[6th PLACE]

How about this for an initial cut at a 6th place amendment:

Moved, to amend section 3.A (Finalist Selection Process) as follows:

3.A.4: After the initial Award ballot is generated, if any finalist(s) are removed for any reason, and this removal would bring the number of finalists below the maximum specified in subsection 3.8.1, 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. that finalist(s) shall be replaced on the final Award ballot by the nominee(s) eliminated in the round(s) immediately preceding the final round.

Just an idea to start...

Kilo

Shouldn't that be "... would bring the number of finalists below the minimum..."? I mean, that's what the minimum's for, right? ;-)

David@2:

I see your point. Elsewhere in the proposal we say "the number of finalists specified in 3.8.1" so maybe that's what we should say here instead of "maximum" or "minimum".

Kilo

We should explicitly state that if the nominee(s) eliminated in the round(s) immediately preceding the final round are in fact tied, then all of them get on the ballot.

Essentially remove the '(s)' from the amended text to avoid confusion.

[6th PLACE]

Okay, incorporating Duncan's and David's suggestions:

Moved, to amend section 3.A (Finalist Selection Process) as follows:

3.A.4: After the initial Award ballot is generated, if any finalist is removed for any reason, and this removal would bring the number of finalists below the number specified in subsection 3.8.1, 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. that finalist shall be replaced on the final Award ballot by the nominee eliminated in the round immediately preceding the final round (continuing backwards through the preceding rounds if there are further finalists removed). If there was a tie in this preceding round, all members of the tie will be placed on the final ballot, extending it if necessary.

It would be *really* useful to have some sort of visual tutorial on how EPH would work in practice, like a graphic or short video or ideally both.

I think if we can show visually how a member's nomination would be treated with EPH compared to the current system (a worked example with "my ballot", with & without slates in play) it would help inform & convince.

Maybe E Pluribus Hugo could take some inspiration from the similarly named Ancient and Honorable Order of E Clampus Vitus. In particular, the most essential element of Clamper regalia is the red shirt, which has the benefits of being simple, traditional, historical, and also distinctive.

[6th PLACE]

If everyone is happy with the 6th place amendment in #5, we'll call it done, and I'll check it off the list. Any last comments or suggestions?

Kilo

Keith "Kilo" Watt #8: It seems straightforward enough to me.

Keith "Kilo" Watt @8

Works for me.

WFM.

[RIBBONS]

Jameson, last I heard you were using the Futura font on the t-shirt design. Is that still true? I'll work on a coordinating ribbon design this week.

Currently we have the following requests for ribbons: Jameson (150), Keith (100), JJ (100), ULTRAGOTHA (50), Adam (100), Joshua (50), and Tammy (50), plus four requests for mailing ribbons. If I don't have you listed already, let me know if you'd like to distribute some at the con (cost will be $9 per 50 at the 1000-ribbon rate; I'll order enough myself to reach 1000 in any case). Current plan is to distribute these at whatever meetup we organize as a group, if it's early enough in the con.

If you're not attending but want a couple ribbons as a keepsake, fill out this form and I'll mail you some. (No charge for that; if you want to pay for it, donate an extra buck to your favorite charity.)

[RIBBONS]

Chris@12:

Thanks for taking the lead on this, Chris! One thing I was thinking about yesterday, is that it would be great if we could make the ribbon visually distinctive enough that it can be recognized (not necessarily read) at a distance. Multiple colors and/or patterns may be too hard, though, I don't know. I'm just thinking that once someone knows a particular color/pattern ribbon is EPH, they would then recognize it everywhere, even if they weren't close enough to read the text. I know people often end up with lots of ribbons attached to their badges, so it'd be cool to not get lost in the noise. There may be a way to do the same thing with just text as well, of course.

Just a thought, and maybe one that isn't practical...

Kilo

Yes, futura. It appears people like bullets between the words, too.

What colors should we go with? I don't care but I think we should coordinate. Perhaps not red or blue because those have political overtones. But it's not that we have to avoid everything reddish or bluish; burgundy or teal or sky blue would be fine.

Somebody with designer knowledge: color nominations pls? Preferably 2-3 nominations, so we can (score) vote on them?

Jameson@14:

Bullets between words definitely gets a thumbs up from me. Looks very... SF.

You should definitely not listen to me on color (so says my wife), but if we had something bright (but not offensively so) that would stand out, that might work well with regards to my #13.

Kilo

14
I'm in favor of light colors: sky blue, silver or pearl gray, like those.

it would be great if we could make the ribbon visually distinctive enough that it can be recognized (not necessarily read) at a distance.

Even without being up on "ribbon culture", I can guess that this way lies an arms race. :-)

[STANDLEE SIGNAL]

Kevin, I don't know if you're following this thread as you did the other one; if not, I'll send you email directly.

I saw your post on File770 that you were thinking of suggesting holding a "Committee of the Whole" to discuss the technical aspects of EPH before we vote to decide if it's a good thing. That sounds great to me, but if you would, please keep me informed. I'm hearing impaired, so I've arranged for sign language interpreters to be at the business meeting. I'd need to make sure that they are there at the time for the COTW as well.

Thanks much!
Kilo

All:

Hmm... It just occurred to me that I don't know that I've ever mentioned the hearing impaired thing to this point. I have really good digital hearing aids with voice recognition (they automatically screen out sounds that aren't speech when I run that program) and 3D scanning microphones, so I do really well in small groups. Not to worry, I won't have any problem communicating with any of you! In large groups, however, they are pretty much useless (nothing close by to lock on to), so I arranged with Sasquan to have ASL interpreters at the business meeting for me (I'm fluent in ASL -- necessity is a hard taskmasker!). The only real effect on my speech is that I seem to have lost most of my Tennessee drawl... ;-)

For the curious, I lost my hearing due to meningitis that I contracted while deployed on my last mission in the Navy. Given that it's normally fatal, the hearing and balance problems are a pretty fair trade as far as I'm concerned! :)

Kilo

[RIBBONS]

Hodges Badge (which has been doing the ribbons for Boskone et al. for decades) has a lot of colors, but a real favorite is the rainbow, where one nice, bright color blurs into the next. Whatever company you're using should have something similar. Although I'm partial to a nice turquoise.

David@17:

LOL, you're probably right about that!

K

[RIBBONS]

Kilo @ 13:

Absent a particularly unusual color scheme, I suspect the best way to guarantee long-distance visibility would be large, bold text.

That said, an unusual color scheme isn't out of the question; my friends who run pronounribbons.org designed them as dark text on bright green ribbons so you can immediately pick that one ribbon out of a large set, once you know to look for it.

I would still prefer something simple - bold text in silver on a black ribbon was my initial thought - but I'm willing to hear other suggestions. A distinctive layout could also help (e.g. a two-line ribbon that says "E PLURIBUS" on one line and then "HUGO" in a larger font on the other line; that text-shape will be more distinctive than a simple single-line-of-text design).

Pfusand @ 20:

The vendor I first looked at didn't have rainbow text, and the second one I found when I went searching for rainbow text recommended against printing it on a dark background, but I foresee legibility problems on a light background. I'll check out your recommendation, though.

[RIBBONS]

Chris @ 22 wrote "The vendor I first looked at didn't have rainbow text, and the second one I found when I went searching for rainbow text..."

Ack, no! I meant rainbow ribbons. Text (ink) is the usual choices.

You [and your friends] are right: Emerald green ribbon would be nice, and eye-catching.

[RIBBONS]

Pfusand @ 23:

Ah, got it. Rainbow ribbons would be eye-catching, and I like the symbolism of a variety of opinions coming together, but I also worry about coopting the Pride symbol. I wouldn't be surprised if there were people at the con handing out Pride ribbons of some sort.

I gather there will be a fairly large number of pronoun ribbons at the event (they were a big hit at Norwescon this year), so I'd like to avoid stepping on their design too. I like turquoise, though. I'll mock up a few different color combinations and post them for comments tonight (or maybe tomorrow night).

Folks -

I just got this from Linda Deneroff, the bid mess meeting secretary. I don't have a problem with it, but I want to see what the group thought. I think I would want the link to the FAQ before the proposal text, since that's actually more important than the amendment itself, I think.

Let me know!

Kilo

======================================
Hi, Keith,

The question was raised about the possibility of creating a separate page just for E Pluribus Hugo FAQ, with a pointer from the New Business page. I'm asking because EPH is so long and complex and people don't realize there is material that follows it, including at least one other constitutional amendment.

Please let me know if this is acceptable to you, and I'll get it taken care.

Bid mess = business

... That was pretty funny. I'm on my phone at the circus, sorry! :-)

K

Kilo,

How about putting E pluribus Hugo as the last item on the agenda page?

That way, no new page needs to be created, also all the information (Proposal + FAQ) is on the same page.

Separate page might be a good idea, so that the whole thing will be together.

Putting EPH always last would require us to constantly be renumbering proposals as new ones arrive. While there is a possibility that the Business Meeting will reorganize the Agenda after the submission deadline, until then we would rather not do so because people will be referring folks to the Agenda page with specific item references, and EPH isn't the only proposal we're juggling.

