say "it's too late for that", or "that's too complex", or "if the tiebreaker is fair sometimes than it's fair always", or "you're the only one who seems to want this"; those are all perfectly valid arguments against my proposal. But "we already dealt with this, and you're just being obtuse" honestly doesn't seem fair to me.

And yes, I realize that there is no way for me to say that without seeming whiny; I've basically given up on this issue, but am still feeling defensive because it seems as if people think I'm re-litigating something that was settled, while I feel as if it wasn't settled because nobody understood the issue until now.

#666 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 08:32 PM:

@664: That is a good answer, thank you.

#667 ::: Annie Y ::: (view all by) ::: June 11, 2015, 08:57 PM:

@665

I suspect that you underestimate at least a few people.:) Plus noone is calling you obtuse - we all are saying that we agreed on how to handle ties, let's stop trying to make it better. So it is not personally against you - it is against the motion to reopen a closed topic on which we had a pretty lively discussion (check the old thread and count how many of the people that have their names on the proposal did not get the tie-breaker they believed to be the best one - and not all of them because they were convinced but because we need an agreement and perfect has a different value for different people when the margins are so small). If you can support this one (and do not believe it to be grossly unfair) - let's call it a day and close and submit that thing, shall we? :)

I would have been surprised if you were not getting defensive considering that voting is your thing really. You think that this makes things better and you are arguing about it. But if we keep doing this, we won't submit this thing until it is time for you to catch the train for Spokane :) And every day we do not submit, someone will come up with an idea to make it better. Or challenge it. And we will keep doing that. We need to close it. Even if it is not perfect according to your standards. Or someone else's. It will never be the perfect system - everyone here knows it. But it is a lot better than the current and it achieves its goals.

Now I am going to stop because I think we do not have anyone unconvinced but you and I had been cross posting a bit so... just wanted to finish my thoughts. Sorry for the somewhat rambling post :)

#668 ::: Brad from Sunnyvale ::: (view all by) ::: June 11, 2015, 10:48 PM:

#637 Jameson -- please be careful with using phrases like 'the only way I can imagine you could say this is if you don't understand the definition of "strategic"' -- I am certain your imagination is not that limited, and this approach does not lead to civil debate.

There are several meanings of strategic. I distinguish collusion (which is what the slates do) from how I usually see the word strategy used in discussion of voting systems, which occurs when a voter deliberately votes other than what the ballot asks them to do to help or avoid hurting certain candidates.

I would go further -- I view collusion by groups as an entirely different class of action from strategic voting by individuals. The latter is almost impossible to detect and prevent, the former I think we all believe we have detected and wish to find a way to prevent.

However, that said, it seems the some of the authors of this proposal do not concur with my conclusion that the proposed system proffers more reward for individual strategic voting than the one it replaces, so I am unlikely to convince them to change the wording. And so perhaps it shall be. Especially when people start arguing about who has more credentials to make their case.

#669 ::: Brad from Sunnyvale ::: (view all by) ::: June 11, 2015, 10:48 PM:

#637 Jameson -- please be careful with using phrases like 'the only way I can imagine you could say this is if you don't understand the definition of "strategic"' -- I am certain your imagination is not that limited, and this approach does not lead to civil debate.

There are several meanings of strategic. I distinguish collusion (which is what the slates do) from how I usually see the word strategy used in discussion of voting systems, which occurs when a voter deliberately votes other than what the ballot asks them to do to help or avoid hurting certain candidates.

I would go further -- I view collusion by groups as an entirely different class of action from strategic voting by individuals. The latter is almost impossible to detect and prevent, the former I think we all believe we have detected and wish to find a way to prevent.

However, that said, it seems the some of the authors of this proposal do not concur with my conclusion that the proposed system proffers more reward for individual strategic voting than the one it replaces, so I am unlikely to convince them to change the wording. And so perhaps it shall be. Especially when people start arguing about who has more credentials to make their case.

#670 ::: Keith "Kilo" Watt ::: (view all by) ::: June 11, 2015, 10:48 PM:

All: Time for tonight's update. I did a little bit of clean-up, but nothing that should be controversial. Our two measures are "point total" and "number of nominations", so I made sure the language was consistent throughout. (There is one place that refers to just "nominations" in the FAQ, but I thought that particular sentence worked better that way.)

Keep those comments and the typo-crushing coming!

Kilo

#671 ::: Keith "Kilo" Watt ::: (view all by) ::: June 11, 2015, 10:51 PM:

Short Title: E Pluribus Hugo (Out of the Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations), section 3.9 (Notification and Acceptance), and section 3.11 (Tallying of Votes) as follows:

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the process described in section 3.A.

Insert new section 3.A after Section 3.8 as follows:
Section 3.A: Finalist Selection Process

3.A.1: For each category, the finalist selection process shall be conducted as elimination rounds consisting of three phases:
(1) Calculation Phase: First, the total number of nominations (the number of ballots on which each nominee appears) from all eligible ballots shall be tallied for each remaining nominee. Next, a single “point” shall be assigned to each nomination ballot. That point shall be divided equally among all remaining nominees on that ballot. Finally, all points from all nomination ballots shall be totaled for each nominee in that category. These two numbers, point total and number of nominations, shall be used in the Selection and Elimination Phases.
(2) Selection Phase: The two nominees with the lowest point totals shall be selected for comparison in the Elimination Phase. (See 3.A.3 for ties.)
(3) Elimination Phase: Nominees chosen in the Selection Phase shall be compared, and the nominee with the fewest number of nominations shall be eliminated and removed from all ballots for the Calculation Phase of all subsequent rounds. (See 3.A.3 for ties.)

3.A.2: The phases described in 3.A.1 are repeated in order for each category until the number of finalists specified in 3.8.1 remain. If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then instead no nominees will be eliminated during that round, and all remaining nominees shall appear on the final ballot, extending it if necessary.

3.A.3: Ties shall be handled as described below:
(1) During the Selection Phase, if two or more nominees are tied for the lowest point total, all such nominees shall be selected for the Elimination Phase.
(2) During the Selection Phase, if one nominee has the lowest point total and two or more nominees are tied for the second-lowest point total, then all such nominees shall be selected for the Elimination Phase.
(3) During the Elimination Phase, if two or more nominees are tied for the fewest number of nominations, the nominee with the lowest point total at that round shall be eliminated.
(4) During the Elimination Phase, if two or more nominees are tied for both fewest number of nominations and lowest point total, then all such nominees tied at that round shall be eliminated.

3.A.4: After the initial Award ballot is generated, if any finalist(s) are removed for any reason, the finalist selection process shall be rerun as though the removed finalist(s) had never been nominee(s). None of the remaining original finalists who have been notified shall be removed as a result of this rerun. The new finalist(s) shall be merged with the original finalists, extending the final ballot if necessary.

Section 3.9: Notification and Acceptance.
3.9.1 Worldcon Committees shall use reasonable efforts to notify the nominees finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each nominee person notified shall be asked at that time to either accept or decline the nomination. If the nominee the person notified declines the nomination, that nominee finalist(s) shall not appear on the final ballot. The procedure for replacement of such finalist(s) is described in subsection 3.A.4.

Section 3.11: Tallying of Votes.
3.11.4: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. During the same period, the results of the last ten rounds of the finalist selection process for each category (or all the rounds if there are fewer than ten) shall also be published.

Submitted by: Keith “Kilo” Watt, Jameson Quinn, Tammy Coxen, Yoana Yotova, Joshua Kronengold, Christopher Battey, David Gallaher, Adam Tilghman, Sara Watt, CJ Cabourne, Steven Halter, P J Evans, David Goldfarb, Seth Gordon, Ginger Tansey, Steve Wright, Catherine Faber, Andy Holloway, Duncan J. Macdonald, Claudia Beach, Derry Earnshaw, Jason Skiles, Soon Lee, David Harmon, Lydia Nickerson, Abigail Sutherland, Lee Billings, Oskari Rantala, Seth Breidbart, Chris Suslowicz, Fragano Ledgister, Lori Coulson, Jeffry Herman, Mark Shier, Buddha Buck, Lenore Jean Jones, David Langford, Christopher Hatton, Rogers Cadenhead

Commentary: “E Pluribus Hugo” – the name of this proposal is also its goal: recognizing the many opinions within fandom as to what nominees might be worthy of the Hugo. It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members: They still list the nominees, unranked, they feel are Hugo-worthy (up to the maximum permitted) in any categories they choose, just as they have in the past. In the past, we have counted the number of nominations each nominee received, and the top five nominees were put on the final ballot. However, because SF fandom typically nominates a variety of different nominees, it was easy for an organized slate to make it so that no other nominees made the final ballot.

Using this system, fandom isn't penalized for nominating a wide variety of nominees. If you nominate something that ends up not having a chance to make the final ballot, then your remaining choices automatically get more of your support instead of just being wasted. In other words, you can safely nominate anything you feel is Hugo-worthy. If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot. In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which finalists should be voted on to be the Hugo winner.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no group of members – of any sort, minority or majority – can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Award.

#672 ::: Keith "Kilo" Watt ::: (view all by) ::: June 11, 2015, 10:51 PM:

FAQ’s:

1. Can you explain the system in plain language?
First, and most importantly, E Pluribus Hugo makes absolutely no changes in how members nominate (no ranking or complex strategies are required). The new system changes only the way nominations are tallied in order to create a more robust measure of which nominees have the broadest and deepest support among WSFS members. The final Hugo voting system, which actually chooses the winner, is unchanged.

The new system narrows down Worldcon members' nominations by knocking out the least supported works in repeated elimination rounds until only five (under current rules) finalists remain. Here are the basic steps, as applied to one category:

a. All the nominations for all nominees (that is, the number of ballots on which each nominee appears) are added up, just like in the current system. This number, called "number of nominations" in the proposal, is one of the criteria used to determine the final nominees with the broadest support.

b. Next, one point is divided equally among all the nominees in a category on each member's ballot. If there are two nominees on a ballot in a category, each gets 1/2 point; if there are three nominees, each gets 1/3 point, and so on.

c. All the points for each nominee from all nomination ballots are then added together. This number, called "point total" in the proposal, is the second of the two criteria used to determine the final nominees with the broadest support.

d. The nominees with the lowest point totals are compared to each other. The nominee with the fewest number of nominations is eliminated, because it is the nominee with the least support.

e. After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed among their remaining nominees. So, if there were four nominees getting 1/4 point each, there are now three nominees getting 1/3 point each, and so on. All points are re-calculated for all nominees. As less-supported nominees are eliminated, the total points for remaining nominees increase.

This process is repeated until the designated number of finalists (five in the current version of the constitution) remains.


2. E Pluribus Hugo?
The designers felt this name accurately reflects what the Hugo nomination system should be trying to achieve. Fandom is the Many. Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in science fiction and fantasy. This is a defining characteristic of fandom, and this system embraces it. Out of all of these many tastes in nominees, this system allows a single set of nominations to slowly emerge as the most popular candidates for the Hugo Award: “Out of the Many, a Hugo.”


3. But surely that’s not what the system is formally called?
Okay, sure. Formally, in the context of election theory, this system would be called “single divisible vote with least popular elimination” or SDV-LPE for short. E Pluribus Hugo is both cooler and geekier, as everything is better with a little Latin.


4. Why are there Selection and Elimination Phases? Couldn’t you simplify the system by doing it all in one step per elimination?
The Selection Phase looks for nominees with the lowest point total, and the Elimination Phase eliminates the one with the fewest number of nominations. Looking at the two different measures of popularity adds robustness to the system. There are strategies to manipulate each of these two measurements, but it is very difficult to manipulate both in the same direction. A strategy that focuses on number of nominations for the Elimination Phase (such as bloc voting) will be stymied by the Selection Phase; a strategy that focuses on points for the Selection Phase (such as nominating only one nominee) will turn out to be counterproductive in the Elimination Phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.


5. Why re-run the system if someone declines a nomination? Why not just take the nominee that was eliminated in the next to last round?
We considered that as an option, including analyzing and simulating the effects on the vote either way. We discovered only one major difference in the results: If the declined nomination is from a slate, then when taking the “6th place” nominee, the declined nomination will tend to be replaced by a non-slate nominee. With our system of re-running the vote, if a slate nominee makes the final ballot and its nomination is declined, that nomination will tend to be replaced by another nominee from the slate. Either way works, but as noted above, our goal is to avoid shutting out any section of fandom from the final ballot. The results of re-running the calculation, seem more in line with that philosophy.


6. Shouldn’t I just nominate one work if I want to give it the best chance to win?
The system was crafted specifically to make most strategies irrelevant. For example, if you wanted to support a single nominee you feel strongly about, and there isn't anything else you feel is Hugo-worthy, you can do that. On the other hand, if you nominate four other things that you feel might be Hugo-worthy, you aren't hurting your favorite. This is because if your other nominees get eliminated, then your full support will go to the remaining nominee, just as if you had never nominated the others. If enough people agree with your other four, some of them might make it, too. If enough people don't agree with your favorite, there's nothing you can do to get it on the final ballot, because you will never have more than one point. In general, the best strategy is simple: nominate as many nominees as you feel are worthy.


7. Isn't it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws -- that is, it's impossible to create a voting system in which strategy never makes a difference. The objective therefore is to choose a system whose flaws are not an issue to the election at hand. In this system, "strategic voting" is technically possible, but extremely risky: There's no good way to know ahead of time whether a given strategy will help, or backfire. Here's a specific example:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. Suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all could submit nomination ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its number of nominations compared in the Elimination Phase. (You have no way to increase the number of nominations that Castle receives.) If you push it up to 5th place, it can still be eliminated by Tentacles, if Castle has fewer nominations. If you could somehow push it all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters so precisely before the election. Maybe Rocket isn’t so popular after all, and it needs your nominations to get over the top. Maybe too many people will assume Rocket is safe, and it loses because they left it off their ballots. Maybe Tentacles has a broad enough base of support that your nominations for Castle aren’t enough to eliminate it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire, and only one, highly-specific way for it to work.

Nominating what you think is Hugo-worthy really is your best strategy.


8. What are E Pluribus Hugo’s flaws?
In rare cases, it is possible that eliminating both members of a tie could change the final ballot slightly from what it would be if the tie were broken. The change is usually in the least popular of the finalists, and requires the two nominees in question to be very close in popularity. Hundreds of simulations were run using real and created data sets to help decide which tie-breaking methods would maintain our goal of supporting the wide range of opinions within fandom. It turned out that all the simulations showed that there was almost no difference in outcomes, no matter how we broke ties. We have chosen, therefore, to break ties in a manner consistent with section 6.4 of the Worldcon constitution. There are a number of additional tie breakers that could be used if it were deemed necessary in the future; however, simulations that have been run by the designers of this system show that it really shouldn’t be required.


9. What are E Pluribus Hugo’s benefits?
Simply put, it reduces the power of bloc voting without eliminating the chance that nominees that do appear on slates will make it to the final ballot. Conversely, it makes it very difficult for slates to prevent non-slate nominees from appearing on the ballot.


10. How does this system eliminate slate or bloc voting?
It doesn't, exactly, nor should a nominee be automatically eliminated just because it appears on a slate. On the other hand, any slate which nominates a full set of five nominees will find that each of its nominations only count 1/5 as much. With “non-slate” nominating, some of your nominees will be slowly eliminated, so your remaining nominees get more and more of your support. Since slate nominees tend to live or die together, they tend to eliminate each other until, in general, only one slate nominee remains. With a large enough support behind the slate (five times as much), the slate may still sweep a category; however, if that many voters support the slate, they arguably deserve to win, and no fair and unbiased system of nomination will prevent that. The answer in that case is, simply, to increase the general pool of voters. Regardless, with E Pluribus Hugo, slates will never receive a disproportionate share of the final ballot, as occurred in the 2015 Hugos.


11. Couldn't supporters of slates just recommend a single nominee for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate has enough supporters, that is certainly a viable possibility – it’s also completely fair. It does not force all other nominees off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn't a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a nominee deserves a Hugo nomination, then it should, in fact, appear on the final ballot.


12. What if there are multiple slates (slate wars, “parties”, etc.)?
As with a single slate, the nominees on each slate will tend to eliminate each other until only the most popular one remains. The end result is that even multiple slates are unable to sweep the nominations.


13. What happens if a broadly popular nominee is nominated by a group of unrelated people?
If it is broadly popular, the system will still select that nominee for the final ballot.


14. What happens if a broadly popular nominee also appears on a slate?
If the nominee garners support from individuals, then the system will select that nominee for the final ballot, even if it is on a slate. In general, slates neither help nor hurt any given nominee.


15. What happens when there are a lot of nominees with no obvious favorites and nomination slates are introduced?
Simulations of this scenario showed that slate nominees did receive a larger proportion of nomination slots than they did otherwise. This is a fair and valid result: If there was no general favorite, then members really had no collective preference. Even in this scenario, simulations showed that non-slate nominees were not completely shut out of the final ballot.


16. How do the results of this system compare to the results under the current nomination system?
Statistical tests showed that this system duplicates historical results under the present system in about (conservatively) 4.5 out of 5 candidates. In the absence of slates, it generally gives identical results to the current nomination system.


17. I think we should just increase the number of nomination slots on the final ballot to a larger number (for example, 6), and decrease the number of slots a member can nominate to a smaller number (for example, 4). Wouldn't that be simpler and easier?
Unfortunately, this simply means that the slate with the largest number of supporters will receive four of the nominations and the slate with the next largest number of supporters will receive the remaining two. It doesn't solve the problem of slates depriving non-slate nominees of the opportunity to compete in the final award vote. In general, we want fandom to nominate as many nominees as they feel are Hugo-worthy, since under E Pluribus Hugo there is no strategic reason not to do so. For this reason, most of the designers of this system would prefer that members not be limited to nominating only four nominees. Keep in mind, however, that E Pluribus Hugo will work with this (or most any other) change as well, so one does not preclude the other.


18. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
This could work. The problem is that then you would have a small group of people serving as literal gatekeepers to the Hugo Awards. In spite of some claims on the Internet, this has never been the case in the past. Establishing it would prove those groups who believed it existed in the past would then be correct. Ultimately, human judgement is fallible. The fairness of a committee’s decisions would forever be subject to opinion, and the prestige of the Hugo Awards would forever be tarnished.


19. I think we should use [insert other mathematical voting system].
We considered essentially every applicable type of voting system currently in the literature, guided by two experts in the field. It should be kept in mind, however, that the goals and requirements for choosing a set of representatives in a political situation are different from those for choosing a set of Hugo finalists. Some of these systems do, in fact, have positive properties that speak for them. None of them were as simple or as intuitive as E Pluribus Hugo, yet E Pluribus Hugo meets all of the stated goals for a Hugo nomination system.


20. Won’t E Pluribus Hugo be complicated to code and implement?
One of our non-experts coded a full simulator for the system in a matter of days. A full web-based app would not be much more difficult to handle.


21. Isn't this change just designed to keep certain people out of the Hugo nomination process?
No. E Pluribus Hugo is designed to fix a flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the award is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.

Although it is true that the discussions for this system were hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resists bloc or slate nominating, treats all individual nominators equally, and allows a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.

#673 ::: Jameson Quinn ::: (view all by) ::: June 11, 2015, 11:11 PM:

@668: You responded gracefully, and so it's my turn to apologize.

Sorry for my abrasive language.

#674 ::: David Dyer-Bennet ::: (view all by) ::: June 11, 2015, 11:16 PM:

Jameson@657: Thanks for this explanation. While, sorry, it doesn't change my position, it does make me understand what's in your head at least a little better, so your positions sound more reasonable.

More broadly, I think it's at least possible that people did understand your alternative tie-breaker proposal previously, and it wasn't argued against because nobody else supported it; so it was just left for dead. Similarly, I think people largely are understanding your recent arguments, we're just not finding them convincing.

#675 ::: Soon Lee ::: (view all by) ::: June 11, 2015, 11:28 PM:

[FAQ4]
Looking at the two different measures of popularity -> Looking at two

[FAQ15]
This is a fair and valid result: If there was no general favorite -> obvious favorite

#676 ::: Annie Y ::: (view all by) ::: June 12, 2015, 02:29 AM:

F18

Establishing it would prove those groups who believed it existed in the past would then be correct.

That sentence sound clunky to me. Attempt at a rewrite:

Establishing it would prove that those groups who believed it existed in the past had been correct.
or
Establishing it would prove that those groups, who believed it existed in the past, had been correct.
#677 ::: David Harmon ::: (view all by) ::: June 12, 2015, 06:58 AM:

Soon Lee #675: Both of those slightly change the sense of the text, and I think for the worse.

"The two different measures" are not just any two measures but the ones that were just defined, and a "general" favorite need not be "obvious".

#678 ::: Soon Lee ::: (view all by) ::: June 12, 2015, 11:23 AM:

@David Harmon:
I see where you're coming from with the former but something about the latter's wording feels off. How about: "If there was no overall favorite..."?

(I can live with current wording)

#679 ::: P J Evans ::: (view all by) ::: June 12, 2015, 11:30 AM:

675-678
I think you might mean 'if there were no overall favorite'. But that's my grammar nitpick.

#680 ::: Kevin Standlee ::: (view all by) ::: June 12, 2015, 12:05 PM:

David Dyer-Bennet @644:

In the copy of the constitution here, which may be a year out of date,...
That is the current Constitution. The reason it says "August 2014" is because it's what took effect as of the end of the 2014 Worldcon. It will be in effect through the end of the 2015 Worldcon.

P J Evans @645:

I checked the description at the Hugo site, which says with a two-way tie for fifth or three-way tie for fourth, they'd all go on the ballot. Doesn't say anything about more than that.
You're mistaking examples of possible cases as if they were the only possibilities. In case of a tie involving the final position, all of the tied works are included. The wording at the Hugo Awards web site gives a few examples, but it can't include every possibility because there are literally an infinite number of them.

Imagine if the nominations run 100, 80, 70, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 48,... and that the 5% rule doesn't. That's an eleven-way tie for fourth place, and thus fourteen works make the final ballot (plus No Award). The work with 48 doesn't make the ballot; it's in fifteenth place. This is only one possible example, and should not be interpreted to mean that the rules say that there is a maximum of fifteen finalists under existing rules.

There is no actual hard limit on the number of works that can appear on the final ballot. Assuming a 5% elimination doesn't happen, all works in the top five places (including any ties) appear on the final ballot.

#681 ::: David Goldfarb ::: (view all by) ::: June 12, 2015, 03:04 PM:

Jameson Quinn @647: What happened to Zeta, Eta, and Theta?

#682 ::: P J Evans ::: (view all by) ::: June 12, 2015, 03:17 PM:

681
Epsilon ate them.

#683 ::: Jameson Quinn ::: (view all by) ::: June 12, 2015, 03:33 PM:

Why is 6 afraid of 7, base 8?

Because 7 10 11.

Why is stigma afraid of zeta?

Because zeta ate a theta.

I'll be here all week. Try the veal.

#684 ::: David Dyer-Bennet ::: (view all by) ::: June 12, 2015, 04:54 PM:

P J Evans@679: one of my favorite grammar nitpicks as well.

#685 ::: P J Evans ::: (view all by) ::: June 12, 2015, 04:58 PM:

683
Why can't programmers tell Halloween from Christmas?

Because 31 Oct = 25 Dec.

The chicken's better.

#686 ::: Soon Lee ::: (view all by) ::: June 12, 2015, 05:03 PM:

[FAQ4] What about:

Looking at the two different measures of popularity adds robustness to the system.

-> Looking at the two different measures of popularity, point total and number of nominations, adds robustness to the system.

@PJ Evans & DD-B,
Thanks for the catch. The were/was usage is one of my blindspots.

#687 ::: P J Evans ::: (view all by) ::: June 12, 2015, 05:20 PM:

686
Subjunctive. It's disappearing from English. (I learned it thoroughly in German in high school: I had to explain it to the third-year students, auf Deutsch, bitte.)

#688 ::: David Dyer-Bennet ::: (view all by) ::: June 12, 2015, 07:15 PM:

[OT] P J Evans@687: Amusingly, that's where I really learned it as well. I had the habit somewhat right in English, but German was where I learned what was really going on.

#689 ::: Brad from Sunnyvale ::: (view all by) ::: June 12, 2015, 07:22 PM:

I'm trying to create a genre of dirty versions of kid's jokes. Here's one:


Why did 6 break up with 7?

Because 7 had an affair with 9.

#690 ::: David Harmon ::: (view all by) ::: June 12, 2015, 09:01 PM:

Soon Lee #678: "Overall" works for me. For #686, I don't think repeating the metric names is really necessary there, they're in the previous sentence and used again later in th paragraph.

#691 ::: P J Evans ::: (view all by) ::: June 12, 2015, 09:07 PM:

688
[OT] Probably any European language other than English would do.

#692 ::: Soon Lee ::: (view all by) ::: June 12, 2015, 10:31 PM:

David Harmon #690:

All right then.

Does this mean we're done?

#693 ::: Keith "Kilo" Watt ::: (view all by) ::: June 12, 2015, 10:57 PM:

All: The only change to this version is from #679. Is this it?

Kilo

#694 ::: Keith "Kilo" Watt ::: (view all by) ::: June 12, 2015, 10:58 PM:

Short Title: E Pluribus Hugo (Out of the Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations), section 3.9 (Notification and Acceptance), and section 3.11 (Tallying of Votes) as follows:

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the process described in section 3.A.

Insert new section 3.A after Section 3.8 as follows:
Section 3.A: Finalist Selection Process

3.A.1: For each category, the finalist selection process shall be conducted as elimination rounds consisting of three phases:
(1) Calculation Phase: First, the total number of nominations (the number of ballots on which each nominee appears) from all eligible ballots shall be tallied for each remaining nominee. Next, a single “point” shall be assigned to each nomination ballot. That point shall be divided equally among all remaining nominees on that ballot. Finally, all points from all nomination ballots shall be totaled for each nominee in that category. These two numbers, point total and number of nominations, shall be used in the Selection and Elimination Phases.
(2) Selection Phase: The two nominees with the lowest point totals shall be selected for comparison in the Elimination Phase. (See 3.A.3 for ties.)
(3) Elimination Phase: Nominees chosen in the Selection Phase shall be compared, and the nominee with the fewest number of nominations shall be eliminated and removed from all ballots for the Calculation Phase of all subsequent rounds. (See 3.A.3 for ties.)

3.A.2: The phases described in 3.A.1 are repeated in order for each category until the number of finalists specified in 3.8.1 remain. If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then instead no nominees will be eliminated during that round, and all remaining nominees shall appear on the final ballot, extending it if necessary.

3.A.3: Ties shall be handled as described below:
(1) During the Selection Phase, if two or more nominees are tied for the lowest point total, all such nominees shall be selected for the Elimination Phase.
(2) During the Selection Phase, if one nominee has the lowest point total and two or more nominees are tied for the second-lowest point total, then all such nominees shall be selected for the Elimination Phase.
(3) During the Elimination Phase, if two or more nominees are tied for the fewest number of nominations, the nominee with the lowest point total at that round shall be eliminated.
(4) During the Elimination Phase, if two or more nominees are tied for both fewest number of nominations and lowest point total, then all such nominees tied at that round shall be eliminated.

3.A.4: After the initial Award ballot is generated, if any finalist(s) are removed for any reason, the finalist selection process shall be rerun as though the removed finalist(s) had never been nominee(s). None of the remaining original finalists who have been notified shall be removed as a result of this rerun. The new finalist(s) shall be merged with the original finalists, extending the final ballot if necessary.

Section 3.9: Notification and Acceptance.
3.9.1 Worldcon Committees shall use reasonable efforts to notify the nominees finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each nominee person notified shall be asked at that time to either accept or decline the nomination. If the nominee the person notified declines the nomination, that nominee finalist(s) shall not appear on the final ballot. The procedure for replacement of such finalist(s) is described in subsection 3.A.4.

Section 3.11: Tallying of Votes.
3.11.4: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. During the same period, the results of the last ten rounds of the finalist selection process for each category (or all the rounds if there are fewer than ten) shall also be published.

Submitted by: Keith “Kilo” Watt, Jameson Quinn, Tammy Coxen, Yoana Yotova, Joshua Kronengold, Christopher Battey, David Gallaher, Adam Tilghman, Sara Watt, CJ Cabourne, Steven Halter, P J Evans, David Goldfarb, Seth Gordon, Ginger Tansey, Steve Wright, Catherine Faber, Andy Holloway, Duncan J. Macdonald, Claudia Beach, Derry Earnshaw, Jason Skiles, Soon Lee, David Harmon, Lydia Nickerson, Abigail Sutherland, Lee Billings, Oskari Rantala, Seth Breidbart, Chris Suslowicz, Fragano Ledgister, Lori Coulson, Jeffry Herman, Mark Shier, Buddha Buck, Lenore Jean Jones, David Langford, Christopher Hatton, Rogers Cadenhead

Commentary: “E Pluribus Hugo” – the name of this proposal is also its goal: recognizing the many opinions within fandom as to what nominees might be worthy of the Hugo. It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members: They still list the nominees, unranked, they feel are Hugo-worthy (up to the maximum permitted) in any categories they choose, just as they have in the past. In the past, we have counted the number of nominations each nominee received, and the top five nominees were put on the final ballot. However, because SF fandom typically nominates a variety of different nominees, it was easy for an organized slate to make it so that no other nominees made the final ballot.

Using this system, fandom isn't penalized for nominating a wide variety of nominees. If you nominate something that ends up not having a chance to make the final ballot, then your remaining choices automatically get more of your support instead of just being wasted. In other words, you can safely nominate anything you feel is Hugo-worthy. If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot. In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which finalists should be voted on to be the Hugo winner.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no group of members – of any sort, minority or majority – can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Award.

#695 ::: Keith "Kilo" Watt ::: (view all by) ::: June 12, 2015, 10:58 PM:

FAQ’s:

1. Can you explain the system in plain language?
First, and most importantly, E Pluribus Hugo makes absolutely no changes in how members nominate (no ranking or complex strategies are required). The new system changes only the way nominations are tallied in order to create a more robust measure of which nominees have the broadest and deepest support among WSFS members. The final Hugo voting system, which actually chooses the winner, is unchanged.

The new system narrows down Worldcon members' nominations by knocking out the least supported works in repeated elimination rounds until only five (under current rules) finalists remain. Here are the basic steps, as applied to one category:

a. All the nominations for all nominees (that is, the number of ballots on which each nominee appears) are added up, just like in the current system. This number, called "number of nominations" in the proposal, is one of the criteria used to determine the final nominees with the broadest support.

b. Next, one point is divided equally among all the nominees in a category on each member's ballot. If there are two nominees on a ballot in a category, each gets 1/2 point; if there are three nominees, each gets 1/3 point, and so on.

c. All the points for each nominee from all nomination ballots are then added together. This number, called "point total" in the proposal, is the second of the two criteria used to determine the final nominees with the broadest support.

d. The nominees with the lowest point totals are compared to each other. The nominee with the fewest number of nominations is eliminated, because it is the nominee with the least support.

e. After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed among their remaining nominees. So, if there were four nominees getting 1/4 point each, there are now three nominees getting 1/3 point each, and so on. All points are re-calculated for all nominees. As less-supported nominees are eliminated, the total points for remaining nominees increase.

This process is repeated until the designated number of finalists (five in the current version of the constitution) remains.


2. E Pluribus Hugo?
The designers felt this name accurately reflects what the Hugo nomination system should be trying to achieve. Fandom is the Many. Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in science fiction and fantasy. This is a defining characteristic of fandom, and this system embraces it. Out of all of these many tastes in nominees, this system allows a single set of nominations to slowly emerge as the most popular candidates for the Hugo Award: “Out of the Many, a Hugo.”


3. But surely that’s not what the system is formally called?
Okay, sure. Formally, in the context of election theory, this system would be called “single divisible vote with least popular elimination” or SDV-LPE for short. E Pluribus Hugo is both cooler and geekier, as everything is better with a little Latin.


4. Why are there Selection and Elimination Phases? Couldn’t you simplify the system by doing it all in one step per elimination?
The Selection Phase looks for nominees with the lowest point total, and the Elimination Phase eliminates the one with the fewest number of nominations. Looking at the two different measures of popularity adds robustness to the system. There are strategies to manipulate each of these two measurements, but it is very difficult to manipulate both in the same direction. A strategy that focuses on number of nominations for the Elimination Phase (such as bloc voting) will be stymied by the Selection Phase; a strategy that focuses on points for the Selection Phase (such as nominating only one nominee) will turn out to be counterproductive in the Elimination Phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.


5. Why re-run the system if someone declines a nomination? Why not just take the nominee that was eliminated in the next to last round?
We considered that as an option, including analyzing and simulating the effects on the vote either way. We discovered only one major difference in the results: If the declined nomination is from a slate, then when taking the “6th place” nominee, the declined nomination will tend to be replaced by a non-slate nominee. With our system of re-running the vote, if a slate nominee makes the final ballot and its nomination is declined, that nomination will tend to be replaced by another nominee from the slate. Either way works, but as noted above, our goal is to avoid shutting out any section of fandom from the final ballot. The results of re-running the calculation, seem more in line with that philosophy.


6. Shouldn’t I just nominate one work if I want to give it the best chance to win?
The system was crafted specifically to make most strategies irrelevant. For example, if you wanted to support a single nominee you feel strongly about, and there isn't anything else you feel is Hugo-worthy, you can do that. On the other hand, if you nominate four other things that you feel might be Hugo-worthy, you aren't hurting your favorite. This is because if your other nominees get eliminated, then your full support will go to the remaining nominee, just as if you had never nominated the others. If enough people agree with your other four, some of them might make it, too. If enough people don't agree with your favorite, there's nothing you can do to get it on the final ballot, because you will never have more than one point. In general, the best strategy is simple: nominate as many nominees as you feel are worthy.


7. Isn't it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws -- that is, it's impossible to create a voting system in which strategy never makes a difference. The objective therefore is to choose a system whose flaws are not an issue to the election at hand. In this system, "strategic voting" is technically possible, but extremely risky: There's no good way to know ahead of time whether a given strategy will help, or backfire. Here's a specific example:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. Suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all could submit nomination ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its number of nominations compared in the Elimination Phase. (You have no way to increase the number of nominations that Castle receives.) If you push it up to 5th place, it can still be eliminated by Tentacles, if Castle has fewer nominations. If you could somehow push it all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters so precisely before the election. Maybe Rocket isn’t so popular after all, and it needs your nominations to get over the top. Maybe too many people will assume Rocket is safe, and it loses because they left it off their ballots. Maybe Tentacles has a broad enough base of support that your nominations for Castle aren’t enough to eliminate it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire, and only one, highly-specific way for it to work.

Nominating what you think is Hugo-worthy really is your best strategy.


8. What are E Pluribus Hugo’s flaws?
In rare cases, it is possible that eliminating both members of a tie could change the final ballot slightly from what it would be if the tie were broken. The change is usually in the least popular of the finalists, and requires the two nominees in question to be very close in popularity. Hundreds of simulations were run using real and created data sets to help decide which tie-breaking methods would maintain our goal of supporting the wide range of opinions within fandom. It turned out that all the simulations showed that there was almost no difference in outcomes, no matter how we broke ties. We have chosen, therefore, to break ties in a manner consistent with section 6.4 of the Worldcon constitution. There are a number of additional tie breakers that could be used if it were deemed necessary in the future; however, simulations that have been run by the designers of this system show that it really shouldn’t be required.


9. What are E Pluribus Hugo’s benefits?
Simply put, it reduces the power of bloc voting without eliminating the chance that nominees that do appear on slates will make it to the final ballot. Conversely, it makes it very difficult for slates to prevent non-slate nominees from appearing on the ballot.