29
Separate page, with a reference from the main agenda page, that's what's being proposed?

It makes sense to do it that way.

Kevin Standlee #29:

In that case, separate page for the FAQ makes more sense.

My only concern with separating the FAQ and the formal proposal is that people might struggle with the proposal language before looking at the FAQ. I think we could precede the proposal text with "We recommend reading the FAQ, found here, for a full explanation before delving into the legalese!" or some such, and that might be okay. We could also put both the FAQ and proposal on a separate page with a link from the main page.

Opinions?

Kilo

Kevin:

If you do reorder the agenda before the meeting, it might be worthwhile to put EPH first. While EPH will work with the other proposals, we might decide some them are not needed if EPH passes.

Also, did you see my #18 in this thread? I just want to make sure the Sasquan folks are up to speed, since they needed to arrange interpreters in advance.

Thanks,
Kilo

Keith "Kilo" Watt #32: "We recommend reading the FAQ, found here, for a full explanation before delving into the legalese!"

Be careful to avoid condescension -- I would hope that anyone reading the BM agenda would be at least somewhat interested in the legalese! Perhaps something more like: "This is the formal amendment text, there is also a plain-English FAQ section [here]."

David Harmon @ 34: "This is the formal amendment text, there is also a plain-English FAQ section [here]."

Love it!

David@34:

That looks good to me, too. So we're okay with having just the formal language on the agenda page and just the FAQ on a separate page, assuming we can get that message ahead of the formal language?

K

Kilo #32 & #36:

Good point, especially given Kevin Standlee's Loncon experience:

"Yes, the proposals include FAQs. Regrettably, many members simply won't read them. My experience last year in London with Popular Ratification was that most of my replies consisted of reading back the commentary included with the proposal."

I got a comment eated by database or something, but I agree with @34-36. I also like turquoise, and wonder what dark color to use for the letters on that.

Jameson Quinn @38, Metallic gold shows well against turquoise. Just sayin'...

Gold and turquoise is now option 1. Any other options?

By the way, the draft PowerPoint doesn't have any design to it. I'm thinking we should make it match whatever we decide on for ribbons...

K

I'd like to put in a vote for emerald green background and whatever text color looks good with that (maybe silver or white?).

Kilo

[RIBBONS]

I haven't heard much in a few days; how are we doing on ribbons?

Kilo

[FAQ]

Given that the graphical demo has proven to be fairly useful to folks, I think we should add an entry in the FAQ that links to it. I also think we should expand the entry about bullet voting, since that seems to be a persistent misconception about the system. Finally, I think we should point out in the FAQ that this system is no more complicated than the instant runoff voting system we use for the finals currently, as I don't think people are making that realization. So, here's some proposed wording. I'd like to get a quick turn-around on these if we can, since I think Linda would like to get that separate FAQ page going.

Thanks!
Kilo

P.S. Note that adding the new FAQ entry renumbers the others.
===============================================

2. I’m a visual person. Is there a graphical way to see how the system works?
Yes! We have created a PowerPoint (and also in PDF form that can be viewed in a browser, but you lose the animations) that explains the system and shows how it limits the effects of slates. In this graphical form, you can watch the eventual finalists’ bar graphs of their point totals get longer and longer as other nominees are eliminated. You can also see how slate nominees tend to compete with and eliminate each other, since they tend to have virtually the same number of both points and nominations.

7. 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 and one nomination for it.

The bottom line is that a low point total isn't what causes a nominee to be eliminated. Not having enough nominations is what eliminates a nominee – exactly as it is under the current system. There is no way that you (alone) can increase the number of nominations your favorite nominee gets – not even by only listing that one nominee. If you only list one nominee (“bullet voting”), you can give it at most an extra 4/5 of a point (but more reasonably only around 1/3 of a point, because you weren’t likely to pick all five finalists in any event). Listing only one nominee would have to give your favorite enough points to put it all the way into fourth place, so that its nominations are never compared for elimination. Since finalists may have well over 100 points, the odds of making much of a difference with this strategy are very, very small – and the price you pay is that you don’t get to nominate any of your other favorites.

In general, the best strategy is simple: nominate as many nominees as you feel are worthy.

21. This system looks really complicated. Won’t E Pluribus Hugo be difficult to code and implement?
We should keep in mind that EPH is actually not any more complicated than the Instant Runoff Voting system we currently use for the Hugo finals. One of our non-experts coded a full simulator for the system in a matter of days. If the current nomination-gathering code can be modified to output a comma-delimited text file, we can use the existing EPH code as-is.

Me@44:

I changed "Since" to "Because" at the end of the second paragraph of FAQ7.

K

All:

I haven't seen any objections or corrections, and I'd like to get this to Linda as soon as possible, so I'm going to assume we're good to go on the modifications. Speak up if you have more!

Kilo

Slide 5 says #28-#32 are finalists under the current system.

Slide 6 goes straight to voter#44 under EPH.

It's too abrupt. There should be something along the lines of "Now let's look at what would have happened under E pluribus Hugo", I think as the title of Slide 6.

The FAQ changes look fine to me.

Soon Lee@47:

Good suggestion; I've uploaded a new version with the slide you suggest.

Thanks!
K

@44: On @7, I'd say "more reasonably around 1/3 of a point on average", or maybe just "on average around 1/3 of a point".

Jameson@49:

Sounds good. I had already submitted a revised document to Linda and Kevin, but I'll make that change for the next version I send. I also made the change in slide #30 of the PowerPoint, which I'll go ahead and update.

Kilo

[RIBBONS]

My apologies for the lack of updates; I've been working 60-70 hours a week lately. I will have some bandwidth soon to put together a ribbon mockup. I like how the gold-on-turquoise idea looks in my head, at least, and I'll give emerald green a try as well.

Chris@51:

Thanks for the update; I was getting concerned. Glad to see you've got it all in hand, and I hope your workload settles down a bit! I'll be interested to see the mockups!

K

[RIBBONS]

Mockups of ribbon designs in four different color styles and four different styles of text.

Colors:
* Silver on green
* Silver on turquoise
* Gold on turquoise
* Silver on black

Styles:
* One line of text
* Two lines, "Hugo" emphasized
* Two lines, "Hugo" emphasized, with dots
* Two lines, same size text

Thoughts?

[RIBBONS]

I showed the mockups to my wife, so here are her comments:

- The turquoise background looks good, but text can't be read. Is there are darker color that could be used?
- The shade of green chosen isn't good, but a darker green might work.
- Of the ones pictured, the black background is easiest to see, but isn't that exciting. One possibility is a darker color like dark purple.
- The third row text is best, but "there's something odd about the dots that bothers me". Getting rid of the dot between E and Pluribus might fix it.

As always, don't listen to me about art, but she's got taste. 😀

Kilo

Clarification: For the turquoise background, is there a darker -text- color that could be used?

K

Personally I like Silver on Green or Turquoise. I'm assuming the actual text colour will be metallic, so stand out better from the background (though the contrast looks fine already on my monitor). Gold on Black also might look nice.

I don't like the single-dot versions; but either Row One or Row Three would work. Maybe Row One with horizontal lines added above and below the text to fill the space a bit more?

---------------
E*Pluribus*Hugo
---------------

Or maybe a five dot version? Probably a bit much.

*E*Pluribus*
  *HUGO*

Of the designs shown, I greatly prefer style 3, silver on turquoise. Yes, the metallic silver will give a much higher contrast. Alternate: same style, but metallic gold on a darker green.

How about mixed case? (Oddly enough, it is easier to read at a distance.)

Style 3, but I do think it needs a little more contrast on the not-black options.

I agree that mixed-case would be easier to read. Could we see some mockups that way?

I was talking with my uncle, a designer, and he recommended putting a planet to center people's attention. Something like this. (The light gray in that picture is supposed to represent metallic silver).

I like that. Although it does add a second color, which might increase printing costs.

My uncle also said that the planet should be done with "sublimation printing", which I'm sure is also more expensive. He's the kind of guy who has very stylish T-shirts, so he'd know how to do it right. But I doubt CafePress or whatever has all the options exactly as he'd want them...

Oh, I also realized that with those colors (base darker than planet) the planet looks better if you turn it 180 degrees. Which I could post some time, but I'm waiting to see if there are other things to fix / requests. (I'm using inkscape and I think it should work to download the file and edit it, if you want to play with it.)

Kevin Standlee just pointed out over on File770 that:

Whether or not to even release the anonymized [Hugo nomination] data is entirely up to this year’s Administrators. They have not said whether they will agree to do so. If they decline to do so, there is no recourse; it’s completely their decision. If they do, I would expect that (assuming EPH gets first passage) that over the period between now and next year’s Worldcon (when EPH would come up for ratification) people will study the data and run it through simulations to find out how it worked out.

The Business Meeting could pass a resolution (which is not binding) recommending or asking the Adminsistrators do release such data. Nobody has yet submitted such a resolution to the New Business queue. The deadline is two weeks before Worldcon starts. I’m actually surprised that no EPH supporters have submitted such a resolution.

Well, here's a rough draft:

Be it resolved that the World Science Fiction Society recommend that the Administrators of the Hugo Awards for Sasquan release the anonymized data from this year's Hugo nominations that it may be studied for the purpose of amending the Hugo nomination process.

Bah -- that should be "recommends" not "recommend" -- drat it!

Jameson@61:

For what it's worth, my wife said she kind of bounced off of the more eye-catching T-shirt design. That's likely in part because she -really- liked your original design, though. She totally gushed about it...

Kilo

All:

Here is some sample language that Kevin suggested when I asked for help from WSFS. What do you think?

Kilo

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

Short Title: Hugo Nominating Data Request

Resolved, That the WSFS Business Meeting requests that the Administrators of the 2015 Hugo Awards make publicly available anonymized raw nominating data from the 2015 Hugo Awards, including the works nominated on each ballot in each category but not including any information that could be used to relate ballots to the members who cast them; and
Resolved, That it is the opinion of the WSFS Business Meeting that releasing such anonymized raw nominating data after the announcement of the results of the 2015 Hugo Awards is not a violation of the privacy of members’ ballots.

Submitted By: [At least two members, supporting and/or attending, of Sasquan]

Commentary: [Justification for requesting this data.]

Keith "Kilo" Watt, #68: For what it's worth, my wife said she kind of bounced off of the more eye-catching T-shirt design. That's likely in part because she -really- liked your original design, though. She totally gushed about it...

My two cents: I find the new design with the planet added very busy and garish, and do not like it nearly as well as the original.

I like the design that's just the rocket; it puts the focus on the stories.

Design people may think that adding the planet draws the eye or whatever and it may even be true on a huge bullboard. But on a shirt or a flyer? It draws the eye to the big ball (call it a planet if you want) and away from the titles in the rocket... the rocket seems to be there to draw attention to the planet and not the other way around.

PS: I and art are rarely on speaking terms so I may be missing something but...

OK, I think that sentiment is generally against the "planet". So I'll drop that idea.

Playing with the colors to develop that proposal, I found that "silver letters on an indigo background" grew on me. What do others think of that combination? I'd also be happy with other metallic colors (copper or gold) on indigo.

How well do metallic inks wash? I honestly don't know. If they're as durable as other inks, that's great.

From quickly googling it, it appears that simply "shiny" metallic inks can be problematic after washing, but "sparkly/glittery" inks last as well as other inks. I'd be OK with sparkly. (The sparkly inks are sold as "shimmer" because lots of people hate glitter.)

@69: I support that proposal and would sign on to it.

Silver on indigo seems nicely "spacey" -- starlight against evening sky.

Kilo #69: I'd support that. I see no technical details, but I can all-too-easily imagine the hazards of putting specifics into a long-term mandate there.

Jameson@79: Okay, good -- we'll need at least two, though of course it'd be great to have more.

We should add a justification in the commentary. Does anyone want to take a first crack at that?

And of course, just because Kevin suggested this text doesn't mean we have to stick to it (though to be honest, I think we'd need a pretty compelling reason not to).

Kilo

Jameson@73: Silver letters sounds cool to me too! Can you do a mockup with the indigo? I'll pass it on to my fashion adviser... ;-)

K

If we can have silver (or at least silvery-sparkly) letters for the rocketwords without compromising on washability, I'm all for it. It evokes the award that much better. And indigo does seem like a logical background color.

I would sign on to a request for the anonymized ballot data.

[RIBBONS]

New mockups.

This set only has silver ink (I dropped gold-on-turquoise). I added three new ribbon colors - teal (darker than turquoise), dark green, and a deep blue that seems like it should match the indigo of Jameson's shirt design pretty well.

I did mixed-case versions of type options 1 and 3.

I personally prefer the all-caps versions, since I think the mixed-case capitalization makes the text look unbalanced (large letters all on one side, plus that one single descending 'g' on the other side); the all-caps also matches Jameson's shirt design.

Thoughts?

Chris@82:

Thanks for putting this together! My first choice would be line 6 in the dark green, though the teal on that same line is nice too. Sally is working, but I'll get her opinion later.

I think the wider two-line ribbon might be a good idea, as I think it's the largest ribbon anyone will try to make. It will help our ribbon stand out a bit when people have several on their badges. I think the silver text on the dark green really jumps out, as well.

Kilo

I agree with Kilo. The dark (forest?) green looks the best, and despite the fact that I normally HATE (heh) all-caps, I like that design best too.

Sally's suggestion is to use the two-line, all caps version, but to make all the dots much smaller, preferably getting rid of the dots on the second line entirely. She was okay with any of the bottom three colors.

Kilo

Chris #82: I prefer the darker colors, blue green or black.

Re the dots (Sally via Keith #85), I agree the dots should be smaller, and I'll add that I find the dots of varying sizes to be jarring (which may be connected to them being too big).

Oh, dear. I hadn't wanted to bring this up, until someone mentioned wearing this ribbon with others. Most ribbons are worn vertically at Worldcon.

This one is horizontally-oriented. (I know, I know. "Pluribus" is a large word, how can it possibly fit on a vertical ribbon?) But think how identifiable the pattern

E
Pluribus
Hugo

would be, she said, while pointing to the silver lining (and lettering). And pretend that the spaces I added before "E" and "Hugo" actually stayed.

@69 ::: Keith "Kilo" Watt

I will be happy for you to affix my name to that resolution.

Here is a rough justification:

Commentary: Testing and development is greatly enhanced by using real-world data. During the creation of the E Pluribus Hugo methodology, a statistically generated nomination dataset was used for testing purposes. It would be beneficial to use an actual nomination dataset for verification of EPH methodology and results.

@87: What about putting the words on a 45º downwards diagonal, to go with the T-shirt? You could still have them on 2 or 3 separate lines...

Jameson Quinn@89:

@87: What about putting the words on a 45º downwards diagonal, to go with the T-shirt? You could still have them on 2 or 3 separate lines...

No, honest, "Pluribus" will fit on one line, either in all upper case or in mixed case. It'll be fine.

Here's my latest mockup.

[Ribbons]

I like any of the three darkest colors, and either of the two-line versions.

As to ribbons being worn vertically at Worldcon, how recent is your information? I don't get to that many Worldcons, but I know that as recently as 2012, to a first approximation all the ribbons I saw were horizontal. And I do get to big Midwestern regional conventions, and the ribbons I've seen Worldcon bids giving away there are all, without exception, horizontal.
Perhaps I'm wrong; like I said, I haven't been to a Worldcon in a few years. Anybody have more recent experience?

Pfusand, #87: Most ribbons are worn vertically? I have to disagree; I see many more horizontal ribbons than I do vertical ones. Among other things, I vividly remember that my "Dealer" ribbon at Denver (2008) was vertical, and by the 3rd day it had been completely obscured by a string of a dozen horizontal ribbons which I had to lift out of the way to demonstrate my right of entry during dealers-only hours. And somewhere in my pictures from Reno (2011) I have one of a person with a string of horizontal ribbons that nearly reaches the floor -- some people make a point of collecting every ribbon they can, and that really doesn't work at all with vertical ribbons.

Jameson, #91: I like that T-shirt design a lot!

For horizontal ribbons, the two-line design works better than anything diagonal.

I think that so far, the strongest options are silver-on-indigo or silver-on-forest-green. I'd add silver-on-black as an option as being neutral ("goes with anything") without being as boring and stain-prone as black-on-white. On a 0-10 scale, I'd vote:

Silver-on-indigo: 10
Silver-on-green: 6
Silver-on-black: 8
As above with white or gold instead of silver:as above minus 1
Everything else mentioned so far or that I can think of: 4 or below. (IE, if somebody brings in a new option, assume my vote on it is 4 unless I say otherwise.)

What do you all think?

94
Silver on indigo, followed by white or pearl/silver grey on indigo.

#94: I pretty much agree with your color rankings.

Lee @93, Pfusand @87, Cally Soukup @92

In London almost all (maybe all - I do not remember seeing even one vertical) the ribbon were designed to be worn horizontally. I do not remember any Vertical ones - stacking horizontal ones on top of each other is easier at the bottom of your Worlcon pass (regardless if you have a few or if they are dragging on the floor when you are walking)... Plus with the sticky side on the bigger side, a lot less likely it to just get unattached I guess.

How about this as an addition to Duncan's #88:

Commentary: Testing and development is greatly enhanced by using real-world data. During the creation of the E Pluribus Hugo methodology, several statistically-generated nomination datasets was used for testing purposes. It would be beneficial to use an actual nomination dataset for verification of EPH methodology and results. This will also help dramatically in providing good, solid information to the members of Worldcon when EPH comes up for ratification in 2016.

While there will always be speculation and "what if?" scenarios, discussions will be more productive if they are based on reality and not speculation. Particularly in an election, transparency (so long as voter privacy isn't violated) is always a good and necessary aspect. Releasing the anonymized ballots will also put to rest any insinuations about impropriety in handling what has been an extraordinarily contentious Hugo season.