10. How does this system eliminate slate or bloc voting?
It doesn't, exactly, nor should a nominee be automatically eliminated just because it appears on a slate. On the other hand, any slate which nominates a full set of five nominees will find that each of its nominations only count 1/5 as much. With “non-slate” nominating, some of your nominees will be slowly eliminated, so your remaining nominees get more and more of your support. Since slate nominees tend to live or die together, they tend to eliminate each other until, in general, only one slate nominee remains. With a large enough support behind the slate (five times as much), the slate may still sweep a category; however, if that many voters support the slate, they arguably deserve to win, and no fair and unbiased system of nomination will prevent that. The answer in that case is, simply, to increase the general pool of voters. Regardless, with E Pluribus Hugo, slates will never receive a disproportionate share of the final ballot, as occurred in the 2015 Hugos.


11. Couldn't supporters of slates just recommend a single nominee for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate has enough supporters, that is certainly a viable possibility – it’s also completely fair. It does not force all other nominees off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn't a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a nominee deserves a Hugo nomination, then it should, in fact, appear on the final ballot.


12. What if there are multiple slates (slate wars, “parties”, etc.)?
As with a single slate, the nominees on each slate will tend to eliminate each other until only the most popular one remains. The end result is that even multiple slates are unable to sweep the nominations.


13. What happens if a broadly popular nominee is nominated by a group of unrelated people?
If it is broadly popular, the system will still select that nominee for the final ballot.


14. What happens if a broadly popular nominee also appears on a slate?
If the nominee garners support from individuals, then the system will select that nominee for the final ballot, even if it is on a slate. In general, slates neither help nor hurt any given nominee.


15. What happens when there are a lot of nominees with no obvious favorites and nomination slates are introduced?
Simulations of this scenario showed that slate nominees did receive a larger proportion of nomination slots than they did otherwise. This is a fair and valid result: If there were no overall favorite, then members really had no collective preference. Even in this scenario, simulations showed that non-slate nominees were not completely shut out of the final ballot.


16. How do the results of this system compare to the results under the current nomination system?
Statistical tests showed that this system duplicates historical results under the present system in about (conservatively) 4.5 out of 5 candidates. In the absence of slates, it generally gives identical results to the current nomination system.


17. I think we should just increase the number of nomination slots on the final ballot to a larger number (for example, 6), and decrease the number of slots a member can nominate to a smaller number (for example, 4). Wouldn't that be simpler and easier?
Unfortunately, this simply means that the slate with the largest number of supporters will receive four of the nominations and the slate with the next largest number of supporters will receive the remaining two. It doesn't solve the problem of slates depriving non-slate nominees of the opportunity to compete in the final award vote. In general, we want fandom to nominate as many nominees as they feel are Hugo-worthy, since under E Pluribus Hugo there is no strategic reason not to do so. For this reason, most of the designers of this system would prefer that members not be limited to nominating only four nominees. Keep in mind, however, that E Pluribus Hugo will work with this (or most any other) change as well, so one does not preclude the other.


18. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
This could work. The problem is that then you would have a small group of people serving as literal gatekeepers to the Hugo Awards. In spite of some claims on the Internet, this has never been the case in the past. Establishing it would prove those groups who believed it existed in the past would then be correct. Ultimately, human judgement is fallible. The fairness of a committee’s decisions would forever be subject to opinion, and the prestige of the Hugo Awards would forever be tarnished.


19. I think we should use [insert other mathematical voting system].
We considered essentially every applicable type of voting system currently in the literature, guided by two experts in the field. It should be kept in mind, however, that the goals and requirements for choosing a set of representatives in a political situation are different from those for choosing a set of Hugo finalists. Some of these systems do, in fact, have positive properties that speak for them. None of them were as simple or as intuitive as E Pluribus Hugo, yet E Pluribus Hugo meets all of the stated goals for a Hugo nomination system.


20. Won’t E Pluribus Hugo be complicated to code and implement?
One of our non-experts coded a full simulator for the system in a matter of days. A full web-based app would not be much more difficult to handle.


21. Isn't this change just designed to keep certain people out of the Hugo nomination process?
No. E Pluribus Hugo is designed to fix a flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the award is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.

Although it is true that the discussions for this system were hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resists bloc or slate nominating, treats all individual nominators equally, and allows a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.

#696 ::: Annie Y ::: (view all by) ::: June 12, 2015, 11:20 PM:

What about 676? If everyone think that sentence is fine, ok - but if it just got lost in the shuffle - can someone look at it? :)

#697 ::: Keith "Kilo" Watt ::: (view all by) ::: June 12, 2015, 11:23 PM:

Annie@696:

I didn't see any thoughts on that one as to which wording was preferred. Of the two you mention, I think "Establishing it would prove that those groups who believed it existed in the past had been correct." is better.

Anybody have an opinion?

K

#698 ::: P J Evans ::: (view all by) ::: June 12, 2015, 11:34 PM:

697
I think it needs to be reworded a little - it's hard to parse as it is. Something like:
'Establishing it would appear to prove correct those groups which believed it had existed in the past.'

#699 ::: Annie Y ::: (view all by) ::: June 12, 2015, 11:47 PM:

@698

I like this one as well.

#700 ::: Soon Lee ::: (view all by) ::: June 12, 2015, 11:50 PM:

18. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
This could work. The problem is that then you would have a small group of people serving as literal gatekeepers to the Hugo Awards. In spite of some claims on the Internet, this has never been the case in the past. Establishing it would prove those groups who believed it existed in the past would then be correct. Ultimately, human judgement is fallible. The fairness of a committee’s decisions would forever be subject to opinion, and the prestige of the Hugo Awards would forever be tarnished.

->
18. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
In spite of some claims on the Internet, there has never been a small group of people serving as literal gatekeepers to the Hugo Awards. Establishing such a committee could work, but ultimately human judgement is fallible. The fairness of a committee’s decisions would forever be subject to opinion, and the prestige of the Hugo Awards would forever be tarnished.

#701 ::: David Harmon ::: (view all by) ::: June 13, 2015, 07:38 AM:

One subtle thing I just spotted:

The whole shebang is mostly called "the Hugo Awards", plural. In #2 it slips into singular, and #22, has "...the integrity of the award is to be maintained...", again singular. #17 also uses a singular "award", but I think it's a different usage where that's appropriate.

#702 ::: Tim Illingworth ::: (view all by) ::: June 13, 2015, 10:36 AM:

Is this us, or someone else?

http://teespring.com/hugo-award-rocket#pid=369&cid=6529&sid=front

Tim

#703 ::: David Goldfarb ::: (view all by) ::: June 13, 2015, 01:13 PM:

It's obviously someone else, if you look closely.

Jameson's design sticks to one title per line, and rearranges the order so that the widths work out. This design has the winners in chronological order by their year. That's cool, but to do it the designer has to wrap some titles between lines, and have some lines with a lot of white space inserted in them.

#704 ::: junego ::: (view all by) ::: June 13, 2015, 01:23 PM:

[FAQ]

Soon Lee @700
18. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
This could work. The problem is that then you would have a small group of people serving as literal gatekeepers to the Hugo Awards. In spite of some claims on the Internet, this has never been the case in the past. Establishing it would prove those groups who believed it existed in the past would then be correct. Ultimately, human judgement is fallible. The fairness of a committee’s decisions would forever be subject to opinion, and the prestige of the Hugo Awards would forever be tarnished.
...
->
18. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
In spite of some claims on the Internet, there has never been a small group of people serving as literal gatekeepers to the Hugo Awards. Establishing such a committee could work, but ultimately human judgement is fallible. The fairness of a committee’s decisions would forever be subject to opinion, and the prestige of the Hugo Awards would forever be tarnished.

I prefer your reword overall. Slight change suggested for sentence starting "Establishing such a committee..."

Establishing such a committee would potentially cause more problems than it might resolve because human judgement is fallible.

#705 ::: Jameson Quinn ::: (view all by) ::: June 13, 2015, 05:06 PM:

@702... I feel used. But, whatev.

#706 ::: Keith "Kilo" Watt ::: (view all by) ::: June 13, 2015, 05:18 PM:

Jameson@705:

Meh. Your design looks better...

K

#707 ::: Soon Lee ::: (view all by) ::: June 13, 2015, 05:41 PM:

junego #704:

That works for me too. "E pluribus consilium"?
(Either Google tranlate gave me the correct word, or my witticism just fell flat)


Jameson #705:

That is... very... annoying.

#708 ::: P J Evans ::: (view all by) ::: June 13, 2015, 06:38 PM:

706
That's true.

#709 ::: Jameson Quinn ::: (view all by) ::: June 13, 2015, 08:46 PM:

Guess it goes to show that it's better to ask forgiveness than permission. I mean, if instead of discussing it here, I'd showed up at Sasquan with my design as a Electology.org fundraiser... I would have apparently sold at least 8 shirts, and nobody would have sicced lawyers on me.

I'm not saying I regret any goodwill I've built up here. I am just saying that working by consensus rather than by initiative has its downsides as well as its upsides.

#710 ::: David Harmon ::: (view all by) ::: June 13, 2015, 09:35 PM:

Jameson Quinn #709: It's not really a given that you'd have sold those shirts -- they may well represent what the designer wanted for themself and their friends. For that matter, didn't one of the iterations posted here have the titles in order but wrapped?

#711 ::: David Harmon ::: (view all by) ::: June 13, 2015, 09:45 PM:

Rather than "designer", I should probably say "the person who uploaded that design".

#712 ::: Keith "Kilo" Watt ::: (view all by) ::: June 13, 2015, 10:12 PM:

Jameson@709:

Not wanting to rehash an old debate, but I don't think anyone had -- or has -- a problem with your making T-shirts as a fundraiser for Electology. The issue was that we didn't want those shirts to be associated with "E Pluribus Hugo" (or Making Light, obviously), as none of us want to see months of work jeopardized. So if you'd shown up with EPH shirts to sell as a fundraiser, that wouldn't have been okay, permission or no. These shirts obviously aren't EPH shirts.

I'm not interested in these shirts, but I will definitely buy one of your EPH shirts if you decide to put them up on CafePress or wherever!

Kilo

#713 ::: Keith "Kilo" Watt ::: (view all by) ::: June 13, 2015, 10:25 PM:

Soon Lee@700:

Do we want to remove the reference to the fact that establishing a review board would mean that the Puppies' claims would then be true? That was a major argument against the "fannish inquisition" to begin with.

David@701:
That's a good point, regarding singular vs. plural. The constitution currently uses both. The actual service mark (section 2.2) is "Award", but the title of article 3 is "Awards". 3.7.1 is an example where they use "Award" where I would use "Awards". The current 3.8.5 refers to the "final Award ballot". It looks to me like they use the singular more often than the plural, but it's not consistent.

What do you say we use "Awards" when talking about the Hugo Awards in general (such as occurs in the proposal commentary), and "Award" everywhere else?

Kilo

#714 ::: Kevin Standlee ::: (view all by) ::: June 13, 2015, 11:22 PM:

The Teespring shirt appears to have been created as something approved by Loncon 3. (Although I wasn't aware of it at the time.) Worldcon committee not only get the right to use the WSFS service marks, but they also get a limited ability to sub-license them. (Example: official convention shirts and other merchandise containing the word "Worldcon," etc.) There's no precedent for such licensing ever ending, and in fact there are licensed designs for WSFS shirts left over from the 2005 Worldcon that are still out there on CafePress. (I know because I helped with the setup and from time to time buy a new polo shirt with the "White Star Federated Spacelines" logo we created for "Spaceport Glasgow.")

Licensing through a Worldcon committee doesn't need WSFS MPC approval because it's covered by the license to the individual Worldcon committees. And Loncon 3 definitely paid for the right to use those marks, including plumping up a fair bit of money to help pay the substantial legal costs associated with defending Worldcon and WSFS from a trademark troll in 2014.

#715 ::: Soon Lee ::: (view all by) ::: June 14, 2015, 12:25 AM:

Keith "Kilo" Watt #713:

Yes, I think we should remove the "establishing a review board would mean that the Puppies' claims would then be true" statement. I think it's clear enough without it, and trying to fit it in would require some time-travel jiu-jitsu with the phrasing - because it would make it (literal Hugo gatekeepers) true in future but not retroactively so, if you understand my meaning.

If we want to include it *and* can come up with a clearer phrasing, I'm good with that too.

#716 ::: JJ ::: (view all by) ::: June 14, 2015, 03:26 AM:

Kevin Standlee, 714: The Teespring shirt appears to have been created as something approved by Loncon 3.

No, it's something just put up by someone who's been following the shirt design discussion here, who realized they could do so under the aegis of Loncon 3's licensing rights. Note the Loncon Hugo winner "Ancillary Justice", appears at the bottom of the design.

This sort of shameless opportunism, stealing someone else's design, seems to me to be incredibly tacky and ill-spirited.

I won't be buying one of those shirts. I would very much like to see our shirt put up soon -- and with a nicer color choice.

#717 ::: abi ::: (view all by) ::: June 14, 2015, 06:12 AM:

I'm not interested in that Teespring shirt, but I am very much keen to buy in the one Jameson designed.

Why? Because one of them is some kind of knockoff, but the other is part of an effort I have been very much admiring, from a contributor whom I respect.

#718 ::: David Harmon ::: (view all by) ::: June 14, 2015, 10:57 AM:

Soon Lee #715: How about "... establishing such a review board would validate their claims".

#719 ::: P J Evans ::: (view all by) ::: June 14, 2015, 11:07 AM:

on the shirt at Teespring: File 770 has the story.

#720 ::: abi ::: (view all by) ::: June 14, 2015, 11:40 AM:

The designer of the Teespring shirt says that he created it before May. I withdraw the "knockoff" comment, though I'm still really only interested in a shirt associated with E Pluribus Hugo.

#721 ::: Keith "Kilo" Watt ::: (view all by) ::: June 14, 2015, 12:46 PM:

abi@717:

I'm not interested in that Teespring shirt, but I am very much keen to buy in the one Jameson designed.

Why? Because one of them is some kind of knockoff, but the other is part of an effort I have been very much admiring, from a contributor whom I respect.

Hear, hear!

K

#722 ::: David Harmon ::: (view all by) ::: June 14, 2015, 01:28 PM:

Keith "Kilo" Watt #713: What do you say we use "Awards" when talking about the Hugo Awards in general (such as occurs in the proposal commentary), and "Award" everywhere else?

Seems reasonable to me. I note that with respect to our proposal, an individual "Hugo Award" is nearly equivalent to a category.

#723 ::: Jameson Quinn ::: (view all by) ::: June 14, 2015, 02:17 PM:

The designer of the other T (@johncoxon on twitter) has convinced me it's a coincidence and not plagiarism. But we should get our act together to make our T available.

#724 ::: Keith "Kilo" Watt ::: (view all by) ::: June 14, 2015, 02:51 PM:

Jameson@723:

Agreed. I'm driving my daughter to camp today, but will be home this evening. I'll upload the latest version tonight, so if everyone could take a hard look at the draft today and make your suggestions, I think we can get this locked in.

Kilo

#725 ::: David Harmon ::: (view all by) ::: June 14, 2015, 03:34 PM:

FAQ 18: It occurs to me that the use of "forever" and prestige are kinda breathless, while paradoxically not being strong enough.

In spite of some claims on the Internet, there has never been a small group of people serving as literal gatekeepers to the Hugo Awards. Establishing such a committee would validate those who alleged gatekeeping in the past, and the committee's decisions would be continuously in question. Ultimately, this would just make trouble for the Hugo Awards.

#726 ::: Soon Lee ::: (view all by) ::: June 14, 2015, 04:56 PM:

[FAQ18] Playing with the wording.

18. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.

In spite of some claims on the Internet, there has never been a small group of people serving as literal gatekeepers to the Hugo Awards. Establishing such a review committee would validate their claims. While it might work, the fairness of a committee’s decisions would be subject to constant questions, and the prestige of the Hugo Awards would be tarnished.

#727 ::: P J Evans ::: (view all by) ::: June 14, 2015, 05:00 PM:

726
I like that version.

#728 ::: David Harmon ::: (view all by) ::: June 14, 2015, 06:06 PM:

Soon Lee #726: Should we really be saying "it might work", in the middle of explaining why it wouldn't? Still not sure about "prestige", but maybe it's just been a while since I saw a genuine concern about prestige.

#729 ::: David Harmon ::: (view all by) ::: June 14, 2015, 06:13 PM:

Also: I suggest removing "literal" -- the meaning of "gatekeeper" we're invoking is well-entrenched, but still figurative.

#730 ::: David Dyer-Bennet ::: (view all by) ::: June 14, 2015, 06:42 PM:

Soon Lee@726: Unclear pronoun reference in "validate their claims". We refer early on to "claims on the internet" but not to the makers of those claims. (Perhaps a vestige of the previous explicit references to puppies.)

I'm not sure if I agree about "literal" gatekeepers being wrong. But I was bothered by reference to it being a "small" group. The size isn't, I think, important.

And I'm not sure I'm actually liking where this FAQ is ending up currently. The arguments pushing it this way I mostly got on with okay, but it's just not feeling like good rhetoric at the moment. I'll try to see if I can figure out why and/or what to do about it, see if it appeals to anybody else.

#731 ::: David Harmon ::: (view all by) ::: June 14, 2015, 08:33 PM:

David Dyer-Bennet #730: I'm not sure if I agree about "literal" gatekeepers being wrong.

"Actual" would work.

About the overall tone: Well, we've been collectively polishing it for a while. Admittedly, we've filed off bits of snark and grandiosity which added sparkle. But then, we've also been armoring the proposal, against both the ordinary skepticism of the Business Meeting, and possible challenges by a Puppy contingent there.

#732 ::: Keith "Kilo" Watt ::: (view all by) ::: June 14, 2015, 09:46 PM:

All:

I'm back home, so I'll upload the latest version. I'm not very keen on major re-writes at this point -- it seems to me we'd just be going around in circles. There's no One Best Way to say any of this, so there are a number of equally-valid approaches we could take. I really think what we have is looking very good, so I'd like to stay focused on any essential tweaks, as opposed to a re-write.

Kilo

#733 ::: Keith "Kilo" Watt ::: (view all by) ::: June 14, 2015, 09:48 PM:

Short Title: E Pluribus Hugo (Out of the Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations), section 3.9 (Notification and Acceptance), and section 3.11 (Tallying of Votes) as follows:

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the process described in section 3.A.

Insert new section 3.A after Section 3.8 as follows:
Section 3.A: Finalist Selection Process

3.A.1: For each category, the finalist selection process shall be conducted as elimination rounds consisting of three phases:
(1) Calculation Phase: First, the total number of nominations (the number of ballots on which each nominee appears) from all eligible ballots shall be tallied for each remaining nominee. Next, a single “point” shall be assigned to each nomination ballot. That point shall be divided equally among all remaining nominees on that ballot. Finally, all points from all nomination ballots shall be totaled for each nominee in that category. These two numbers, point total and number of nominations, shall be used in the Selection and Elimination Phases.
(2) Selection Phase: The two nominees with the lowest point totals shall be selected for comparison in the Elimination Phase. (See 3.A.3 for ties.)
(3) Elimination Phase: Nominees chosen in the Selection Phase shall be compared, and the nominee with the fewest number of nominations shall be eliminated and removed from all ballots for the Calculation Phase of all subsequent rounds. (See 3.A.3 for ties.)

3.A.2: The phases described in 3.A.1 are repeated in order for each category until the number of finalists specified in 3.8.1 remain. If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then instead no nominees will be eliminated during that round, and all remaining nominees shall appear on the final ballot, extending it if necessary.

3.A.3: Ties shall be handled as described below:
(1) During the Selection Phase, if two or more nominees are tied for the lowest point total, all such nominees shall be selected for the Elimination Phase.
(2) During the Selection Phase, if one nominee has the lowest point total and two or more nominees are tied for the second-lowest point total, then all such nominees shall be selected for the Elimination Phase.
(3) During the Elimination Phase, if two or more nominees are tied for the fewest number of nominations, the nominee with the lowest point total at that round shall be eliminated.
(4) During the Elimination Phase, if two or more nominees are tied for both fewest number of nominations and lowest point total, then all such nominees tied at that round shall be eliminated.

3.A.4: After the initial Award ballot is generated, if any finalist(s) are removed for any reason, the finalist selection process shall be rerun as though the removed finalist(s) had never been nominee(s). None of the remaining original finalists who have been notified shall be removed as a result of this rerun. The new finalist(s) shall be merged with the original finalists, extending the final ballot if necessary.

Section 3.9: Notification and Acceptance.
3.9.1 Worldcon Committees shall use reasonable efforts to notify the nominees finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each nominee person notified shall be asked at that time to either accept or decline the nomination. If the nominee the person notified declines the nomination, that nominee finalist(s) shall not appear on the final ballot. The procedure for replacement of such finalist(s) is described in subsection 3.A.4.

Section 3.11: Tallying of Votes.
3.11.4: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. During the same period, the results of the last ten rounds of the finalist selection process for each category (or all the rounds if there are fewer than ten) shall also be published.

Submitted by: Keith “Kilo” Watt, Jameson Quinn, Tammy Coxen, Yoana Yotova, Joshua Kronengold, Christopher Battey, David Gallaher, Adam Tilghman, Sara Watt, CJ Cabourne, Steven Halter, P J Evans, David Goldfarb, Seth Gordon, Ginger Tansey, Steve Wright, Catherine Faber, Andy Holloway, Duncan J. Macdonald, Claudia Beach, Derry Earnshaw, Jason Skiles, Soon Lee, David Harmon, Lydia Nickerson, Abigail Sutherland, Lee Billings, Oskari Rantala, Seth Breidbart, Chris Suslowicz, Fragano Ledgister, Lori Coulson, Jeffry Herman, Mark Shier, Buddha Buck, Lenore Jean Jones, David Langford, Christopher Hatton, Rogers Cadenhead

Commentary: “E Pluribus Hugo” – the name of this proposal is also its goal: recognizing the many opinions within fandom as to what nominees might be worthy of the Hugo. It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members: They still list the nominees, unranked, they feel are Hugo-worthy (up to the maximum permitted) in any categories they choose, just as they have in the past. In the past, we have counted the number of nominations each nominee received, and the top five nominees were put on the final ballot. However, because SF fandom typically nominates a variety of different nominees, it was easy for an organized slate to make it so that no other nominees made the final ballot.

Using this system, fandom isn't penalized for nominating a wide variety of nominees. If you nominate something that ends up not having a chance to make the final ballot, then your remaining choices automatically get more of your support instead of just being wasted. In other words, you can safely nominate anything you feel is Hugo-worthy. If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot. In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which finalists should be voted on to be the Hugo winner.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no group of members – of any sort, minority or majority – can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Awards.

#734 ::: Keith "Kilo" Watt ::: (view all by) ::: June 14, 2015, 09:51 PM:

FAQ’s:

1. Can you explain the system in plain language?
First, and most importantly, E Pluribus Hugo makes absolutely no changes in how members nominate (no ranking or complex strategies are required). The new system changes only the way nominations are tallied in order to create a more robust measure of which nominees have the broadest and deepest support among WSFS members. The final Hugo voting system, which actually chooses the winner, is unchanged.

The new system narrows down Worldcon members' nominations by knocking out the least supported works in repeated elimination rounds until only five (under current rules) finalists remain. Here are the basic steps, as applied to one category:

a. All the nominations for all nominees (that is, the number of ballots on which each nominee appears) are added up, just like in the current system. This number, called "number of nominations" in the proposal, is one of the criteria used to determine the final nominees with the broadest support.

b. Next, one point is divided equally among all the nominees in a category on each member's ballot. If there are two nominees on a ballot in a category, each gets 1/2 point; if there are three nominees, each gets 1/3 point, and so on.

c. All the points for each nominee from all nomination ballots are then added together. This number, called "point total" in the proposal, is the second of the two criteria used to determine the final nominees with the broadest support.

d. The nominees with the lowest point totals are compared to each other. The nominee with the fewest number of nominations is eliminated, because it is the nominee with the least support.

e. After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed among their remaining nominees. So, if there were four nominees getting 1/4 point each, there are now three nominees getting 1/3 point each, and so on. All points are re-calculated for all nominees. As less-supported nominees are eliminated, the total points for remaining nominees increase.

This process is repeated until the designated number of finalists (five in the current version of the constitution) remains.


2. E Pluribus Hugo?
The designers felt this name accurately reflects what the Hugo nomination system should be trying to achieve. Fandom is the Many. Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in science fiction and fantasy. This is a defining characteristic of fandom, and this system embraces it. Out of all of these many tastes in nominees, this system allows a single set of nominations to slowly emerge as the most popular candidates for the Hugo Awards: “Out of the Many, a Hugo.”


3. But surely that’s not what the system is formally called?
Okay, sure. Formally, in the context of election theory, this system would be called “single divisible vote with least popular elimination” or SDV-LPE for short. E Pluribus Hugo is both cooler and geekier, as everything is better with a little Latin.


4. Why are there Selection and Elimination Phases? Couldn’t you simplify the system by doing it all in one step per elimination?
The Selection Phase looks for nominees with the lowest point total, and the Elimination Phase eliminates the one with the fewest number of nominations. Looking at the two different measures of popularity adds robustness to the system. There are strategies to manipulate each of these two measurements, but it is very difficult to manipulate both in the same direction. A strategy that focuses on number of nominations for the Elimination Phase (such as bloc voting) will be stymied by the Selection Phase; a strategy that focuses on points for the Selection Phase (such as nominating only one nominee) will turn out to be counterproductive in the Elimination Phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.


5. Why re-run the system if someone declines a nomination? Why not just take the nominee that was eliminated in the next to last round?
We considered that as an option, including analyzing and simulating the effects on the vote either way. We discovered only one major difference in the results: If the declined nomination is from a slate, then when taking the “6th place” nominee, the declined nomination will tend to be replaced by a non-slate nominee. With our system of re-running the vote, if a slate nominee makes the final ballot and its nomination is declined, that nomination will tend to be replaced by another nominee from the slate. Either way works, but as noted above, our goal is to avoid shutting out any section of fandom from the final ballot. The results of re-running the calculation, seem more in line with that philosophy.


6. Shouldn’t I just nominate one work if I want to give it the best chance to win?
The system was crafted specifically to make most strategies irrelevant. For example, if you wanted to support a single nominee you feel strongly about, and there isn't anything else you feel is Hugo-worthy, you can do that. On the other hand, if you nominate four other things that you feel might be Hugo-worthy, you aren't hurting your favorite. This is because if your other nominees get eliminated, then your full support will go to the remaining nominee, just as if you had never nominated the others. If enough people agree with your other four, some of them might make it, too. If enough people don't agree with your favorite, there's nothing you can do to get it on the final ballot, because you will never have more than one point. In general, the best strategy is simple: nominate as many nominees as you feel are worthy.


7. Isn't it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws -- that is, it's impossible to create a voting system in which strategy never makes a difference. The objective therefore is to choose a system whose flaws are not an issue to the election at hand. In this system, "strategic voting" is technically possible, but extremely risky: There's no good way to know ahead of time whether a given strategy will help, or backfire. Here's a specific example:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. Suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all could submit nomination ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its number of nominations compared in the Elimination Phase. (You have no way to increase the number of nominations that Castle receives.) If you push it up to 5th place, it can still be eliminated by Tentacles, if Castle has fewer nominations. If you could somehow push it all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters so precisely before the election. Maybe Rocket isn’t so popular after all, and it needs your nominations to get over the top. Maybe too many people will assume Rocket is safe, and it loses because they left it off their ballots. Maybe Tentacles has a broad enough base of support that your nominations for Castle aren’t enough to eliminate it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire, and only one, highly-specific way for it to work.

Nominating what you think is Hugo-worthy really is your best strategy.


8. What are E Pluribus Hugo’s flaws?
In rare cases, it is possible that eliminating both members of a tie could change the final ballot slightly from what it would be if the tie were broken. The change is usually in the least popular of the finalists, and requires the two nominees in question to be very close in popularity. Hundreds of simulations were run using real and created data sets to help decide which tie-breaking methods would maintain our goal of supporting the wide range of opinions within fandom. It turned out that all the simulations showed that there was almost no difference in outcomes, no matter how we broke ties. We have chosen, therefore, to break ties in a manner consistent with section 6.4 of the Worldcon constitution. There are a number of additional tie breakers that could be used if it were deemed necessary in the future; however, simulations that have been run by the designers of this system show that it really shouldn’t be required.


9. What are E Pluribus Hugo’s benefits?
Simply put, it reduces the power of bloc voting without eliminating the chance that nominees that do appear on slates will make it to the final ballot. Conversely, it makes it very difficult for slates to prevent non-slate nominees from appearing on the ballot.


10. How does this system eliminate slate or bloc voting?
It doesn't, exactly, nor should a nominee be automatically eliminated just because it appears on a slate. On the other hand, any slate which nominates a full set of five nominees will find that each of its nominations only count 1/5 as much. With “non-slate” nominating, some of your nominees will be slowly eliminated, so your remaining nominees get more and more of your support. Since slate nominees tend to live or die together, they tend to eliminate each other until, in general, only one slate nominee remains. With a large enough support behind the slate (five times as much), the slate may still sweep a category; however, if that many voters support the slate, they arguably deserve to win, and no fair and unbiased system of nomination will prevent that. The answer in that case is, simply, to increase the general pool of voters. Regardless, with E Pluribus Hugo, slates will never receive a disproportionate share of the final ballot, as occurred in the 2015 Hugos.


11. Couldn't supporters of slates just recommend a single nominee for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate has enough supporters, that is certainly a viable possibility – it’s also completely fair. It does not force all other nominees off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn't a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a nominee deserves a Hugo nomination, then it should, in fact, appear on the final ballot.


12. What if there are multiple slates (slate wars, “parties”, etc.)?
As with a single slate, the nominees on each slate will tend to eliminate each other until only the most popular one remains. The end result is that even multiple slates are unable to sweep the nominations.


13. What happens if a broadly popular nominee is nominated by a group of unrelated people?
If it is broadly popular, the system will still select that nominee for the final ballot.


14. What happens if a broadly popular nominee also appears on a slate?
If the nominee garners support from individuals, then the system will select that nominee for the final ballot, even if it is on a slate. In general, slates neither help nor hurt any given nominee.


15. What happens when there are a lot of nominees with no obvious favorites and nomination slates are introduced?
Simulations of this scenario showed that slate nominees did receive a larger proportion of nomination slots than they did otherwise. This is a fair and valid result: If there were no overall favorite, then members really had no collective preference. Even in this scenario, simulations showed that non-slate nominees were not completely shut out of the final ballot.


16. How do the results of this system compare to the results under the current nomination system?
Statistical tests showed that this system duplicates historical results under the present system in about (conservatively) 4.5 out of 5 candidates. In the absence of slates, it generally gives identical results to the current nomination system.


17. I think we should just increase the number of nomination slots on the final ballot to a larger number (for example, 6), and decrease the number of slots a member can nominate to a smaller number (for example, 4). Wouldn't that be simpler and easier?
Unfortunately, this simply means that the slate with the largest number of supporters will receive four of the nominations and the slate with the next largest number of supporters will receive the remaining two. It doesn't solve the problem of slates depriving non-slate nominees of the opportunity to compete in the final Award vote. In general, we want fandom to nominate as many nominees as they feel are Hugo-worthy, since under E Pluribus Hugo there is no strategic reason not to do so. For this reason, most of the designers of this system would prefer that members not be limited to nominating only four nominees. Keep in mind, however, that E Pluribus Hugo will work with this (or most any other) change as well, so one does not preclude the other.


18. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
In spite of some claims on the Internet, there has never been a small group of people serving as actual gatekeepers to the Hugo Awards. Establishing such a review committee would validate those claims. The fairness of a committee’s decisions would be subject to constant questions, and the prestige of the Hugo Awards would be tarnished.


19. I think we should use [insert other mathematical voting system].
We considered essentially every applicable type of voting system currently in the literature, guided by two experts in the field. It should be kept in mind, however, that the goals and requirements for choosing a set of representatives in a political situation are different from those for choosing a set of Hugo finalists. Some of these systems do, in fact, have positive properties that speak for them. None of them were as simple or as intuitive as E Pluribus Hugo, yet E Pluribus Hugo meets all of the stated goals for a Hugo nomination system.


20. Won’t E Pluribus Hugo be complicated to code and implement?
One of our non-experts coded a full simulator for the system in a matter of days. A full web-based app would not be much more difficult to handle.


21. Isn't this change just designed to keep certain people out of the Hugo nomination process?
No. E Pluribus Hugo is designed to fix a flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the Hugo Awards is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.

Although it is true that the discussions for this system were hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resists bloc or slate nominating, treats all individual nominators equally, and allows a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.

#735 ::: David Harmon ::: (view all by) ::: June 14, 2015, 10:39 PM:

I almost hate to toss another one in, but how about:

As less-supported nominees are eliminated, the total points for remaining nominees increase.
-->
As less-supported nominees are eliminated, their points are transferred to the remaining nominees.

This avoids the implication of the "total points" increasing, when in fact they are only coalescing around the survivors (but I wouldn't use the word "coalescing" in the FAQ, it's a stop-and-think word.)

#736 ::: Soon Lee ::: (view all by) ::: June 14, 2015, 10:42 PM:

Do we need a period after "Rogers Cadenhead"?

I would be happy for this version to be submitted.

#737 ::: Keith "Kilo" Watt ::: (view all by) ::: June 15, 2015, 02:00 AM:

Soon Lee@736:

It's not a complete sentence, but I don't think it has to be, so I think we're fine without the period.

David@735:

The point total (we should probably change to this term to be consistent) does actually increase for each nominee, so maybe we should say it that way. Saying "transferred to the remaining nominees" makes me think that you can choose which nominee to transfer the points to. Would adding "each" fix the issue for you?

Kilo

#738 ::: David Dyer-Bennet ::: (view all by) ::: June 15, 2015, 03:39 AM:

David@735: "points are transferred" suggests a simple shift, whereas what's really happening is that the points on each ballot are redistributed among the remaining nominees on that ballot. So how about just using that term -- "redistributed among"?

Keith@737: No it doesn't. It doesn't decrease for any nominee, but when a nominee is eliminated, only other nominees on the same ballots as this one see an increase in points (not all nominees).

#739 ::: David Harmon ::: (view all by) ::: June 15, 2015, 07:19 AM:

#737-738:

Hmm. Kilo: ISTM Adding "each" wouldn't fix the (valid) issue you point out. David: "redistributed among" would be equivalent to my attempt. But I think we can combine the ideas, maybe with a vocabulary simplification too. How about:

As each least-supported nominee is eliminated, its points go to the remaining nominees from its ballots.

That still doesn't cover ballots which get completely eliminated, but I think that's OK.

#740 ::: Keith "Kilo" Watt ::: (view all by) ::: June 15, 2015, 01:23 PM:

DavidDB@738:

Unless a nominee appears on nothing but bullet-voted ballots (which is highly unlikely), then all surviving nominee's points -will- increase over time (which is what I was trying to say in that phrase). But it's true that they won't increase each -round-, which is what I think you're saying.

What if we made that explicit:

"After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed among their remaining nominees. So, if there were four nominees getting 1/4 point each, there are now three nominees getting 1/3 point each, and so on. All points are re-calculated for all nominees. As rounds progress, the point total for remaining nominees will increase over time as less-supported nominees are eliminated."

Kilo

#741 ::: David Harmon ::: (view all by) ::: June 15, 2015, 01:59 PM:

Keith #740: The problem here is that "the point total for remaining nominees" is ambiguous. You appear to be reading it as "each nominee's point total", while to me it consistently reads as "the total points of all remaining nominees", which clearly is not increasing. (In fact, it decreases as ballots get eliminated with the last of their entries.)