Regarding colors, are we talking about T-shirts, ribbons, or both?

K

Both, I think. I thought we'd agreed that we want things to match.

@69: we should probably add a "without connecting the separate categories of a given ballot" in there somewhere, in case somebody announced some categories but wanted to keep some categories private.

@98:
Also, as Standlee said at File 770, we should say: "This does not compromise anonymity, any more than manually recounting ballots would in any election."

"including the works nominated on each ballot in each category"

I don't think we want that. All we need is 13 datasets which identify the nominees consistently: thus "Ballot 101: Novella 1, Novella 12, Novella 15, Novella 20, Novella 21" is all we need, and there's no reason Ballot 101 in Novella needs to be the same nominator as Ballot 101 in Novel.

102
Number the ballots from 1 to whatever, in each category. Or random-number them between 1 and whatever-plus-ten, in each category.

Tim @102 and PJ @ 103:

Would that require the administrator(s) to manually replace the nominated work with "Novel 1", "Novel 2", etc.? If so, that is additional work above removing Personally Identifying Information (PII). They may not want to do that.

102, 103, 104: I don't think any of you have any idea how much work is connected with figuring out what the nominators are nominating. People are not picking from a list, or a pull-down menu: they're entering titles, sometimes with authors, sometimes with the place where it appeared, sometimes from memory and misspelled/containing erroneous words.

The original data set is ugly. Cleaning it up is a lot of work. Matches are not simple.

If you want to work with real-world data, you're probably going to have to work with real-world data.

Tom@105:

The original data set is ugly. Cleaning it up is a lot of work. Matches are not simple.

If you want to work with real-world data, you're probably going to have to work with real-world data.

I have to agree here. I would think that pulling off the name or Worldcon membership number or whatever is used to identify a ballot would be pretty easy, but I don't see any way to automate the process of figuring out that The Return of the King, Return of the Kng[sic], and Lord of the Rings, Book 3 are all the same nomination.

That said, I assume the Hugo admins do this at some point, don't they, Tom? I'm not 100% sure where in the process that clean up happens. If EPH passes and gets ratified, it may turn out that the very simplest approach is to have the current nomination database (once it's cleaned up) spit out a formatted file that EPH then uses. But for transparency purposes, I don't see how the Hugo admins can release anything other than the raw, uncleaned (but anonymized) ballots. When testing, I think we're going to have to do that clean-up ourselves.

Kilo

I don't know either, Kilo, but that's the way I'd bet. There may be something intermediate, with the mail ballots partially cleaned -- that data has to get entered, where the online nominations are already entered. It's still likely to be ugly, and it may be very ugly if you get really raw data.

106
It may not happen until it's absolutely necessary - the best tool still seems to be the human brain.

I spent five years doing data extraction, entry, and QC for a utility company, where making-it-match was critical. It could be really interesting as a figure-it-out, when the data went back decades and was subject to spell-by-ear as well as handwriting difficulties and simple street-name changes. Those were almost the least of the problems. (I was quite happy when I had to get into the extracted data some 25 years later, and found it was mostly still solid.)

Tim Illingworth @ 102

In a perfect world - yes. But I highly doubt that the nominations ever go down to that level - considering that only counts are important now (Throne of the Bad and Throne of Badness as only 1 nomination each are just fine to stay on their own under the current count but need to be collated together if they are the same one for EPH for example because it will make a difference in the early rounds - and if they are not, people may scream for unfairness even if it is easy to see that these eliminations will always be very early), chances are that some of the illegible scribbles that are obviously not one of the front runners never get resolved (I may be wrong but I had run an open ballot a few times and when something is obviously not going to have an influence, you just leave it on its own - you leave those ballots for last so you know if that thing starting with D is one of the works for which numbers matter.)

So technically, the easy way is to get the ratty data (moles, broken nails and all:) ) and have a volunteer team move it to a clean computer data set. We cannot ask the Hugo admins to do the work for us on that.

So the best we can ask for is to get the ballots and then if we want to work with the data, we work with the data.

Duncan @ #88 and Keith @ #98:

I think this request should stand completely separate from EPH, since the data is useful outside of EPH. Having that data not only allows for the validation of EPH as a system, but also for potentially crafting something even better, and retains that value even were EPH to be rejected.

I don't have a proposal for exact wording, but I'd keep it very general, and at most relate it back to EPH by a small phrase somewhere along the lines of "such as during the development of the solution which eventually became the EPH proposal".

106, 108 It strikes me that consistency of spelling of nominations may be the true mark of a slate and its use of copy-pasting. I worked with real world data too:(

I completely agree with Aan@110 here -- and the data should be made widely available if it's made available at all. Among other things, it will give enthusiasts for the various voting methods proposed at the start of the EPH debate a chance to test their hypotheses against a medium-sized real-world data set. That can only result in good information (which is Humpty-Dumpty speak for a real knock down argument, I fear).

I think it is an excellent idea to submit a resolution to the Business Meeting asking for the anonymized nomination data for this year to be released. I would be happy to add my name to it.

I think it would be a good idea to have it available for general testing of Hugo nomination algorithms. I think EPH is best, but if someone came up with something even better it would be good to have the data set available. Furthermore it would be morbidly interesting to verify the actual numbers of lock step voters among the Puppies. (I'm not sure that would be a selling point to the business meeting, though?)

As a fallback position, if the Hugo Administrators aren't willing to release the data, I wonder if they'd be willing to run EPH on it themselves? It wouldn't be as transparent, but it would still be a test of the system.

Cat@113:

I'm still hoping they will release the data (though it seems they might have to scrub the nominees that got less than 5% of the vote -- it's not clear). But if they don't, I also really hope they'd be willing to at lest run our implementation of EPH and see what would have come up. It doesn't given the open testing of EPH that I'd like to have, but I think it would at least make the point that EPH does what it claims to.

Kilo

All:

We need to submit our resolution for the ballot data. Here is the current version I have. Any last-minute comments? Also, if anyone else wants to sign, let me know (Cat, I need your real-world name and I'll add you).

Thanks!
Kilo
===========================
Short Title: Hugo Nominating Data Request

Resolved, That the WSFS Business Meeting requests that the Administrators of the 2015 Hugo Awards make publicly available anonymized raw nominating data from the 2015 Hugo Awards, including the works nominated on each ballot in each category but not including any information that could be used to relate ballots to the members who cast them; and
Resolved, That it is the opinion of the WSFS Business Meeting that releasing such anonymized raw nominating data after the announcement of the results of the 2015 Hugo Awards is not a violation of the privacy of members’ ballots.

Submitted By: Keith "Kilo" Watt, Jameson Quinn, David Harmon, Duncan J. Macdonald

Commentary: While there will always be speculation and "what if?" scenarios, discussions will be more productive if they are based on reality. Particularly in an election, transparency (so long as voter privacy isn't violated) is always a good and necessary aspect. Releasing the anonymized ballots will also put to rest any insinuations about impropriety in handling what has been an extraordinarily contentious Hugo season. This request does not compromise anonymity, any more than manually recounting ballots would in any election.

Testing and development is greatly enhanced by using real-world data. During the creation of the E Pluribus Hugo methodology, several statistically-generated nomination datasets was used for testing purposes. It would be beneficial to use an actual nomination dataset for verification of EPH methodology and results, as well as for any other nomination system which may be proposed.

[RIBBONS/T-SHIRTS]

Does anyone have a status report on these? It seems to me we're going to have to get these ordered soon if we're going to have them ready in time for Worldcon.

Let me know...
Kilo

115
Sounds good. (And not just because I was wanting a standard set of test data back in '84, for the counting program for the final ballots. You shouldn't have to rely on the existence of a prior data set, with results, to test important programs.)

I'll sign on to the data request. I'd really like to know what this year's ballot would have looked like if EPH had been in effect.

Keith "Kilo" Watt @115: I think nomination datasets was used should be nomination datasets were used.

I'll sign on the data request too.

I did mention over on File770 that this year's ballots would be of historical interest, though I'm not sure if including that reason would help or hinder our case.

I did not help in any way for crafting the data request but if that is not really a requirement, I'd sign it too. :)

[DATA RESOLUTION]

Here's the next version... (Once I get Cat's full name, I'll include that as well.)

Kilo

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

Short Title: Hugo Nominating Data Request

Resolved, That the WSFS Business Meeting requests that the Administrators of the 2015 Hugo Awards make publicly available anonymized raw nominating data from the 2015 Hugo Awards, including the works nominated on each ballot in each category but not including any information that could be used to relate ballots to the members who cast them; and
Resolved, That it is the opinion of the WSFS Business Meeting that releasing such anonymized raw nominating data after the announcement of the results of the 2015 Hugo Awards is not a violation of the privacy of members’ ballots.