#742 ::: Keith "Kilo" Watt ::: (view all by) ::: June 15, 2015, 02:10 PM:

David@741:

Yes, that's exactly what I mean: "each nominee's point total". The total points in the election as a whole is pretty much irrelevant. I guess that's why I don't see that it could be interpreted that way. The fact that you do, though, says we should probably clear it up.

Why don't we put the "each" that I suggested originally back in:

"After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed among their remaining nominees. So, if there were four nominees getting 1/4 point each, there are now three nominees getting 1/3 point each, and so on. All points are re-calculated for all nominees. As rounds progress, the point total for each of the remaining nominees will increase over time as less-supported nominees are eliminated."

#743 ::: Cassy B. ::: (view all by) ::: June 15, 2015, 02:14 PM:

@740 and @741, How about making the following (bolded) changes?

"After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed equally among their remaining nominees on that ballot. So, if a ballot had four nominees getting 1/4 point each, it now has three nominees getting 1/3 point each, and so on. All points are re-calculated for all nominees. As rounds progress, the point total for each remaining nominee will increase over time as less-supported nominees are eliminated."

Also, I'd rather thought we were eliminating the word "nominee", except for the very final, confirmed ballot? Or am I misremembering?

#744 ::: Keith "Kilo" Watt ::: (view all by) ::: June 15, 2015, 02:36 PM:

Cassy@743:

I like your changes. Does that work for others?

We were to change "works" to "nominees" and call nominees that make the final ballot "finalists".

K

#745 ::: Soon Lee ::: (view all by) ::: June 15, 2015, 03:08 PM:

Yes, the Cassy@743 version is clearer.

#746 ::: Jameson Quinn ::: (view all by) ::: June 15, 2015, 05:44 PM:

I'm good with @743.

#747 ::: David Harmon ::: (view all by) ::: June 15, 2015, 05:48 PM:

Cassy #743: That looks good, but with a quibble: "Those ballots that had the eliminated nominee then have their one point redistributed equally among their remaining nominees on that ballot" is betwixt and between; it should either be:
"... among their remaining nominees." OR "... among the remaining nominees on that ballot." I would prefer the latter.

#748 ::: Cassy B. ::: (view all by) ::: June 15, 2015, 05:57 PM:

David Harmon @747, you're quite right. I missed that when I made the change to clarify that it was per-ballot. Good catch.

#749 ::: Cassy B. ::: (view all by) ::: June 15, 2015, 05:57 PM:

David Harmon @747, you're quite right. I missed that when I made the change to clarify that it was per-ballot. Good catch.

#750 ::: Keith "Kilo" Watt ::: (view all by) ::: June 15, 2015, 11:23 PM:

[Proposal is unchanged; reposted here for convenience...]

Short Title: E Pluribus Hugo (Out of the Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations), section 3.9 (Notification and Acceptance), and section 3.11 (Tallying of Votes) as follows:

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the process described in section 3.A.

Insert new section 3.A after Section 3.8 as follows:
Section 3.A: Finalist Selection Process

3.A.1: For each category, the finalist selection process shall be conducted as elimination rounds consisting of three phases:
(1) Calculation Phase: First, the total number of nominations (the number of ballots on which each nominee appears) from all eligible ballots shall be tallied for each remaining nominee. Next, a single “point” shall be assigned to each nomination ballot. That point shall be divided equally among all remaining nominees on that ballot. Finally, all points from all nomination ballots shall be totaled for each nominee in that category. These two numbers, point total and number of nominations, shall be used in the Selection and Elimination Phases.
(2) Selection Phase: The two nominees with the lowest point totals shall be selected for comparison in the Elimination Phase. (See 3.A.3 for ties.)
(3) Elimination Phase: Nominees chosen in the Selection Phase shall be compared, and the nominee with the fewest number of nominations shall be eliminated and removed from all ballots for the Calculation Phase of all subsequent rounds. (See 3.A.3 for ties.)

3.A.2: The phases described in 3.A.1 are repeated in order for each category until the number of finalists specified in 3.8.1 remain. If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then instead no nominees will be eliminated during that round, and all remaining nominees shall appear on the final ballot, extending it if necessary.

3.A.3: Ties shall be handled as described below:
(1) During the Selection Phase, if two or more nominees are tied for the lowest point total, all such nominees shall be selected for the Elimination Phase.
(2) During the Selection Phase, if one nominee has the lowest point total and two or more nominees are tied for the second-lowest point total, then all such nominees shall be selected for the Elimination Phase.
(3) During the Elimination Phase, if two or more nominees are tied for the fewest number of nominations, the nominee with the lowest point total at that round shall be eliminated.
(4) During the Elimination Phase, if two or more nominees are tied for both fewest number of nominations and lowest point total, then all such nominees tied at that round shall be eliminated.

3.A.4: After the initial Award ballot is generated, if any finalist(s) are removed for any reason, the finalist selection process shall be rerun as though the removed finalist(s) had never been nominee(s). None of the remaining original finalists who have been notified shall be removed as a result of this rerun. The new finalist(s) shall be merged with the original finalists, extending the final ballot if necessary.

Section 3.9: Notification and Acceptance.
3.9.1 Worldcon Committees shall use reasonable efforts to notify the nominees finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each nominee person notified shall be asked at that time to either accept or decline the nomination. If the nominee the person notified declines the nomination, that nominee finalist(s) shall not appear on the final ballot. The procedure for replacement of such finalist(s) is described in subsection 3.A.4.

Section 3.11: Tallying of Votes.
3.11.4: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. During the same period, the results of the last ten rounds of the finalist selection process for each category (or all the rounds if there are fewer than ten) shall also be published.

Submitted by: Keith “Kilo” Watt, Jameson Quinn, Tammy Coxen, Yoana Yotova, Joshua Kronengold, Christopher Battey, David Gallaher, Adam Tilghman, Sara Watt, CJ Cabourne, Steven Halter, P J Evans, David Goldfarb, Seth Gordon, Ginger Tansey, Steve Wright, Catherine Faber, Andy Holloway, Duncan J. Macdonald, Claudia Beach, Derry Earnshaw, Jason Skiles, Soon Lee, David Harmon, Lydia Nickerson, Abigail Sutherland, Lee Billings, Oskari Rantala, Seth Breidbart, Chris Suslowicz, Fragano Ledgister, Lori Coulson, Jeffry Herman, Mark Shier, Buddha Buck, Lenore Jean Jones, David Langford, Christopher Hatton, Rogers Cadenhead

Commentary: “E Pluribus Hugo” – the name of this proposal is also its goal: recognizing the many opinions within fandom as to what nominees might be worthy of the Hugo. It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members: They still list the nominees, unranked, they feel are Hugo-worthy (up to the maximum permitted) in any categories they choose, just as they have in the past. In the past, we have counted the number of nominations each nominee received, and the top five nominees were put on the final ballot. However, because SF fandom typically nominates a variety of different nominees, it was easy for an organized slate to make it so that no other nominees made the final ballot.

Using this system, fandom isn't penalized for nominating a wide variety of nominees. If you nominate something that ends up not having a chance to make the final ballot, then your remaining choices automatically get more of your support instead of just being wasted. In other words, you can safely nominate anything you feel is Hugo-worthy. If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot. In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which finalists should be voted on to be the Hugo winner.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no group of members – of any sort, minority or majority – can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Awards.

#751 ::: Keith "Kilo" Watt ::: (view all by) ::: June 15, 2015, 11:24 PM:

FAQ’s:

1. Can you explain the system in plain language?
First, and most importantly, E Pluribus Hugo makes absolutely no changes in how members nominate (no ranking or complex strategies are required). The new system changes only the way nominations are tallied in order to create a more robust measure of which nominees have the broadest and deepest support among WSFS members. The final Hugo voting system, which actually chooses the winner, is unchanged.

The new system narrows down Worldcon members' nominations by knocking out the least supported works in repeated elimination rounds until only five (under current rules) finalists remain. Here are the basic steps, as applied to one category:

a. All the nominations for all nominees (that is, the number of ballots on which each nominee appears) are added up, just like in the current system. This number, called "number of nominations" in the proposal, is one of the criteria used to determine the final nominees with the broadest support.

b. Next, one point is divided equally among all the nominees in a category on each member's ballot. If there are two nominees on a ballot in a category, each gets 1/2 point; if there are three nominees, each gets 1/3 point, and so on.

c. All the points for each nominee from all nomination ballots are then added together. This number, called "point total" in the proposal, is the second of the two criteria used to determine the final nominees with the broadest support.

d. The nominees with the lowest point totals are compared to each other. The nominee with the fewest number of nominations is eliminated, because it is the nominee with the least support.

e. After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed equally among the remaining nominees on that ballot. So, if a ballot had four nominees getting 1/4 point each, it now has three nominees getting 1/3 point each, and so on. All points are re-calculated for all nominees. As rounds progress, the point total for each remaining nominee will increase over time as less-supported nominees are eliminated.

This process is repeated until the designated number of finalists (five in the current version of the constitution) remains.


2. E Pluribus Hugo?
The designers felt this name accurately reflects what the Hugo nomination system should be trying to achieve. Fandom is the Many. Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in science fiction and fantasy. This is a defining characteristic of fandom, and this system embraces it. Out of all of these many tastes in nominees, this system allows a single set of nominations to slowly emerge as the most popular candidates for the Hugo Awards: “Out of the Many, a Hugo.”


3. But surely that’s not what the system is formally called?
Okay, sure. Formally, in the context of election theory, this system would be called “single divisible vote with least popular elimination” or SDV-LPE for short. E Pluribus Hugo is both cooler and geekier, as everything is better with a little Latin.


4. Why are there Selection and Elimination Phases? Couldn’t you simplify the system by doing it all in one step per elimination?
The Selection Phase looks for nominees with the lowest point total, and the Elimination Phase eliminates the one with the fewest number of nominations. Looking at the two different measures of popularity adds robustness to the system. There are strategies to manipulate each of these two measurements, but it is very difficult to manipulate both in the same direction. A strategy that focuses on number of nominations for the Elimination Phase (such as bloc voting) will be stymied by the Selection Phase; a strategy that focuses on points for the Selection Phase (such as nominating only one nominee) will turn out to be counterproductive in the Elimination Phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.


5. Why re-run the system if someone declines a nomination? Why not just take the nominee that was eliminated in the next to last round?
We considered that as an option, including analyzing and simulating the effects on the vote either way. We discovered only one major difference in the results: If the declined nomination is from a slate, then when taking the “6th place” nominee, the declined nomination will tend to be replaced by a non-slate nominee. With our system of re-running the vote, if a slate nominee makes the final ballot and its nomination is declined, that nomination will tend to be replaced by another nominee from the slate. Either way works, but as noted above, our goal is to avoid shutting out any section of fandom from the final ballot. The results of re-running the calculation, seem more in line with that philosophy.


6. Shouldn’t I just nominate one work if I want to give it the best chance to win?
The system was crafted specifically to make most strategies irrelevant. For example, if you wanted to support a single nominee you feel strongly about, and there isn't anything else you feel is Hugo-worthy, you can do that. On the other hand, if you nominate four other things that you feel might be Hugo-worthy, you aren't hurting your favorite. This is because if your other nominees get eliminated, then your full support will go to the remaining nominee, just as if you had never nominated the others. If enough people agree with your other four, some of them might make it, too. If enough people don't agree with your favorite, there's nothing you can do to get it on the final ballot, because you will never have more than one point. In general, the best strategy is simple: nominate as many nominees as you feel are worthy.


7. Isn't it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws -- that is, it's impossible to create a voting system in which strategy never makes a difference. The objective therefore is to choose a system whose flaws are not an issue to the election at hand. In this system, "strategic voting" is technically possible, but extremely risky: There's no good way to know ahead of time whether a given strategy will help, or backfire. Here's a specific example:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. Suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all could submit nomination ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its number of nominations compared in the Elimination Phase. (You have no way to increase the number of nominations that Castle receives.) If you push it up to 5th place, it can still be eliminated by Tentacles, if Castle has fewer nominations. If you could somehow push it all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters so precisely before the election. Maybe Rocket isn’t so popular after all, and it needs your nominations to get over the top. Maybe too many people will assume Rocket is safe, and it loses because they left it off their ballots. Maybe Tentacles has a broad enough base of support that your nominations for Castle aren’t enough to eliminate it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire, and only one, highly-specific way for it to work.

Nominating what you think is Hugo-worthy really is your best strategy.


8. What are E Pluribus Hugo’s flaws?
In rare cases, it is possible that eliminating both members of a tie could change the final ballot slightly from what it would be if the tie were broken. The change is usually in the least popular of the finalists, and requires the two nominees in question to be very close in popularity. Hundreds of simulations were run using real and created data sets to help decide which tie-breaking methods would maintain our goal of supporting the wide range of opinions within fandom. It turned out that all the simulations showed that there was almost no difference in outcomes, no matter how we broke ties. We have chosen, therefore, to break ties in a manner consistent with section 6.4 of the Worldcon constitution. There are a number of additional tie breakers that could be used if it were deemed necessary in the future; however, simulations that have been run by the designers of this system show that it really shouldn’t be required.


9. What are E Pluribus Hugo’s benefits?
Simply put, it reduces the power of bloc voting without eliminating the chance that nominees that do appear on slates will make it to the final ballot. Conversely, it makes it very difficult for slates to prevent non-slate nominees from appearing on the ballot.


10. How does this system eliminate slate or bloc voting?
It doesn't, exactly, nor should a nominee be automatically eliminated just because it appears on a slate. On the other hand, any slate which nominates a full set of five nominees will find that each of its nominations only count 1/5 as much. With “non-slate” nominating, some of your nominees will be slowly eliminated, so your remaining nominees get more and more of your support. Since slate nominees tend to live or die together, they tend to eliminate each other until, in general, only one slate nominee remains. With a large enough support behind the slate (five times as much), the slate may still sweep a category; however, if that many voters support the slate, they arguably deserve to win, and no fair and unbiased system of nomination will prevent that. The answer in that case is, simply, to increase the general pool of voters. Regardless, with E Pluribus Hugo, slates will never receive a disproportionate share of the final ballot, as occurred in the 2015 Hugos.


11. Couldn't supporters of slates just recommend a single nominee for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate has enough supporters, that is certainly a viable possibility – it’s also completely fair. It does not force all other nominees off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn't a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a nominee deserves a Hugo nomination, then it should, in fact, appear on the final ballot.


12. What if there are multiple slates (slate wars, “parties”, etc.)?
As with a single slate, the nominees on each slate will tend to eliminate each other until only the most popular one remains. The end result is that even multiple slates are unable to sweep the nominations.


13. What happens if a broadly popular nominee is nominated by a group of unrelated people?
If it is broadly popular, the system will still select that nominee for the final ballot.


14. What happens if a broadly popular nominee also appears on a slate?
If the nominee garners support from individuals, then the system will select that nominee for the final ballot, even if it is on a slate. In general, slates neither help nor hurt any given nominee.


15. What happens when there are a lot of nominees with no obvious favorites and nomination slates are introduced?
Simulations of this scenario showed that slate nominees did receive a larger proportion of nomination slots than they did otherwise. This is a fair and valid result: If there were no overall favorite, then members really had no collective preference. Even in this scenario, simulations showed that non-slate nominees were not completely shut out of the final ballot.


16. How do the results of this system compare to the results under the current nomination system?
Statistical tests showed that this system duplicates historical results under the present system in about (conservatively) 4.5 out of 5 candidates. In the absence of slates, it generally gives identical results to the current nomination system.


17. I think we should just increase the number of nomination slots on the final ballot to a larger number (for example, 6), and decrease the number of slots a member can nominate to a smaller number (for example, 4). Wouldn't that be simpler and easier?
Unfortunately, this simply means that the slate with the largest number of supporters will receive four of the nominations and the slate with the next largest number of supporters will receive the remaining two. It doesn't solve the problem of slates depriving non-slate nominees of the opportunity to compete in the final Award vote. In general, we want fandom to nominate as many nominees as they feel are Hugo-worthy, since under E Pluribus Hugo there is no strategic reason not to do so. For this reason, most of the designers of this system would prefer that members not be limited to nominating only four nominees. Keep in mind, however, that E Pluribus Hugo will work with this (or most any other) change as well, so one does not preclude the other.


18. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
In spite of some claims on the Internet, there has never been a small group of people serving as actual gatekeepers to the Hugo Awards. Establishing such a review committee would validate those claims. The fairness of a committee’s decisions would be subject to constant questions, and the prestige of the Hugo Awards would be tarnished.


19. I think we should use [insert other mathematical voting system].
We considered essentially every applicable type of voting system currently in the literature, guided by two experts in the field. It should be kept in mind, however, that the goals and requirements for choosing a set of representatives in a political situation are different from those for choosing a set of Hugo finalists. Some of these systems do, in fact, have positive properties that speak for them. None of them were as simple or as intuitive as E Pluribus Hugo, yet E Pluribus Hugo meets all of the stated goals for a Hugo nomination system.


20. Won’t E Pluribus Hugo be complicated to code and implement?
One of our non-experts coded a full simulator for the system in a matter of days. A full web-based app would not be much more difficult to handle.


21. Isn't this change just designed to keep certain people out of the Hugo nomination process?
No. E Pluribus Hugo is designed to fix a flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the Hugo Awards is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.

Although it is true that the discussions for this system were hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resists bloc or slate nominating, treats all individual nominators equally, and allows a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.

#752 ::: Doire ::: (view all by) ::: June 16, 2015, 07:57 AM:
In 3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the process described in section 3.A.
Should "nominees" be "finalists" in this section?

I've looked back but may have missed any discussion of this. Sorry if I'm rehashing old points.

#753 ::: Cadbury Moose ::: (view all by) ::: June 16, 2015, 10:31 AM:

Doire @ #752

I think "nominees" is correct at this point, they only become "finalists" once they have been placed on the final award ballot, i.e. at the end of this step.