Submitted By: Keith "Kilo" Watt, Jameson Quinn, David Harmon, Duncan J. Macdonald, David Goldfarb, Soon Lee

Commentary: While there will always be speculation and "what if?" scenarios, discussions will be more productive if they are based on reality. Particularly in an election, transparency (so long as voter privacy isn't violated) is always a good and necessary aspect. Releasing the anonymized ballots will also put to rest any insinuations about impropriety in handling what has been an extraordinarily contentious Hugo season. Perhaps more importantly, the nomination ballot data have significant historical value, and should be preserved in any event. This request does not compromise anonymity, any more than manually recounting ballots would in any election.

Testing and development of voting systems are greatly enhanced by using real-world data. During the creation of the E Pluribus Hugo methodology, several statistically-generated nomination datasets were used for testing purposes. It would be beneficial to use an actual nomination dataset for verification of EPH methodology and results, as well as for any other nomination system which may be proposed.

I'm happy with the wording.

If you haven't sent it yet, you can add my name, too. Lenore Jean Jones

Would it be a good idea to include in the resolution that the names of any works which don't qualify for the published nomination voting totals as per section 3.11.4 should also be anonymized? We don't particularly need to know what those works were (EPH is highly unlikely to promote any of them to finalist), and making the resolution as uncontroversial as possible seems sensible to me.

felice@125:

They could anonymize them or leave them out completely, I'd think. As you pointed out, either is fine for EPH. We can't mandate that they do anything, so we don't have to specify what they give us -- ultimately, they can release as much or as little as they want. But I can see the point that we might suggest that alternative to them in case they don't think of it themselves. (However, I'm reasonably certain the Hugo admins are reading this...) The question is, do we want to specifically write the request that way? Obviously it's easier for everyone if they just release the raw ballot file, but there is the question of what 3.11.4 permits and what it doesn't.

Opinions?

Kilo

Signing also. I came back to this thread today to propose that we put just such a resolution together, only to find that you guys were already on it. This version sounds good to me.

[DATA RESOLUTION]

Okay, I haven't seen much more feedback, so I'm going to submit in the morning unless anyone has changes. The one remaining issue was whether we should suggest to the admins that they could hide the bottom 5% if they felt that was a problem. Kevin doesn't think it is, but regardless, it's my understanding that admins are aware of the option, should they desire it. I think we're safe to leave it off.

Let me know if you have any more suggestions!

Kilo

[DATA RESOLUTION]

It's done then!

I just submitted the resolution to the business meeting staff, so we should see it online in the next day or two. As with the EPH proposal, we can still submit changes up until the deadline if we feel a pressing need to do so. Thanks to all of you for helping to put this together!

Kilo

I'm not sure if this belongs here (it concerns Hugo voting strategy, rather than EPH), but since this seems to be the only active Hugo thread here right now, I'd like to note that I've written a comment over at the Obsidian Wings Hugo voting thread on my thoughts about "How to Vote ABD (Anyone But Day) for the Hugo Editor Categories."

Finally done with my time-sucking work project. (I ran an event with nearly 2000 participants across our various global offices. It was awesome, but I was putting in 70-hour weeks at the end there.) I will be more responsive from here on out.

[RIBBONS]

It sounds like we have a pretty good consensus around this design (silver ink on indigo ribbon, two lines, all caps).

I'm willing to tweak it a bit more, but opinions seem to be more divided from this point (e.g. include the dots versus remove them versus make them smaller). Anyone feel particularly strongly one way or the other?

I for one am in favor of the dots here because I think they're an interesting visual element that will help the design stand out without being too busy, and because otherwise there's only the one dot in the first line which ends up looking a little out of place.

And finally: last call for ribbon requests...

131
[RIBBONS]
Go for it!

Chris Battey #131: otherwise there's only the one dot in the first line which ends up looking a little out of place.

Um, if you were to remove the dots, why not skip all three?

Chris Battey #131
I like the colours.
I'd take the top dot out, but leave the bottom ones for balance.

Actually, if I could make the bottom ones tiny rockets I'd try. What's the unicode for a rocket? :)

From a distance it'd be just a bullet point, but closer a little detail. Though as I haven't said anything about this before it's a bit late for me to start now.

If we were to make the dots rockets, then obviously we'd want to use proper Hugo rockets, and we decided not to go there because of respect for the Mark Protection Committee and its neutrality.
Personally, I like the dots, but if I'm in the minority, it doesn't bother me if they're not there.

If you've never attended a WSFS Business Meeting and plan to attend this one, I recommend watching the Business Meeting Basics video that I created with the help of the members of the Westercon 68 Business Meeting. This video demonstrates some basic procedures including how to be recognized to make motions and debate, and the different ways that we vote on motions.

Kevin, that's a great video. Thank you very much for making it.

David @ 133:

The dot in place of the space in "E PLURIBUS" would be to keep to the same style as Jameson's t-shirt (which does dots-for-spaces in the same way).

Doire @ 134:

You may be joking, but it's U+1F680. Some fonts have pretty nice versions of them, generally in the form factor of the emoji on that page.

[RIBBONS]

All -

Okay, I think we're coming down to the wire, so we need to make a call about ribbons. The argument for the design Chris posted in #131 was that it would match Jameson's t-shirt, but I think we've got some flexibility in whether and how many of the dots we include. The simplest thing, obviously, is to just leave them all in, since that's what the t-shirt design has. I think we could get away with pretty much any combination of dots, however.

So! Would everyone please vote for which (or all) of the dots you think looks best on the ribbons. I think the ribbons are very important, since that is one of the main ways we will engender support for the proposal, so we need to make sure we've got a design that stands out.

Thanks!
Kilo

[T-SHIRTS]

Jameson, do you know what the status is on these? Are we going to go with CafePress or some other company?

Kilo

[RIBBONS]
all the dots, like the one linked in 131, please.

Personally, I think the dot on the first line looks out of place (and unbalanced), even with the ones on the bottom. That is, I agree with Doire #134.) It might just be scaled too big. But as you say, we could get away with almost any combination -- I wouldn't worry too much about matching the T-shirt.

Posting for my partner, Martin.

Keith @140 I am not going to Sasquan, and I came in too late to the voting methods discussion to make any useful suggestions. But I support the EPH proposal, and I really, really want the t-shirt. Please post a link as soon as there is one available.

Thanks!

[RIBBONS]

Okay, looking back I don't see a clear consensus for changing the dots: P J and I say to keep it as it is; David and Doire say to drop the dot on the first line; Sally said earlier to make the dots smaller or drop the bottom dots entirely (and David pointed out that the different sizes of dots is jarring too); several other people who have generally approved of the design haven't commented specifically on the dot issue.

So I'm going to make a minor change to reduce the size of the larger dots on line 2 to match the smaller one on line 1, and otherwise leave the three dots in place. I'll make the order with that design, silver ink on indigo ribbons.

Thanks for the feedback, everyone.

That works for me, Chris. Thanks for taking care of this!

Kilo

[RIBBONS]

And the order is in! Should arrive at my place with at least a week to spare; I'll divide them into bags for easy distribution at the con.

Chris Battey #144: Sounds good, looking forward to seeing them!

I'm talking to Customink about silver for the T-Shirts, and whether glitter ink will work with such fine detail. I should have an answer today.

Great job on the ribbons! Looking forward to 'em.

Jameson Quinn @148, please confirm with them that their silver ink will last after repeated washings (I've had bad luck previously with metallic inks on t-shirts). That said, glad to hear the t-shirts are coming along; do let us know when they can be ordered!

So it seems that there aren't really good options for silver. The "flat silver" options wash out, and the "glitter" options do not work with such fine text (and for reasons I don't understand, don't combine well, so we couldn't even do glitter for just the larger text).

So, digital printed shirts cost around $13.68 per shirt, while silkscreened cost around $6.70 for orders of around 100. The disadvantage of silkscreening is that we'd have to know exactly how many of each size we'd want up front (and probably deal with any shipping ourselves, aside from the stuff to be picked up at worldcon.)

150
A light grey might work as a substitute for silver. It won't be as pretty as a metallic, but it should hold up better.

[T-shirts]

It seems to me a light grey would work as well. Is this a possibility?

Regarding digital vs. silk screening, I seriously doubt we would need 100 shirts, so it seems to me that digital is the way we'd need to go...

Kilo

I've ordered a pair of shirts; as I noted on the form, I won't be there, but I've got a friend who will be. When you know, please let me know here where and when would be good times and places to connect with you or your agents to pick up the shirts.

Ok, I've tried to order shirts twice. As far as I can tell it's not submitting the form. At least, I've not gotten any email receipt as I usually do. Do they give email receipts? I'm using Linux, so maybe there's something weird that's not working right. For the record, I'm trying to order a 2x and a Large.

154
i got one...of course, I forgot to put in shipping info.

P J Evans: Did they email you a receipt? Because either I ordered two sets of teeshirts or zero; based on the lack of an email receipt I'm guessing zero. (I was trying to order one set.)

156
Yes, it was almost immediate. I got a screen, too, saying that it had gone in. (I did order two - one for my sister.)

Thanks. Looks like I can't order it on my computer. I'll get my sister to do it.

My shirt order is in, thanks Jameson!