#754 ::: Jameson Quinn ::: (view all by) ::: June 16, 2015, 10:33 AM:

@754: If you wanted to be explicit, you could say "...the five finalists, which are the five eligible nominees...". In other words, this is the section which turns them from nominees into finalists, so it's hard to know which to call them in the moment of transition. I'm sure there are other wordsmiths here who can thread that needle better than I can.

/me imagines a blacksmith threading a needle...

#755 ::: David Dyer-Bennet ::: (view all by) ::: June 16, 2015, 03:02 PM:

Just to be sure I've got it right -- is this the way we're supposed to be using the terms?

nominee: any work appearing on a Hugo Awards nominating ballot.

finalist: any work appearing on the final Hugo Awards ballot.

#756 ::: David Dyer-Bennet ::: (view all by) ::: June 16, 2015, 03:08 PM:

743 still contains the statement "As rounds progress, the point total for each remaining nominee will increase over time as less-supported nominees are eliminated."

Yes, statistically, in real-world cases, it's likely that over the sequence of elimination rounds, the point total for each nominee will increase.

But that's different from stating it as an explaining principle. Trying to use it as an explaining principle will cause people to try to figure out how it happens in detail, and it doesn't happen in detail; it doesn't happen to each nominee each round. It only happens statistically.

It's a bad explanation that will cause confusion to people trying to figure out what's really going on.

(Yes, I've read Keith@743).

#757 ::: Keith "Kilo" Watt ::: (view all by) ::: June 16, 2015, 04:20 PM:

DavidDB@756:

Well, I may just be too close to it. To me, it looks extremely obvious what's going on, especially if you put it in the context of the rest of the steps.

Am I correct that your problem is that in some rounds a particular work may not increase its point total because no works it shared a ballot with happened to be eliminated that round? I just don't see this as being confusing, but if you can propose other language, we can look at it. My fear is that we're going to start stringing clauses and exceptions together that -will- make it confusing.

What's written is correct, so long as you aren't reading it in the sense that the points increase every single round. Over time they do increase, though I'll stipulate a very bizarre exception when you have a work that was -only- bullet voted in the whole set of ballots.

Kilo

#758 ::: Cally Soukup ::: (view all by) ::: June 16, 2015, 05:09 PM:

I don't see a problem with that wording, either. This is, after all, an explanation of how it actually works, not a dissertation on every single possible edge case. Anyone who wants to know how a particular edge case works out can actually run through the steps themselves, and find out.

#759 ::: David Dyer-Bennet ::: (view all by) ::: June 16, 2015, 05:12 PM:

Keith@757: You're basically right about what I object to. But to me it's a black-and-white, blatant, thing. This statement is definitely not true in the simple obvious interpretation. As a statistical abstraction, okay, it's true that the system works by building support for some candidates as others are eliminated; but it's not the case that support for anything automatically or inherently increases as other nominees are eliminated. And the statement is tucked into a detailed step-by-step list of what's done; it's in a detail-centered context, a high-level statistical interpretation is by no means the obvious one to apply.

And, for people who take it this way, if they do attempt to run a simple example in their head -- they'll quickly see it's not true. So they'll be left wondering what they've misunderstood, and it can begin the process of their seeing this method as complex and confusing.

(By the way, this is FAQ 1.e; took me a while to find it when I wanted to see it in context just now.)

Now, for suggested alternate wording, how about:

e. After a nominee is eliminated it is removed from all ballots that it appeared on. Those ballots that had the eliminated nominee then have their one point redistributed among their remaining nominees. So, if a ballot had four nominees (getting 1/4 point each), there are now three nominees (getting 1/3 point each), and so on.

This process is repeated until the designated number of finalists (five in the current version of the constitution) remains.

(Looking too closely is always dangerous. I fixed a tense problem or two, and I think clarified the recalculation step. And simply eliminated the statement that started the discussion; I don't think it actually adds anything. And I removed "All points are re-calculated for all nominees" because it conflicts with the earlier statement "Those ballots that had the eliminated nominee then have...". Whether you recalculate all, or only those that have changed, is an implementation detail, not a part of the process specification.)

Anyway, not much sign that anybody except Keith and I care much about this detail. That probably means it doesn't matter much; small numbers perhaps best served by one version or the other, most people not bothered either way.

#760 ::: Annie Y ::: (view all by) ::: June 16, 2015, 06:11 PM:

Taking a sentence out of context will always make it unclear. It is part of the steps and if you read the whole thing, it is clear how points get redistributed. If we try to precise each sentence to the point where it is absolutely clear in every single situation, we will end up with something 3 times longer than it is now...

I think the current wording is clear enough...

#761 ::: P J Evans ::: (view all by) ::: June 16, 2015, 08:50 PM:

I agree that the current wording is clear enough. It isn't misleading, and it isn't confusing - unless you want to be misled or confused.

#762 ::: David Harmon ::: (view all by) ::: June 16, 2015, 11:22 PM:

David Dyer-Bennet #759:

I'd say the current wording is fairly clear. The statement about "points increasing over time" is not meant to be a mathematical theorem, but a general statement of how the numbers develop over time. And the shift to the general is signalled by "as rounds progress".

That said, I note the statement isn't really necessary, and your version makes the form of 1.e closer the pattern set by 1.a through 1.d (One sentence of summary, one of explanation). Even combining sentences, I think we still end up needing an extra one, or at least a parenthesis:

After a nominee is eliminated, it is removed from all ballots that it appeared on, and those ballots get their point redistributed among their remaining nominees. E.g., if a ballot had four nominees with 1/4 point each, it now has three nominees with 1/3 point each (If no nominees are left, the ballot is eliminated and its point is effectively lost.)

#763 ::: David Harmon ::: (view all by) ::: June 16, 2015, 11:27 PM:

Ohnosecond: Muffed the punctuation on the trailing clause, easier to repost fixed:

After a nominee is eliminated, it is removed from all ballots that it appeared on, and those ballots get their point redistributed among their remaining nominees. E.g., if a ballot had four nominees with 1/4 point each, it now has three nominees with 1/3 point each (if no nominees are left, the ballot is eliminated and its point is effectively lost).

#764 ::: Keith "Kilo" Watt ::: (view all by) ::: June 16, 2015, 11:43 PM:

David@763:

I don't think the parenthesis is really necessary, since I think it's pretty obvious what happens if all your nominees are eliminated. I still think there's value in mentioning that point totals increase over time, as I don't think we mention that explicitly anywhere else. The increase is referenced in #7, so I fear that might be ambiguous without a previous introduction to the idea. And I still just don't see the harm in including it here.

Kilo

#765 ::: David Dyer-Bennet ::: (view all by) ::: June 17, 2015, 12:23 AM:

P.J. Evans@761: Thanks for the personal jab!

#766 ::: David Dyer-Bennet ::: (view all by) ::: June 17, 2015, 12:25 AM:

Annie@760: I'm taking it in context. It appears as part of the detailed step-by-step description of what's going on. In that context, it is false.

There is an out-of-context interpretation which is sort-of true, sure. Out of context.

#767 ::: Soon Lee ::: (view all by) ::: June 17, 2015, 12:27 AM:

[FAQ1]@751

I think the current wording is fine but I have the advantage of familiarity with it. Perhaps it's time to solicit the opinion of a few people who haven't seen the proposal before, to get a fresh perspective?

#768 ::: David Dyer-Bennet ::: (view all by) ::: June 17, 2015, 01:09 AM:

FAQ 1.e: So, I suggested just dropping the statement about increasing point totals. Shorter and simpler certainly, and not misleading to me. But that doesn't seem popular. (I hope the other improvements from the same proposal don't get lost in the shuffle, by the way.)

I don't get the issue with needing to tell people that numbers sometimes go up. I don't see how it's relevant to understanding the tallying procedure. Is this somehow bearing on something deep? There's no magic boundary we're chasing (the way there is for "majority" in the final ballot counting for example). Does it represent "progress" in people's heads somehow? I'm not getting the point; it seems kind of a weird statement to insert into the description. It seems disconnected, basically a non-sequitur. (And that, I guess, may be why I've examined it closely and found that, to me and in the context it occurs in, it's false. You look more closely at things that seem strange.)

Is it useful, surprising, informative, reassuring, or something, to people to be told that sometimes the point totals for a nominee increase? I mean, given that the essence of the procedure is elimination of nominees and re-distribution of points, the points clearly change each and every round. Is it specifically increase that it's valuable to people to have pointed out, or is it the change that's really being communicated?

(Trying to understand, to either find another way to do it, or perhaps convince me that it's worth doing this way.)

#769 ::: David Harmon ::: (view all by) ::: June 17, 2015, 01:12 AM:

Keith "Kilo" Watt #764: Re: For the "obvious" point, having the reciprocal in there just makes me itch to specify the zero case.

It's too evening my time for me to write it properly, but perhaps the shifting of points among the ballots increase could be described more precisely in a new 1.f? Something like "As nominees are eliminated, some ballots will be eliminated entirely, but the survivors will concentrate their points on fewer nominees." ?

#770 ::: P J Evans ::: (view all by) ::: June 17, 2015, 01:37 AM:

766/768
Possibly it's helpful to remember that FAQs are not detailed design specs, and aren't supposed to be.

#771 ::: David Harmon ::: (view all by) ::: June 17, 2015, 06:13 AM:

P J Evans #770: It's still going to be going to the Business Meeting. The trick is to find the proper levels of detail and precision.

#772 ::: Buddha Buck ::: (view all by) ::: June 17, 2015, 07:30 AM:

The points for any individual nominee are nondecreasing over time. They may not increase, but they never decrease. If you don't like "increase over time" because it's not correct, perhaps consider nondecreasing.


#773 ::: David Harmon ::: (view all by) ::: June 17, 2015, 09:50 AM:

A try with an additional entry. The new one separates out the increase issue; the question is whether we want to add it.

1e: After a nominee is eliminated, it is removed from the ballots that it appeared on, each of which gets their point redistributed among its remaining nominees (until none are left, when the ballot is eliminated). E.g., if a ballot had four nominees with 1/4 point each, it now has three nominees with 1/3 point each.

1f: Over successive rounds, each surviving nominee's total point score can increase as its ballots get shorter, but can't decrease short of elimination. Eventually, the finalists will have all the points from those ballots which nominated them.

#774 ::: David Harmon ::: (view all by) ::: June 17, 2015, 09:53 AM:

An alternate option for "... those ballots which nominated them" would be "... all remaining ballots".

#775 ::: David Harmon ::: (view all by) ::: June 17, 2015, 09:59 AM:

Sorry for multiposting; Noticed an echo, decided to rework that sentence.

1f: Over successive rounds, a surviving nominee's total point score can't decrease, but can increase as its ballots get shorter. Eventually, the finalists will have all the points from all remaining ballots.

#776 ::: David Dyer-Bennet ::: (view all by) ::: June 17, 2015, 11:19 AM:

David@773: That works for me. It eliminates the bit that stuck me as false, plus it's clearly a statement about trends over the course of the tallying process rather than possibly being a statement about a single round.

Thanks to everybody for putting up with my being stuck on this issue!

(Yes, I'm aware people haven't agreed to adopt the version I like better there. But now there is one!)

#777 ::: Keith "Kilo" Watt ::: (view all by) ::: June 17, 2015, 01:08 PM:

David@773:

I can live with that, if we can make just a couple of tweaks. I really don't think the parenthesis is necessary, and my first response was, "Wait, my ballot gets eliminated??" It doesn't of course, all my -nominees- got eliminated. I don't think the parenthesis is necessary at all, though.

Secondly, instead of a 1f, why don't we just add it to the closing line of that section. When separated out like that, it's not really a "step" in the process. Also, "ballots getting shorter" is kind of an odd image for me (I'm picturing someone with scissors), so I'd like to reword that just a bit.

So, what about something like this (I also made some small terminology changes to match the rest of the proposal):


"1e: After a nominee is eliminated, it is removed from the ballots that it appeared on, each of which gets its point redistributed among its remaining nominees. For example, if a ballot had four nominees with 1/4 point each, it now has three nominees with 1/3 point each, and so on.

Over successive rounds, a surviving nominee's point total can't decrease, but it can increase as the nominees it shares a ballot with are eliminated. Eventually, the finalists will have all the points from all remaining ballots. This process is repeated until the designated number of finalists (five in the current version of the constitution) remains."


What do you think?

Kilo

#778 ::: David Dyer-Bennet ::: (view all by) ::: June 17, 2015, 02:10 PM:

Kilo@777: Yes. That works okay for me.

#779 ::: David Harmon ::: (view all by) ::: June 17, 2015, 02:39 PM:

Keith "Kilo" Watt #777: Okay, that looks good.

#780 ::: Soon Lee ::: (view all by) ::: June 17, 2015, 03:05 PM:

#777: Good for me too.

#781 ::: David Dyer-Bennet ::: (view all by) ::: June 17, 2015, 03:43 PM:

Oops, didn't mean to be that emphatic (triple posted!). First three times posting here from my phone!

#783 ::: Doire ::: (view all by) ::: June 17, 2015, 06:30 PM:

I asked someone to have a look over the proposal and they said, "I've read through those and I couldn't see any typos, and I can confirm that once you take them together, the rules and the FAQ make sense. It took me a couple of readings of the rules to understand them, but I'd expect that."

So, at least one uninvolved reader can follow the proposal, which is a good sign.

#784 ::: Tim Illingworth ::: (view all by) ::: June 17, 2015, 08:13 PM:

Minor nitpick: "Insert new section 3.A after Section 3.8 as follows:" should be in plain text rather than italics, since it's not part of the text to be inserted.

#785 ::: Keith "Kilo" Watt ::: (view all by) ::: June 17, 2015, 10:42 PM:

Short Title: E Pluribus Hugo (Out of the Many, a Hugo)
Moved, to amend section 3.8 (Tallying of Nominations), section 3.9 (Notification and Acceptance), and section 3.11 (Tallying of Votes) as follows:

Section 3.8: Tallying of Nominations.
3.8.1: Except as provided below, the final Award ballots shall list in each category the five eligible nominees receiving the most nominations. If there is a tie including fifth place, all the tied eligible nominees shall be listed. determined by the process described in section 3.A.

Insert new section 3.A after Section 3.8 as follows:
Section 3.A: Finalist Selection Process

3.A.1: For each category, the finalist selection process shall be conducted as elimination rounds consisting of three phases:
(1) Calculation Phase: First, the total number of nominations (the number of ballots on which each nominee appears) from all eligible ballots shall be tallied for each remaining nominee. Next, a single “point” shall be assigned to each nomination ballot. That point shall be divided equally among all remaining nominees on that ballot. Finally, all points from all nomination ballots shall be totaled for each nominee in that category. These two numbers, point total and number of nominations, shall be used in the Selection and Elimination Phases.
(2) Selection Phase: The two nominees with the lowest point totals shall be selected for comparison in the Elimination Phase. (See 3.A.3 for ties.)
(3) Elimination Phase: Nominees chosen in the Selection Phase shall be compared, and the nominee with the fewest number of nominations shall be eliminated and removed from all ballots for the Calculation Phase of all subsequent rounds. (See 3.A.3 for ties.)

3.A.2: The phases described in 3.A.1 are repeated in order for each category until the number of finalists specified in 3.8.1 remain. If elimination would reduce the number of finalists to fewer than the number specified in section 3.8.1, then instead no nominees will be eliminated during that round, and all remaining nominees shall appear on the final ballot, extending it if necessary.

3.A.3: Ties shall be handled as described below:
(1) During the Selection Phase, if two or more nominees are tied for the lowest point total, all such nominees shall be selected for the Elimination Phase.
(2) During the Selection Phase, if one nominee has the lowest point total and two or more nominees are tied for the second-lowest point total, then all such nominees shall be selected for the Elimination Phase.
(3) During the Elimination Phase, if two or more nominees are tied for the fewest number of nominations, the nominee with the lowest point total at that round shall be eliminated.
(4) During the Elimination Phase, if two or more nominees are tied for both fewest number of nominations and lowest point total, then all such nominees tied at that round shall be eliminated.

3.A.4: After the initial Award ballot is generated, if any finalist(s) are removed for any reason, the finalist selection process shall be rerun as though the removed finalist(s) had never been nominee(s). None of the remaining original finalists who have been notified shall be removed as a result of this rerun. The new finalist(s) shall be merged with the original finalists, extending the final ballot if necessary.

Section 3.9: Notification and Acceptance.
3.9.1 Worldcon Committees shall use reasonable efforts to notify the nominees finalists, or in the case of deceased or incapacitated persons, their heirs, assigns, or legal guardians, in each category prior to the release of such information. Each nominee person notified shall be asked at that time to either accept or decline the nomination. If the nominee the person notified declines the nomination, that nominee finalist(s) shall not appear on the final ballot. The procedure for replacement of such finalist(s) is described in subsection 3.A.4.

Section 3.11: Tallying of Votes.
3.11.4: The complete numerical vote totals, including all preliminary tallies for first, second, … places, shall be made public by the Worldcon Committee within ninety (90) days after the Worldcon. During the same period the nomination voting totals shall also be published, including in each category the vote counts for at least the fifteen highest vote-getters and any other candidate receiving a number of votes equal to at least five percent (5%) of the nomination ballots cast in that category, but not including any candidate receiving fewer than five votes. During the same period, the results of the last ten rounds of the finalist selection process for each category (or all the rounds if there are fewer than ten) shall also be published.