K

Whoops, it seems that the order form link failed to post here. Here it is:

Please order your E Pluribus Hugo t-shirts ASAP using this link. They cost $8.04 per shirt, if you're there at the convention to pick them up. If not, please still order now, and we'll figure out what shipping costs later (probably a reasonable amount; but it certainly won't be free, since we have no profit margin on the shirts themselves).

In the notes on your order please tell us:

-Your name and email.
-If you will be picking the shirt up.
-If so, which day you'd like to get it.
-If not, your mailing address. We'll contact you by email to work out shipping payment.

As for the receipts: I'm not sure exactly how it works, but the payments don't go through until I finalize the order. But I don't see Cally Soukop on the list of people who have made orders.

It's been a while since I was able to check in (real life has been interfering). I've ordered my EPH shirt, and I'm still in for 100 ribbons, if Christ still had me down for them (if not, I promise I won't cry).

Thanks so much to Kilo, Jameson, Chris, and everyone else for all your hard work on this. I'm really looking forward to Worldcon.

Jameson, I also thought I placed an order, but have not received a confirmation. Should I try again?

Jamison, when Cally's order failed, I ordered her shirt and mine from my Windows machine. So you should have a receipt from (rot-13) Pynhqvn Ornpu - if so, don't worry about Cally's order; it's covered. (I'm pretty sure you have mine; I got an email confirmation.)

Oops, Chris -- sorry about the inadvertent promotion in status! ;-)

A question not t-shirt related: Does EPH have a tame parliamentarian? If not, can you get one? While, in general, my understanding is that WSFS tries really hard to make sure that voices are heard, Robert's rules can be arcane, and some of the players are experts. Having a tame parliamentarian might be helpful.

And, t-shirt related, my second attempt went through! Yay!

166
I think Kevin can handle it.

[RIBBONS]

Ribbons are in! They turned out to be a slightly lighter blue than I'd inferred from the online image - more like a "true blue" than an "indigo", but I think they turned out really nicely; the silver pops out against the blue like we'd hoped. I'll try to get a picture up soon.

When I was computing estimates I forgot to account for shipping, which bumped the cost for the whole 1000-ribbon order just over $200, so final costs will be an even $20 per 100. (And yes, JJ, I have a hundred reserved for you.) I'll get the handful of individual ribbons requested in the mail soon too.

Chris, is it possible for those not attending to get a ribbon? I seem to have missed the sign up...

Thanks Chris! Can't wait to see the ribbons. How do you want to be paid for them?

Kilo

[RIBBONS]

Kilo @ 171: Cash at the con would be easiest; if that doesn't work I can take a check.

Lori @ 170: I'm happy to mail small numbers of ribbons to people who can't attend the con - I have a request form where you can provide your mailing address. I'm not asking for payment for the mailed ribbons, but feel free to donate a buck or two to your favorite charity if you want to pay for them.

[MEETUP]

Abi has posted the Sasquan Gathering of Light thread, but in the interest of not monopolizing the conversation at those gatherings with EPH business, I think we should meet separately as well.

It would be nice to meet before the first business meeting (Thursday at 10 AM), unless we're sure that the EPH proposal won't be taken up until Friday. Who will be at the con in time for dinner on Wednesday?

I arrive in Tue night so I can be there for a Wed dinner.

Chris, I remember I'm in for $10 worth of ribbons. I'll be at all the business meetings at Sasquan, do you want me to pay you there?

.

Jameson, I've also completed an order for two shirts to be delivered at Sasquan. The receipt showed up immediately in my in box.

.

Anyone with any stake in any business on the agenda really ought to come to the Thursday meeting and not wait until Friday. The EPH proposal could very well come up on Thursday in an attempt to remove it from the agenda.

The Thursday meeting is where someone can propose to have any item on the agenda postponed indefinitely. It takes a 2/3 vote to postpone an item (that is, remove it from the agenda entirely for this WorldCon, in which case it's dead unless re-proposed in 2016). Every vote to KEEP it helps.

We are getting in on Tuesday also (well, Monday night but that's being spent with family).

Ive been following this and previous threads and wanted to say how much I appreciate all the work that's gone into EPH. Thank you!

Yep, at the business meeting(s) is fine. I'd much rather just take cash at the con than mess around with PayPal and such.

Xopher and I arrive Tuesday morning, so Wednesday night works.

We won't be getting in until fairly late Wednesday night (definitely after dinner time). I could meet for breakfast, though...

And cash at the con for ribbons works for me, Chris, thanks.

Kilo

PJ Evans @ 168: I absolutely trust Kevin to do well by everyone at the business meeting. I was just thinking that his hands are a bit tied by being the chairman, and that if things go deeply into the Roberts Rules Weeds of Order, a tracker might be helpful. I certainly didn't mean to cast aspersions on Kevin or on the business meeting participants.

Lydy@179: If things get overcomplex, that's Jared Dashoff's role as Deputy to Kevin. So it's handled no problem.

[RIBBONS]

Here's a picture of a ribbon! Like I said, a lighter blue than I'd expected, but I'm still quite happy with how they turned out.

Looks great, Chris!!

K

Oh yeah, those will be very visible and well-suited to the purpose!

I'm also getting in on Tuesday, so I'm good for a Wednesday night strategy session.

Any chance we could plan on Thursday morning? I'd obviously like to be there...

Kilo

The problem with Thursday evening is that there is a chance that it will be all over by then - the Thu meeting has an option to remove the proposal from the agenda for this Worldcon. Not very likely but...

I could (groan) do Thursday morning, sure.

I'm almost certain that there will be a motion to Postpone Indefinitely (kill for the duration of the current Worldcon) EPH. That takes a 2/3 vote against consideration at the Thursday meeting. There will be 4 minutes of debate on the question, with 2 minutes for those who think the matter shouldn't be considered and 2 minute for those who want to consider it. The lead sponsors of EPH need to be prepared to succinctly debate the question of "Why should we even consider this?"

Thanks for the head's up, Kevin. I'll make sure we're prepared.

Annie, I was actually referring to Thursday morning not night. I won't be arriving until after 11 PM on Wednesday, I'm afraid.

Kilo

So any supporters who can be at the Thursday meeting should be, to be Rhode Island. :-)

Lee@190:

Yes, I think as many people as we can get on Thursday and Sunday for the final vote would be a good idea. I think if Kevin decides to call a committee of the whole for EPH Q&A on another day we'll have that covered. My wife and I are making this trip pretty much strictly for the business meetings (and the parties :-) ), so I'll be there every morning. I consider the other panels I make to be bonus...

Kilo

I think the question of "Why should we consider this?" can be answered quite easily in well under two minutes: "This year's Hugo nominations prove that slates are a problem. E Pluribus Hugo is the best available technical solution. It reduces the power of all slates impartially, while imposing no additional burden on nominators, and only a very small one on the administrators."

I wonder if it'd be useful to mention that the exploitability of the current system has been (viz P J Evans in several comments throughout the discussion) known to awards admins for decades? - that is, the Puppy slates didn't break the system, it was already broken (or rather, is inherently broken).

Alternately, this could be exactly the wrong direction to take; some folks might reason that if it took this long before anyone exploited the hole, the existence of exploiters must be a unique and transitory phenomenon. (This, I suspect, is the perspective that a certain prolific naysayer is working hard to cultivate.)

For me personally, the former was significant in shifting me from a reluctance to mess about with a system that had worked well until now, to a conviction that EPH was the best way to ensure that it kept working. But that might just be me.

SunflowerP@193 Thanks for the insights into what convinced you that EPH was the way to go. I think these kinds of insights (along with the elevator pitches we were developing up-thread) are invaluable in preparing for the preliminary meeting. Please keep them coming!

Kilo

On the question to Postpone Indefinitely, the people who want to kill the proposal without further debate get the first shot, but the "burden of proof" is upon them to establish why we should suppress the proposal, and they need a 2/3 vote to do so. This is quite different from the main debate (if you get that far), where the backers of EPH have to convince a majority that that the change is worth making.

And Lee: "I saw what you did there, Mr. Hopkins." (This reference will make no sense to anyone who hasn't seen the musical 1776. I have for many years considered a fannish version of it built around the WSFS Business Meeting, and I have several of the musical numbers written. Because my name scans into "Richard Henry Lee," I'd have to cast myself in that role, which is something of a pity because Lee leaves at the end of Act 1 and is not heard from again.)

193
I gather from comments at File770 that there was a much bigger problem in 1989. And I really don't think ballot-box-stuffing and slates are improvements to the Hugos in any way.

Kilo @189

Oops. I misread. Apologies. :) Chances of me making a meeting before the BM are very slim (not a morning person... not at all. Especially in days I am not working) but if everyone else is up for it, I may try...

SunflowerP@193,

I do not think that the current system is broken and I am reluctant to use the term. It is exploitable - as is any system that is honor based (and/or based on unwritten rules). But saying that it had always been broken will alienate people - it did work for quite a lot of time. Just my 2 cents.

Yes, "exploitable" is a better term than "broken". The system has always been in the nature of a gentleman's agreement. The problem is that any gentleman's agreement works only when everyone involved with it behaves like a gentleman. Or, as David Gerrold put it, because "nobody wanted to be that asshole".

Well, now somebody wants to be that asshole, and has declared their intention of continuing to be that asshole, so now the problem is no longer temporary and needs to be addressed.

Kevin, #195: I figured there were enough 1776 fans here who would get the joke, but thank you for the explanation anyhow. I couldn't find a YouTube clip of that particular scene to link.

Lee @190:

"New York abstains...courteously."

Annie@197, Lee@198: Yeah, that's a technical usage of 'broken' that doesn't carry over well to general use (and now I have a deja-vu-ish feeling that this very point came up way back in these discussions, quite possibly in the Schneier-led voting systems threads); 'exploitable' is definitely better... not least, because it puts the focus right back on 'there are people who are exploiting the system'.

So what are the plans for meeting? Wed dinner? Thu breakfast? Both? None?

On a separate topic - I just realized I never put my name for the ribbons - oops... Anyone have at least one to share for my badge (and I will pay for it of course) or 50/100 so I can distribute some while meeting people(again, I will pay for them)?

Annie@201:

How about we plan on breakfast on Thursday morning? Shutters Cafe is in the DoubleTree, which puts us close to the convention center. The business meeting starts at 10:00, so maybe we should meet at 8:00? The big thing to discuss is to be ready for the preliminary session as well as just trading contact info. If Kevin calls a committee of the whole, we'll need to make sure our ducks are in a row for that, but otherwise, it's mostly just talk up EPH for the vote on Sunday.

Thoughts?

Kilo

If we must... :) Now more seriously - sure. 8 am is fine. I am staying in the Grand Hotel on the other side of the convention center so easy enough to get to the DoubleTree.

Kilo, #202: Should be doable, although I won't like it any more than Annie does. :-)

Same here.

Will we be able to get our tshirts at breakfast?

The PBM looks busy: Kevin has some estimated timings

So it looks like the committee of the whole will be on Friday. Good to know... I have a power point ready for that. I'll send it on to the businesses meeting staff (they need it installed ahead of time) Monday morning. For Sunday I'm planning a very brief summary of the technical aspects of the system, then focus on why we need EPH. I'm putting that one together tomorrow, so I'll send it on Monday as well.

So breakfast is a go on Thursday at 8:00. I'm the tall blonde guy with hearing aids, so I'm pretty easy to spot, but I'll let the cafe know to expect people. Just ask for the EPH group. Hopefully we can get a supply of ribbons and our t shirts as well!

I think we're good to go; am I forgetting anything?

Kilo

Good luck! I am sorely tempted to rewrite a certain famous farewell speech to apply to the Quest of the Algorithm-Bearer which here begins ...

I'll plan to be at the breakfast Thursday also. My son and I are also staying at the Grand.

I've got a couple of thoughts related to the Business Meeting. First, just in case there turns out to be opposition to the possibility of a perfectly disciplined slate eliminating all its members in a perfect tie, would it be good to have some amendment language available to break such a tie alphabetically, or something like that? To be clear, I prefer the way EPH stands as submitted, as (1) this will probably never happen in practice, and (2) I don't mind having a slight disincentive for putting together such a slate in the first place. What I'm suggesting would just be a "have language available in the back pocket" possibility in case it turns out that a significant fraction of the meeting objects on this point, and we needed to amend to get their votes. But maybe this is too remote to worry about?

Second, am I correct that a motion to Postpone Indefinitely does not apply to Business Passed On from the previous Worldcon? The effort to get EPH explained to people has convinced me that A.1 Popular Ratification is a bad idea for the future, although it wouldn't apply to EPH itself or other proposals voted on at Sasquan. I'm just appreciating how hard it is to explain anything at all complicated even when you can get everyone in the same room for a presentation, and how much harder it would be with a diffuse electorate who aren't even meeting in person or reading the same blogs. I don't want to defocus attention on getting EPH passed, though, so even if such a motion (to Postpone Indefinitely A.1) was in order, it might not be a good idea.

David @209: Are we all at the Grand then? My wife and I are booked there.

As to PI: that was certainly the intention. However, if you could get 2/3 to PI something at the PBM, you should have no trouble getting 1/2 to defeat it at the MBM.

Maybe Rule 5.2 needs a tidyup?

As the person who wrote the current version of Postpone Indefinitely, I would say that I do not think it was ever intended to apply to constitutional amendments pending ratification, because PI is intended to be the kinder, gentler version of Objection to Consideration, and we've never allowed an OTC against a pending ratification.

However, as I am not going to be presiding over the ratification debate (because I'm one of the lead sponsors), I won't be in a position to rule on it; the Deputy Presiding Officer will do so if necessary. If you are opposed to PR's ratification, I think you are better off focusing on defeating its ratification at the main meeting.

Now that I've slept, I think that Rule 1.2 covers this: "Rule 1.2: Preliminary Business Meeting(s). ... The Preliminary Business Meeting may not postpone consideration of a Constitutional amendment beyond the last Preliminary Business Meeting. "

PI would seem to violate this :-)

Where are we breakfasting on Thursday? And I requested ribbons long ago on some previous thread (50 of them, I think) - am I still on the list?

Tammy@213:

Breakfast at 8:00 at Shutters Cafe, which is inside the DoubleTree hotel.

Incidentally, my wife and I (and our friend) are staying at the Davenport Tower, which is a bit of a walk from the convention (though it is the "party hotel", supposedly). I'm not sure if that's the same as the Grand or not...

Kilo

David@209:

I can have something on standby if people have a serious problem with eliminating perfect slates, but honestly, I think it will be a non-issue.

I'm also opposed to popular ratification. There are just too many people who don't have time to get informed on the issues, yet would still be able to ratify or kill an amendment. I'm all for the internet age, but these proposals affect all of Worldcon, not just the Hugos. If you aren't coming to Worldcon, i really don't think you should be dictating how it's run. The Hugos are for everyone, of course, but the convention should be for, well, the people at the convention. Just my thoughts, though.

Kilo

Keith @214: The Historic Davenport is not the same as the Davenport Tower -- they're kitty-corner across the street. The Davenport Grand is completely different from either: a newly-completed very Modern hotel, which is connected to the convention center by a skyway (and is therefore a bit under a mile closer to the convention center than the other Davenports).

Looking forward to meeting you this coming week!

David Wallace @209: "just in case there turns out to be opposition to the possibility of a perfectly disciplined slate eliminating all its members in a perfect tie, would it be good to have some amendment language available to break such a tie alphabetically, or something like that?"

I'd be strongly in favour of that. A bigger problem than a perfect slate is a coincidental tie between a pair of non-slate potential finalists, which would be a rare occurrence but probably will happen at some point (something like two Doctor Who episodes being the most likely candidates).

I'll have a friend picking up my teeshirts at the con; please mention here as soon as you know good time(s) and place(s) for him to pick them up for me.

Good luck guys! I'll be rooting for you from afar.

Catching up from a weekend offline...

[MEETUP]

I will be at breakfast (potentially a couple minutes late - mornings are hard), and will bring the ribbons there!

[RIBBONS]

Annie Y @ 201: I now have a hundred ribbons set aside for you. (I bought extra ribbons to get us up to the cheaper per-unit price at 1000 and also to have spares for last-minute requests.)

Tammy @ 213: You are still on the list for 50.

I failed to mail out the individual ribbon requests this weekend; I intend to do so tonight. Sorry for the delay!

Chris Battey,

If you are arriving Wed or earlier and you would rather give me the ribbons before Thu, we can arrange a meeting of a sort. Not at 8 am on Wed though please :)

I'll be arriving Wednesday afternoon and would happily distribute ribbons that day as well; I don't know my specific schedule yet, as it depends (on that day in particular) on the mood of a three-year-old, but I'll probably be attending at least some of the First Night festivities.

[T-SHIRTS]

We just a call from the Davenport saying that the box of T-shirts arrived safely. They are holding it at the desk for me, and I'll get it when I check in tomorrow night. I'll bring them to breakfast Thursday morning.

[PRESENTATIONS]

I submitted four PowerPoints to Donald Eastlake, who will be running the AV equipment at the Business Meeting. He tested them out on the computer that will be used, and they all seemed to be fine. There is a PowerPoint for Friday (a longer one, intended to be used as reference for the Q&A/Committee of the Whole) that focuses on the mechanics, a shorter canned one for Sunday that focuses on the question of "why should we do this" (plus some selections from the FAQ, since there won't be any Q&A at this session), an animation showing the entire 2013 data run, and then a PowerPoint that contains our EPH proposal text, the data request proposal, and the "6th place" and tie-breaker amendments if we need them.

So, if my plane goes down in flames, someone knows what's available... ;-)

Thanks,
Kilo

Everyone at Sasquan, enjoy yourself, have fun, and good luck in the business meeting/s.

I am here, and will be attending the First Night party (which started at 4, but I'll be there at 6ish). I'm in a blue denim kilt and a blue Deathly Hallows shirt. I have a few of the ribbon packs with me.