Submitted by: Keith “Kilo” Watt, Jameson Quinn, Tammy Coxen, Yoana Yotova, Joshua Kronengold, Christopher Battey, David Gallaher, Adam Tilghman, Sara Watt, CJ Cabourne, Steven Halter, P J Evans, David Goldfarb, Seth Gordon, Ginger Tansey, Steve Wright, Catherine Faber, Andy Holloway, Duncan J. Macdonald, Claudia Beach, Derry Earnshaw, Jason Skiles, Soon Lee, David Harmon, Lydia Nickerson, Abigail Sutherland, Lee Billings, Oskari Rantala, Seth Breidbart, Chris Suslowicz, Fragano Ledgister, Lori Coulson, Jeffry Herman, Mark Shier, Buddha Buck, Lenore Jean Jones, David Langford, Christopher Hatton, Rogers Cadenhead

Commentary: “E Pluribus Hugo” – the name of this proposal is also its goal: recognizing the many opinions within fandom as to what nominees might be worthy of the Hugo. It is to be emphasized that this proposal does not change the nominating process from the perspective of Worldcon members: They still list the nominees, unranked, they feel are Hugo-worthy (up to the maximum permitted) in any categories they choose, just as they have in the past. In the past, we have counted the number of nominations each nominee received, and the top five nominees were put on the final ballot. However, because SF fandom typically nominates a variety of different nominees, it was easy for an organized slate to make it so that no other nominees made the final ballot.

Using this system, fandom isn't penalized for nominating a wide variety of nominees. If you nominate something that ends up not having a chance to make the final ballot, then your remaining choices automatically get more of your support instead of just being wasted. In other words, you can safely nominate anything you feel is Hugo-worthy. If enough people agree with you, it will make the final ballot. If they don't, that's okay – when that nominee is eliminated, your other choices will have a greater chance of making the final ballot. In this way, by eliminating the least popular candidates each round, fandom slowly converges to a consensus as to which finalists should be voted on to be the Hugo winner.

It is also an explicit goal of this proposal not to disenfranchise anyone. Rather, this proposal seeks to ensure that no group of members – of any sort, minority or majority – can disproportionately dominate an entire category. This system allows the broadest range of nominees that are popular with fandom-at-large to be considered for the Hugo Awards.

#786 ::: Keith "Kilo" Watt ::: (view all by) ::: June 17, 2015, 10:43 PM:

FAQ’s:

1. Can you explain the system in plain language?
First, and most importantly, E Pluribus Hugo makes absolutely no changes in how members nominate (no ranking or complex strategies are required). The new system changes only the way nominations are tallied in order to create a more robust measure of which nominees have the broadest and deepest support among WSFS members. The final Hugo voting system, which actually chooses the winner, is unchanged.

The new system narrows down Worldcon members' nominations by knocking out the least supported works in repeated elimination rounds until only five (under current rules) finalists remain. Here are the basic steps, as applied to one category:

a. All the nominations for all nominees (that is, the number of ballots on which each nominee appears) are added up, just like in the current system. This number, called "number of nominations" in the proposal, is one of the criteria used to determine the final nominees with the broadest support.

b. Next, one point is divided equally among all the nominees in a category on each member's ballot. If there are two nominees on a ballot in a category, each gets 1/2 point; if there are three nominees, each gets 1/3 point, and so on.

c. All the points for each nominee from all nomination ballots are then added together. This number, called "point total" in the proposal, is the second of the two criteria used to determine the final nominees with the broadest support.

d. The nominees with the lowest point totals are compared to each other. The nominee with the fewest number of nominations is eliminated, because it is the nominee with the least support.

e. After a nominee is eliminated, it is removed from the ballots that it appeared on, each of which gets its point redistributed among its remaining nominees. For example, if a ballot had four nominees with 1/4 point each, it now has three nominees with 1/3 point each, and so on.

Over successive rounds, a surviving nominee's point total can't decrease, but it can increase as the nominees it shares a ballot with are eliminated. Eventually, the finalists will have all the points from all remaining ballots. This process is repeated until the designated number of finalists (five in the current version of the constitution) remains.


2. E Pluribus Hugo?
The designers felt this name accurately reflects what the Hugo nomination system should be trying to achieve. Fandom is the Many. Members of fandom have a wide range of interests, tastes, and desires in what they appreciate in science fiction and fantasy. This is a defining characteristic of fandom, and this system embraces it. Out of all of these many tastes in nominees, this system allows a single set of nominations to slowly emerge as the most popular candidates for the Hugo Awards: “Out of the Many, a Hugo.”


3. But surely that’s not what the system is formally called?
Okay, sure. Formally, in the context of election theory, this system would be called “single divisible vote with least popular elimination” or SDV-LPE for short. E Pluribus Hugo is both cooler and geekier, as everything is better with a little Latin.


4. Why are there Selection and Elimination Phases? Couldn’t you simplify the system by doing it all in one step per elimination?
The Selection Phase looks for nominees with the lowest point total, and the Elimination Phase eliminates the one with the fewest number of nominations. Looking at the two different measures of popularity adds robustness to the system. There are strategies to manipulate each of these two measurements, but it is very difficult to manipulate both in the same direction. A strategy that focuses on number of nominations for the Elimination Phase (such as bloc voting) will be stymied by the Selection Phase; a strategy that focuses on points for the Selection Phase (such as nominating only one nominee) will turn out to be counterproductive in the Elimination Phase. While no voting system can be completely strategy-proof, this two-step approach makes this system as close to that as possible.


5. Why re-run the system if someone declines a nomination? Why not just take the nominee that was eliminated in the next to last round?
We considered that as an option, including analyzing and simulating the effects on the vote either way. We discovered only one major difference in the results: If the declined nomination is from a slate, then when taking the “6th place” nominee, the declined nomination will tend to be replaced by a non-slate nominee. With our system of re-running the vote, if a slate nominee makes the final ballot and its nomination is declined, that nomination will tend to be replaced by another nominee from the slate. Either way works, but as noted above, our goal is to avoid shutting out any section of fandom from the final ballot. The results of re-running the calculation, seem more in line with that philosophy.


6. Shouldn’t I just nominate one work if I want to give it the best chance to win?
The system was crafted specifically to make most strategies irrelevant. For example, if you wanted to support a single nominee you feel strongly about, and there isn't anything else you feel is Hugo-worthy, you can do that. On the other hand, if you nominate four other things that you feel might be Hugo-worthy, you aren't hurting your favorite. This is because if your other nominees get eliminated, then your full support will go to the remaining nominee, just as if you had never nominated the others. If enough people agree with your other four, some of them might make it, too. If enough people don't agree with your favorite, there's nothing you can do to get it on the final ballot, because you will never have more than one point. In general, the best strategy is simple: nominate as many nominees as you feel are worthy.


7. Isn't it true that any voting system can be gamed (or strategized, etc.)?
Yes, there is a theorem which proves that all voting systems must have inherent flaws -- that is, it's impossible to create a voting system in which strategy never makes a difference. The objective therefore is to choose a system whose flaws are not an issue to the election at hand. In this system, "strategic voting" is technically possible, but extremely risky: There's no good way to know ahead of time whether a given strategy will help, or backfire. Here's a specific example:

In theory: Suppose that you and your friends wanted to see both the novels Rocket and Castle on the final ballot, and you really don’t want to see Tentacles. Suppose that you knew that without your nominations, Rocket would be a shoo-in, Tentacles would just barely make it to fifth place, and Castle would just barely miss. In that case, you all could submit nomination ballots that only have Castle in the Best Novel category, leaving off Rocket. By doing that, and giving your undivided points to Castle, you increase its points and maybe help it to avoid having its number of nominations compared in the Elimination Phase. (You have no way to increase the number of nominations that Castle receives.) If you push it up to 5th place, it can still be eliminated by Tentacles, if Castle has fewer nominations. If you could somehow push it all the way up to 4th place, it would safely make the final ballot.

In practice: You won’t actually know the preferences of all the other Hugo voters so precisely before the election. Maybe Rocket isn’t so popular after all, and it needs your nominations to get over the top. Maybe too many people will assume Rocket is safe, and it loses because they left it off their ballots. Maybe Tentacles has a broad enough base of support that your nominations for Castle aren’t enough to eliminate it. Maybe Castle is so unpopular outside your own circle of friends that it never had a chance. Maybe you're the only Castle fan who also likes Rocket, so your strategy won't change Castle's score by enough to put it in 4th place. Half a point isn't likely to make enough of a difference to give it that much of a boost (and the cost is that you are giving up nominating any of your other favorites). There are many ways for your strategy to fail, or even backfire, and only one, highly-specific way for it to work.

Nominating what you think is Hugo-worthy really is your best strategy.


8. What are E Pluribus Hugo’s flaws?
In rare cases, it is possible that eliminating both members of a tie could change the final ballot slightly from what it would be if the tie were broken. The change is usually in the least popular of the finalists, and requires the two nominees in question to be very close in popularity. Hundreds of simulations were run using real and created data sets to help decide which tie-breaking methods would maintain our goal of supporting the wide range of opinions within fandom. It turned out that all the simulations showed that there was almost no difference in outcomes, no matter how we broke ties. We have chosen, therefore, to break ties in a manner consistent with section 6.4 of the Worldcon constitution. There are a number of additional tie breakers that could be used if it were deemed necessary in the future; however, simulations that have been run by the designers of this system show that it really shouldn’t be required.


9. What are E Pluribus Hugo’s benefits?
Simply put, it reduces the power of bloc voting without eliminating the chance that nominees that do appear on slates will make it to the final ballot. Conversely, it makes it very difficult for slates to prevent non-slate nominees from appearing on the ballot.


10. How does this system eliminate slate or bloc voting?
It doesn't, exactly, nor should a nominee be automatically eliminated just because it appears on a slate. On the other hand, any slate which nominates a full set of five nominees will find that each of its nominations only count 1/5 as much. With “non-slate” nominating, some of your nominees will be slowly eliminated, so your remaining nominees get more and more of your support. Since slate nominees tend to live or die together, they tend to eliminate each other until, in general, only one slate nominee remains. With a large enough support behind the slate (five times as much), the slate may still sweep a category; however, if that many voters support the slate, they arguably deserve to win, and no fair and unbiased system of nomination will prevent that. The answer in that case is, simply, to increase the general pool of voters. Regardless, with E Pluribus Hugo, slates will never receive a disproportionate share of the final ballot, as occurred in the 2015 Hugos.


11. Couldn't supporters of slates just recommend a single nominee for a candidate, and it will automatically appear on the final ballot?
Yes, if a slate has enough supporters, that is certainly a viable possibility – it’s also completely fair. It does not force all other nominees off of the final ballot, and the final Hugo winner is determined by the same voting process we have always had. Just appearing on the final ballot isn't a guarantee of winning a Hugo. However, if any large section of fandom strongly believes that a nominee deserves a Hugo nomination, then it should, in fact, appear on the final ballot.


12. What if there are multiple slates (slate wars, “parties”, etc.)?
As with a single slate, the nominees on each slate will tend to eliminate each other until only the most popular one remains. The end result is that even multiple slates are unable to sweep the nominations.


13. What happens if a broadly popular nominee is nominated by a group of unrelated people?
If it is broadly popular, the system will still select that nominee for the final ballot.


14. What happens if a broadly popular nominee also appears on a slate?
If the nominee garners support from individuals, then the system will select that nominee for the final ballot, even if it is on a slate. In general, slates neither help nor hurt any given nominee.


15. What happens when there are a lot of nominees with no obvious favorites and nomination slates are introduced?
Simulations of this scenario showed that slate nominees did receive a larger proportion of nomination slots than they did otherwise. This is a fair and valid result: If there were no overall favorite, then members really had no collective preference. Even in this scenario, simulations showed that non-slate nominees were not completely shut out of the final ballot.


16. How do the results of this system compare to the results under the current nomination system?
Statistical tests showed that this system duplicates historical results under the present system in about (conservatively) 4.5 out of 5 candidates. In the absence of slates, it generally gives identical results to the current nomination system.


17. I think we should just increase the number of nomination slots on the final ballot to a larger number (for example, 6), and decrease the number of slots a member can nominate to a smaller number (for example, 4). Wouldn't that be simpler and easier?
Unfortunately, this simply means that the slate with the largest number of supporters will receive four of the nominations and the slate with the next largest number of supporters will receive the remaining two. It doesn't solve the problem of slates depriving non-slate nominees of the opportunity to compete in the final Award vote. In general, we want fandom to nominate as many nominees as they feel are Hugo-worthy, since under E Pluribus Hugo there is no strategic reason not to do so. For this reason, most of the designers of this system would prefer that members not be limited to nominating only four nominees. Keep in mind, however, that E Pluribus Hugo will work with this (or most any other) change as well, so one does not preclude the other.


18. I think we should set up a committee to handle these situations as they occur. The committee would be empowered to add nomination slots or throw out slate-influenced ballots as required.
In spite of some claims on the Internet, there has never been a small group of people serving as actual gatekeepers to the Hugo Awards. Establishing such a review committee would validate those claims. The fairness of a committee’s decisions would be subject to constant questions, and the prestige of the Hugo Awards would be tarnished.


19. I think we should use [insert other mathematical voting system].
We considered essentially every applicable type of voting system currently in the literature, guided by two experts in the field. It should be kept in mind, however, that the goals and requirements for choosing a set of representatives in a political situation are different from those for choosing a set of Hugo finalists. Some of these systems do, in fact, have positive properties that speak for them. None of them were as simple or as intuitive as E Pluribus Hugo, yet E Pluribus Hugo meets all of the stated goals for a Hugo nomination system.


20. Won’t E Pluribus Hugo be complicated to code and implement?
One of our non-experts coded a full simulator for the system in a matter of days. A full web-based app would not be much more difficult to handle.


21. Isn't this change just designed to keep certain people out of the Hugo nomination process?
No. E Pluribus Hugo is designed to fix a flaw in the current Hugo nominating process, which was highlighted by the Sad and Rabid Puppies slates in 2015. This flaw allowed a small percentage of nominators to dominate the final ballot, shutting out all others from entire categories. This is a major flaw that must be fixed if the integrity of the Hugo Awards is to be maintained. E Pluribus Hugo dilutes the effect of slate or bloc nominating, and works against any group trying to control the nominations. If some hypothetical cabal were trying to maintain its dominance, they would not propose this system, as E Pluribus Hugo works against that purpose.

Although it is true that the discussions for this system were hosted at the “Making Light” blog, owned by Teresa and Patrick Nielsen Hayden, the discussions were conducted openly and inclusively. Those of us who worked on the system were very clear that our goals were to create a system that resists bloc or slate nominating, treats all individual nominators equally, and allows a consensus final ballot to emerge from all the nominations submitted. E Pluribus Hugo accomplishes those goals.

#787 ::: David Harmon ::: (view all by) ::: June 18, 2015, 12:49 PM:

Looks good to me!

#788 ::: Cally Soukup ::: (view all by) ::: June 18, 2015, 01:35 PM:

I say we submit it.

If nothing else, we'll almost certainly get feedback from people who haven't been involved in these discussions, and we'll have time to revise (or not) accordingly.

#789 ::: abi ::: (view all by) ::: June 18, 2015, 01:46 PM:

Do we want a front-page post with it and links to all the previous discussion threads? Are we there yet?

#790 ::: Jameson Quinn ::: (view all by) ::: June 18, 2015, 02:21 PM:

@789: It's Keith's call, but I vote yes.

#791 ::: Annie Y ::: (view all by) ::: June 18, 2015, 02:30 PM:

Looks good to me.

#792 ::: Cassy B. ::: (view all by) ::: June 18, 2015, 02:38 PM:

abi @789, Kilo's done the heavy lifting, so I defer to to him... but in my opinion, once he says we're good to go, a front-page post with links to all discussions would be an excellent idea, in the interests of transparency.

#793 ::: Cadbury Moose ::: (view all by) ::: June 18, 2015, 03:58 PM:

Fully approved by this moose. It's clearly written, easy to understand, and the FAQ looks like it covers anything that could sensibly be asked. (I stand ready to be proved wrong in that last regard.)

Well done everybody who contributed!

3:O)>

#794 ::: David Dyer-Bennet ::: (view all by) ::: June 18, 2015, 04:00 PM:

We asked a while ago if it was okay to include a link to the discussion here in the FAQ, and unless there was a private answer to the people actually maintaining the current document, we haven't had one. That's one thing that should be closed before we submit; but the only one I've thought of so far.

Yes, there will be responses to the submission, some of which will involve re-arguing old arguments, which will make some people tired. Possibly, though, there will be useful new thoughts, or at least observations about the explanations in the FAQ.

#795 ::: Keith "Kilo" Watt ::: (view all by) ::: June 18, 2015, 04:16 PM:

All:

I'm really liking how this looks as well. Unless there is serious objection, I think we're ready to submit. Remember, Kevin has said we can always modify up until the deadline if we really need to, but I think we're at the point where nothing but fresh eyes will lead to any substantive changes. It's time to release it to the wild.

Abi, what I'd actually like to see is a new topic for Q&A (or, hopefully, voices of support) from the community after we submit and it's published on the Sasquan site. I think it'd be helpful to have a central place dedicated to that. One question I have for the group, though: I think it still might be useful to have a place to discuss the 6th place amendment, swag/T shirts, and strategy for the actual business meeting presentation. I'm thinking that would be better as a topic separate from the community Q&A, but we've only got 200 or so posts left in this one. Is there an issue with starting two new threads?

Regarding the link to these discussions, once we publish, this will be picked up by File770, I'm sure (and I will contact Mike to make certain it will), so I don't think a link in the proposal text itself is needed. I can see a potential issue in the far future of having a dead link hardwired in the proposal, so it might be better not to explicitly include that.

Really great job, folks. This is just outstanding, and I honestly believe we will have saved the Hugos if we can get this passed. Hats off to all of you for the hard work and discussion!

Thanks,
Kilo

#796 ::: Soon Lee ::: (view all by) ::: June 18, 2015, 04:24 PM:

I think we're there.

(...said while resisting the urge to polish the wording just a tiny bit more...)

#797 ::: abi ::: (view all by) ::: June 18, 2015, 04:30 PM:

There's no issue with starting two threads. I'll probably keep the longest text behind the cut.

Please be clear and explicit what you want posted, though. I'm a little confused by what you're describing right here.

#798 ::: David Dyer-Bennet ::: (view all by) ::: June 18, 2015, 04:31 PM:

Kilo@795: Ookay!

Agree on being ready to submit, agree that we can in fact make changes for a while yet if absolutely necessary. Getting some outside feedback, and getting some buzz going, are valuable, and I think it's time.

A separate thread for the new discussion sounds right to me, should that be possible. We're 80% of the way to 1000 right here already, on top of other considerations.

The link would be in the FAQ, not something with that long a life, but whatever; once we've said "Making Light" it's trivial to find in any case. I do think having the link right there emphasizes the transparency slightly, is all.

#799 ::: Tim Illingworth ::: (view all by) ::: June 18, 2015, 08:00 PM:

e-pluribus-hugo.com is available

#800 ::: Doire ::: (view all by) ::: June 19, 2015, 03:49 AM:

David Dyer-Bennet @798 Perhaps a date, or dates, of the post start(s). That would direct people even if the URL changed.

#801 ::: junego ::: (view all by) ::: June 19, 2015, 08:44 AM:

I'm in on sending this puppy off to Sasquan!

#802 ::: David Wallace ::: (view all by) ::: June 19, 2015, 11:48 AM:

Is it too late to add my name to the supporters?

[SIGNING] and [ATTENDING].

After many years on what could charitably be described as the outer penumbra of fandom, I've ponied up for two attending memberships for myself and my son specifically so I can come to the business meeting and vote for this proposal (and vote for the Hugos themselves, of course). This will be our first WorldCon, and first SF con of any kind. I trust we'll have a good time, and I suspect it won't be our last.

The Hugos used to mean a lot to me, back in the days when I was in school and looking for recommendations of quality SF to read. I haven't been reading much new fiction in recent years, but the Puppies controversy this year has made me realize just how much I care about the significance of the Hugos. I am pleased to be able to support this proposal.

#803 ::: Keith "Kilo" Watt ::: (view all by) ::: June 19, 2015, 03:05 PM:

All:

It's done then!


I just submitted the Word document of the proposal to Kevin, so hopefully we'll see that on the site soon. If any of you have contacts or mailing lists that you'd be willing to put the word out to, I'd greatly appreciate it. I will be doing the same with my contacts. The proposal should appear at the following link:

http://sasquan.org/business-meeting/agenda/


Abi, I'll make two more posts for the two threads that I think we should start. Please let me know if you need anything else or if there are issues.

Congratulations to everyone!

Kilo


#804 ::: Keith "Kilo" Watt ::: (view all by) ::: June 19, 2015, 03:06 PM:

[THREAD 1 INTRO]

E Pluribus Hugo: Post-Proposal Planning and To-Dos

Now that the proposal has been submitted, we can turn our attention to the other matters that we need to complete. In particular, we need to look at the following items:

  • Prepare a "6th place" amendment for handling declined nominations, in the event it is needed
  • Discuss swag/ribbons/T-shirts for handing out at Sasquan
  • Strategize ideas for the actual proposal presentation at the business meeting itself
  • Plan a meet-up for those who will be attending Sasquan
  • Discuss places/parties at Sasquan where we can campaign for the proposal
There may be other issues that need to be discussed as well -- this is the place for them!

I've listed these iems roughly in order of importance, but as long as we are using [TAGS], I think we can multi-task without too many problems.

Let's get to it!

Kilo

#805 ::: Keith "Kilo" Watt ::: (view all by) ::: June 19, 2015, 03:09 PM:

[THREAD 2 INTRO]

E Pluribus Hugo: Community Q&A

After over 3000 messages and nearly two and half months of discussions, we have put together a simple Hugo Award nomination system that meets our primary goal:


No group -- whomever that group may be -- should be able to absolutely prevent nominees from having the chance to be considered for the Hugo Award.


This works both ways. Not only should a slate not be able to force all other nominees off the final ballot, but just because a nominee appears on a slate, it should not be disqualified from appearing on the final ballot. No "shadowy cabal" should be able to prevent a nominee from being considered, either. All nominees deserve a fair chance. E Pluribus Hugo accomplishes this goal.

The system was developed by fans with a wide range of interests in science fiction and fantasy, including those who prefer the genres supported by the Sad Puppies 3 slate. We hope that it will truly meet the needs of all varieties of SFF fans.

The purpose of this proposal is explicitly not to change the rules so as to prevent the "new voices" of 2015 from being heard. There is no shadowy cabal here. All discussions, including our missteps and paths not taken, have been conducted in a completely open forum. You can find all of those discussions in the following threads:

On voting systems: a guest post from Bruce Schneier
Discussing Specific Changes to the Hugo Nomination Election: Another Guest Post By Bruce Schneier
Discussing Specific Changes to the Hugo Nomination Election: A Post Not By Bruce Schneier
E Pluribus Hugo: Out of Many, A Hugo

This thread is for questions and answers (and hopefully voices of support!) from the community about the system. We hope that the FAQ attached to the proposal will answer many of your questions, but we want to be available for any other questions that may arise. We sincerely hope that you will give this proposal a thoughtful and fair hearing. In this effort we are not Social Justice Warriors, we are not Puppies, and we don't hate anyone. We are simply fans who want to save the Hugo from sinking into irrelevance as a result of the flaw in the nomination system that has been pointed out this year.

The official proposal can be found here once it has been posted by the Sasquan business meeting chair. We look forward to hearing from you.

#806 ::: Keith "Kilo" Watt ::: (view all by) ::: June 19, 2015, 03:10 PM:

David@802: I added you to the signer list -- thanks for supporting the proposal!

Abi: In the thread 2 intro, I forgot to bold the title, "E Pluribus Hugo: Community Q&A" Sorry about that!

K

#807 ::: Kevin Standlee ::: (view all by) ::: June 19, 2015, 03:37 PM:

Keith "Kilo" Watt @803:

I just submitted the Word document of the proposal to Kevin,....

As of 12:35 PDT, nothing has arrived in my WSFS Business Meeting mailbox. You did send it to wsfs-business [at] sasquan.org, right? That sends it to the entire WSFS Business Meeting team, including the WSFS Secretary, Linda Deneroff, who is the person who actually adds the proposal to the Agenda.

#808 ::: abi ::: (view all by) ::: June 19, 2015, 04:43 PM:

Sorry for the delay - I was waylaid by a lap teenager. I'll get the entries posted now.

#809 ::: abi ::: (view all by) ::: June 19, 2015, 04:50 PM:

And done!

#810 ::: Soon Lee ::: (view all by) ::: June 19, 2015, 06:25 PM:

Thanks Abi!

Thanks also to Kilo, Jameson, & everyone else who has participated!

#811 ::: Keith "Kilo" Watt ::: (view all by) ::: June 19, 2015, 06:43 PM:

Kevin@807:

That was the address I used, but I'll double check when I get home. We're on the road bringing my daughter back from camp at the moment.

K

#812 ::: Keith "Kilo" Watt ::: (view all by) ::: June 19, 2015, 07:59 PM:

Kevin@807:

It looks like I had the address correct, but I resent another copy anyway. Please let me know if you get it!

Thanks,
Kilo

#813 ::: Kevin Standlee ::: (view all by) ::: June 19, 2015, 10:25 PM:

Keith "Kilo" Watt @812:

It looks like I had the address correct, but I resent another copy anyway. Please let me know if you get it!
For some reason, the first copy took most of the day before it arrived. Both copies have arrived and the Secretary has acknowledged receipt. Thank you!

Remember that if you decide to make any changes, you need to send them to the wsfs-business address and that while I do try to keep an eye on these discussions, the Secretary does not, and only official submissions to wsfs-business before the deadline count.

Parliamentary neepery: once the deadline has passed, the proposal doesn't technically belong to you anymore, but instead belongs to the meeting, and can only be changed or withdrawn with the consent of the meeting. This may sound academic, but it did actually come up a couple of years ago when the primary sponsor of a proposal tried to withdraw it, but wasn't allowed to do so because she would have needed the consent of the meeting, and the members present voted against allowing the withdrawal.

#814 ::: Keith "Kilo" Watt ::: (view all by) ::: June 19, 2015, 11:32 PM:

Kevin@813:

Thanks much! I just heard from Linda and I see that the proposal is on the agenda page. Looks like everything is good to go.

Thanks again!
Kilo

#815 ::: David Goldfarb ::: (view all by) ::: June 20, 2015, 01:57 AM:

It looks like the version on the Sasquan website does not include David Wallace as a signer. Possibly one of the ones Kilo sent had him, and one didn't, and the wrong one was used?

#816 ::: Kevin Standlee ::: (view all by) ::: June 20, 2015, 02:16 AM:

David Goldfarb @815:

It looks like the version on the Sasquan website does not include David Wallace as a signer.
David Wallace's name appears between Adam Tilghman and Sara Watt. The names were not listed alphabetically.

#817 ::: David Goldfarb ::: (view all by) ::: June 20, 2015, 03:39 AM:

So it does.

Most of the list is ordered by the time in which people posted their endorsements, which is why I'm relatively early and Rogers Cadenhead is last. David Wallace was the latest one I've seen, so I looked for him at the end. I didn't expect him to have been stuck in the middle.

#818 ::: Keith "Kilo" Watt ::: (view all by) ::: June 20, 2015, 10:18 AM:

David@817:

The signers are in the order they posted, but those who appear before my wife's name (Sara) indicated they were attending the business meeting as well. It was just a way for me to stay organized is all.

Kilo

#819 ::: Brad from Sunnyvale ::: (view all by) ::: June 20, 2015, 08:06 PM:

I know that I ticked a few off here why my strong belief that an algorithmic approach (with some flaws) is not the correct answer, and further I know that a lot of the discussion here has already passed, most of it while I was on the road in remote places.

But strangely, even after so much discussion an entirely new solution came to me today. Making Light may not be the place for it to be discussed, but after my participation, I felt I should put it forward.

It's quite simple, with two components:

1) The ability for some VERY large group of fans to request a very limited review of the Hugo nomination process, such review to publish in that case the names of nominees who may have missed the ballot due to impropriety, and a description of the reasoning behind that.

and

2) The ability for voters on the final ballot to enter a write-in instead of an official ballot entry in any or all of the ranking slots.

----------------------------------

#1 actually is technically allowed under the constitution today. It would simply be frowned upon. Without write-ins it would have fairly minimal effect. A constitutional change would simply formalize the doing of it in extreme situations.

#2 is of course found in many election systems, and only rarely makes a difference in them. It is fully democratic.


So what would have happened this year?

Certainly a very large group of fans would have asked for the review. Probably more fans than have ever voted in the Hugos in HISTORY, would ask for the review, that's how high the threshold could be.

One hypothetical course for the reviewers would have been to examine ballots which contained X% or more Puppy slate choices, perhaps X=60%, and removed said puppy slate choices from those ballots, followed by a recalculation. (Sad/Rabid candidates on ballots which only had 2 such nominees and 3 others might have still been counted, for example.) The review would have released the names of any nominees who now would have made this recalculated ballot, and the methodology, but perhaps not the names of the slate choices.

The existing Hugo final ballot would have been published, but many fans would have elected to offer write-in choices, and many of them would have written in the names published after the review. In all probability, no slate choice not also very popular with the rest of fandom would win an award, even in categories where all 5 choices were slate-promoted.

Why is this better?

a) It's pretty simple. Nothing changes in nomination.
b) It's robust. While I have deliberately not described the process for calling for and conducting a review, it would be a human process, as all justice systems in the world have, so it could respond to many attack strategies against the awards.
c) It's a deterrent -- slates would be less likely to act and get people to buy supporting memberships for no other reason than to buy the Hugos, because they would know it was unlikely they could win. With pure algorithms, the attacker can know if an attack will work in advance, and thus judge if they should buy memberships.
d) It can be combined with E Pluribus Hugo, if desired. Ie. nominate that way but still have review and write-in to fix problem (e)
e) It can prevent the slates from even getting in the top 5, if fans wish to vote such a rebuke. To get several of your slate on a ballot to have them all finish below 5th place (possibly those below 5th would not even be published but I like transparency more than the snub.)
f) It assures no worthy candidate goes unrecognized or gets no chance because of slates.

What are its downsides?

a) A modest amount of extra work in vote calculation (nothing worse than doing nomination which is all write-in.)
b) A very limited ability of the group doing the review to exercise a bias in the course of their review. However, the most they can do is expose the names of works they judge inappropriately excluded from the ballot, they can't undo anything about the appearance of works that got there via approval or E-Pluribus.
c) It does nothing about attempts to steal the final ballot.

This differs from plans to just increase the number of nominations in a couple of ways.

a) It should, ideally never happen, because its deterrent capability should make collusion much less fruitful, though not entirely.
b) The write-in works are still at a disadvantage over those that made the ballot.

Variant B with less change: Add only write-in. Let fans propose and promote their own lists of who to write in, though that gets close to slate vs. slate which I don't like. I have some fixes for that if need be.

Who would do the review? Well, it could be the Hugo Committee but there are other options which would relieve them of that burden.

#820 ::: P J Evans ::: (view all by) ::: June 20, 2015, 08:35 PM:

818
Why propose these at the end?)

#821 ::: Keith "Kilo" Watt ::: (view all by) ::: June 20, 2015, 11:07 PM:

Brad@819:

It's an interesting idea, and as you say it certainly doesn't preclude working with EPH. The problem I see is that unlike with human justice systems, there's no definitive right or wrong here in most cases. That's no "finding of fact" that can be done. So let me propose a couple of situations for you that I'm not sure how your idea would handle.

Over on File770, I posted an example involving Dr. Who in the Dramatic Short Form category. This category often gets swept by Dr. Who fans, keeping all other works off the ballot. Trouble is, they aren't a slate -- or are they? There's no collusion, no organisation. It's just that they still tend to pick the same five episodes as the best of the season, so even though they are nominating independently, they function as a slate. So the question is, should this trigger your modified inquisition? Are the fans who complain that their works didn't have a chance justified? Did the Dr. Who fans do anything wrong that causes them to deserve to have their choices removed? I don't think there's a clear and definitive answer to this, no matter who gets to make the decision. Saying, "the majority gets to decide" doesn't mean the majority is necessarily right.

You cite human justice systems, but even in the context of elections, a committee decision can still be contentious for years afterward. Look at the election of GW Bush, decided (effectively) by the Supreme Court for a classic example. There is -still- controversy over that one. It seems to me that the Hugo system needs to be impartial if it is to maintain its integrity. Ultimately, no human can be absolutely impartial, so I have a hard time seeing how out it can work.

Just for what it's worth...

Kilo

#822 ::: Idumea Arbacoochee, Gardener of Threads ::: (view all by) ::: June 21, 2015, 05:48 AM:

At the request of the commenters, I have removed the comments that were mistakenly published in this thread. They're over in the Q&A thread where they belong.

#823 ::: Brad from Sunnyvale ::: (view all by) ::: June 21, 2015, 06:24 PM:

I don't pretend any human judgement will be perfect, far from it. I just believe you can't fight human attackers with purely algorithmic and strict-letter-of-the-law approaches. You must have a dynamic response to dynamic attack. The WSFS constitution is not, to paraphrase, a suicide pact. It was meant to be hard to change (taking 2 years) to keep things stable, but it was not at all designed to deal with overt attack.

I do not think that Doctor Who fandom dominating nominations would cause for thousands of fans to call for a review, and any review would probably find no wrongdoing requiring action. And even so, the write-in ability would allow other fans to vote for other works, if the DW domination is indeed a failure of the system to represent true fan opinion.

No, I want human justice to come in when it's overt and coordinated, and would design it to not come in when it's subtle and non-coordinated.

When John C. Wright gets more Hugo nominations than any other writer ever, it's hardly subtle. It's practically impossible to do major collusion in secret.

But in this proposal, in the end, it is Hugo final voters who decide. If they don't like the 5 episodes of DW, they can vote in something else. (For example, Black Mirror: White Christmas may be the best SF TV show ever made, so I would happily write it in this year!)

#824 ::: David Harmon ::: (view all by) ::: June 22, 2015, 08:09 AM:

Idumea #822: ... the comments that were mistakenly published in this thread. They're over in the Q&A thread where they belong.

Not anymore they aren't. Is this a mod reversal or a glitch? In any case, it might be worth closing this thread.

#825 ::: Richard Ristow ::: (view all by) ::: June 24, 2015, 04:26 PM:

I'm very late to this discussion -- I haven't been here, but have been following the postings on File770.

I see one possible 'instability' (as I'll call it) in the EPH system.

I can imagine there being some hugely popular work, and fans failing to include it on their ballots, thinking
(1) It will inevitably be nominated
(2) There are one or two less-popular works I'd like to see nominated, and including the highly popular work will dilute my vote.

This opens the possibility of a work failing to be nominated, precisely because everyone thinks it will be.

#826 ::: P J Evans ::: (view all by) ::: June 24, 2015, 04:27 PM:

825
I'd describe that as very, very unlikely. Fans tend not to nominate that way.

Welcome to Making Light's comment section. The moderators are Avram Grumer, Teresa & Patrick Nielsen Hayden, and Abi Sutherland. Abi is the moderator most frequently onsite. She's also the kindest. Teresa is the theoretician. Are you feeling lucky?

Comments containing more than seven URLs will be held for approval. If you want to comment on a thread that's been closed, please post to the most recent "Open Thread" discussion.

You can subscribe (via RSS) to this particular comment thread. (If this option is baffling, here's a quick introduction.)

Post a comment.
(Real e-mail addresses and URLs only, please.)

HTML Tags:
<strong>Strong</strong> = Strong
<em>Emphasized</em> = Emphasized
<a href="http://www.url.com">Linked text</a> = Linked text

Spelling reference:
Tolkien. Minuscule. Gandhi. Millennium. Delany. Embarrassment. Publishers Weekly. Occurrence. Asimov. Weird. Connoisseur. Accommodate. Hierarchy. Deity. Etiquette. Pharaoh. Teresa. Its. Macdonald. Nielsen Hayden. It's. Fluorosphere. Barack. More here.















(You must preview before posting.)

Dire legal notice
Making Light copyright 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020 by Patrick & Teresa Nielsen Hayden. All rights reserved.