Reminder: when you know where and when would be good time(s) and place(s) to pick up teeshirts, please post it so I can text my local agent. I have no idea what his convention schedule is (nor whether he's there yet), so finding out as early as possible would be good.

Hm. You know, I was going to have my t-shirt shipped, but it occurs to me that I know Lee, and she's going to be there. I should see if she can pick it up for me.

[T SHIRTS]

I'll have some t shirts at breakfast tomorrow if you want to get them then. 8:00 at Shutters Cafe in the DoubleTree. If I had an estimate of how many to bring, that'd be helpful. I don't know how big the box is, so I don't know if it's easy to transport. I'm still in PHX airport, getting ready to board, so I'll be offline soon.

Kilo

Honestly, I don't know when my agent (Joe P) will be awake tomorrow, and it's awfully late to text him. Is there another time/place that would work?

Cally: Well, I'll obviously be at the business meeting from 10:00-12:45 in CC 300B, I'll be ibn touch here if we need to coordinate further.

All: Made it safely into Spokane, and we're now at the Davenport Tower. I have the shirts in hand, and they look outstanding!

Hope to see some of you in the morning!

Kilo

Keith: I'll text him. How will he know who you are?

The journey of a thousand nominations begins with a single business meeting...

Excelsior!

Then again, there prolly won't be a mob at this first business meeting. I've texted him; we'll see if he shows up.

Jameson and I are in Shutters in the back. I'm wearing a green crew shirt and have my us navy leather flight jacket.

Kilo

Ok, I've texted him your description. He's a tall white guy medium build with medium-short brown hair and the usual clever fannish teeshirt. Not helpful, I know. (Don't know which shirt, but that's the way to bet)

Tammy: Jameson is wondering where he can store the t shirts. Can you give us directions to your table? You can text me and I'll relay.

K

Sorry guys, a migraine got me out today and I had to stay in a dark room for most of the day if I did not want a bad headache for the whole week. Will see you all at the business meeting tomorrow morning.

Annie: Hope you're feeling better!

All: First day went well. As predicted, there was an attempt to kill both EPH and 4 and 6, but both survived. There was then a proposal to substitute EPH's amendment for one that would send EPH to a study committee. The best rebuttal was made that we've essentially been in committee for months. That one was voted down as well.

The Hugo admin came out against EPH because of its complexity adding to the workload. I made the point that it doesn't change anything the admins do now, but he wasn't buying it. Still, that's definitely a point that needs to be emphasized tomorrow.

Final vote on EPH will be Sunday, as expected. The only possible wrinkle is that my plane leaves at 3:00, so we should really leave for the airport by 1:00. If they finish on time, there won't be an issue, though. It's 25 min to the airport, so I can probably push it if need be. There are plenty others of our group here if needed, and I have a feeling a lot of people will be making up their kids ode way or the other tomorrow.

Kilo

For those who haven't yet gotten the word, EPH survived the motions at the preliminary business meeting today, and will have a final debate and vote on Sunday. The business meeting will go into committee of the whole tomorrow (Friday) some time after 11AM for up to 1/2 hr to discuss it and the 4/6 proposal with relaxed debate rules.

We knew that EPH was likely to face a motion to postpone indefinitely today, which would have required a 2/3 vote to pass. What we (or at least I) didn't anticipate was that there would also be a motion to amend it by substituting a new motion to refer the whole issue to a committee for another year, which would also have killed EPH for this year. That motion only required a simple majority - fortunately, it didn't pass. There was also a motion earlier in the meeting to take up the proposals in the written order (order of submission), rather than the revised order proposed by the leadership. That motion wouldn't have killed EPH directly, but would have made it harder to get enough time to explain and debate it.

For those who couldn't be at the Preliminary Business Meeting, the video recording is available on YouTube - search on "2015 wsfs business meeting". The video team was very good about getting the segments up promptly - the first few segments were put up while the meeting was still going on. Most of the EPH manuvering is probably in segments 4 or 5 of the Thursday session - the final vote on EPH-related motions came about 9 minutes before the end of the meeting. You should also be able to follow subsequent sessions of the business meeting via this YouTube channel.

Kilo,

Yeah. Thanks. I was out and about by 2 pm. As long as I do not push it and do not try to run around despite it, my migraines tend to subside in a few hours. :)

The only additional work for the admins would be that all titles that are illegible or cannot be determined cleanly need to be resolved before EPH starts its work. That should not be that hard. And technically even under the current rules, they are supposed to anyway. Unless if the counting at the moment is by hand and the ballots are never entered electronically. Which if true, maybe should be changing anyway.

Oh well - I should be at the meeting tomorrow. Greta work today, guys.

Kilo and David are too polite to mention that Ben Yalow was leading the effort to kill EPH. The "refer to committee" substitution was, obviously, a transparent effort to get a "postpone indefinitely" passed with only a simple majority (a rules exploit worthy of the Puppies themselves, and something that probably also needs fixing), but Yalow also seems determined to waste as much of the Business Meeting's time as possible, calling for votes he knows he can't win, and so on.

I used to like that guy. Not now.

Huh. I would have expected better of him.

It's possible that he just really doesn't like EPH. Though I'd have preferred that to result in a discussion rather than parliamentary maneuvers.

243
A discussion would certainly be more honest than trying to kill it without discussion and by sneaky means.

Yalow wasn't the only one to try to kill a proposal through amendment-by-substitution. Kate Secor tried to kill the electronic signatures amendment that way, and I think Kent Bloom tried to kill 4 and 6 that way. Agree that the tactics remind me of the Puppies: just because a procedure is within the rules doesn't mean it's acceptable.

At the moment, I'm inclined to propose a change to the WSFS Standing Rules at next year's Business Meeting to ban the tactic, on the grounds that the Preliminary Business Meeting is not the place for debating proposals: either kill it by a strong 2/3 majority without debate, or let it go to the Main Business Meeting. But there'll be plenty of time to discuss next year's Business Meeting after this year's is over.

And the "amend by substitution" for an entire proposal, where the amendment is equivalent to a "postpone indefinitely," ought to a) require a two-thirds vote, like the PI, and b) be out of order when a PI has already been debated and defeated.

Ben Yalow is certainly not a Puppy. But he's carrying their water, and I'm going to remember that.

Will the t-shirts be anywhere like the dealer's room where someone can pick up the shirts we ordered?

Popular ratification - if that's what they mean by "2+1" - failed, 60-something to 90-something.

EPHers:

My phone died, so I'm on a borrowed one...

For all who can, let's meet tomorrow (Saturday) at 8:00 at shutters cafe. I'd like to talk about where we are and some plans for Sunday.

The presentation today went really well! Thanks to all of you for helping us get some momentum. The 2016 Hugo admin wants EPH to happen (though in fairness, he'd rather wait a year) and wants to work with us, so no matter which way it goes on Sunday, I think EPH is going to be a reality. Now to see if we can get it passed this year rather next. Now that we will have 2015 and 2016 data (and likely 2014 as well), we will have even more to work with. I was stopped a number of times at the con with well wishes, so I really think we made an impression.

Thanks again for all you've done!
Kilo

All:

Slight change in plans: let's plan on getting together Sunday morning instead of tomorrow (Saturday). It occurs to me that I'll have the 2015 Hugo data by then. The big thing we do need to discuss is a backup plan if the Sunday meeting goes over and I have to head to on the airport. We can still at Shutters at 8 on Sunday. If anyone has any issues that really should be discussed tomorrow, let me know here, otherwise we'll just edit until Sunday.

By the way, I gave the code to someone at the business meeting, and he coded his own version of EPH in just a few hours. He said he gets the same results we do with our test data, so looks like his code is good to go.

Thanks,
Kilo

Here is the youtube playlist of the Sasquan Business Meeting. I expect other segments will be added as they become available.

And Kilo, I thought you did a great job at the meeting (and it was also nice to finally put a voice to a name; the video wasn't hi-res enough for proper facial recognition though it caught your gestures well).

The person who did their own implementation of EPH in Perl was Galen Charlton (who also posted his findings in the other thread). He has also posted a Visual Basic implementation of EPH.

#WSFSBM appears to be the Twitter hashtag for the Sasquan Business Meeting which I have been following with interest. I have also been enjoying Racheal Acks' liveblog of the meetings.

Wishing you all the best for Sunday!

Edit: Whoops! The Visual Basic implementation is Kilo's which Galen posted with permission.

P J Evans - 99 to 69. And yeah - this was the 2+1/Popular Ratification. :)

Kilo - see you at Shutters at 8 in Sunday then. Great work today!

Unofficially it's Helsinki in 2017. If you can get someone from their bid to confirm they wouldn't find EPH too onerous it could be a big help.

If there is anyone from that bid sober enough to speak that is. :)

Doire @254:
If there is anyone from that bid sober enough to speak that is.

I've gone drinking with Finns. The final result is usually that even the Poles and the Brits (the other champion imbibers in my usual context) are under the table while they still sit there, smiling and chatting.

But maybe let them have their moment of good cheer before the politics resume.

I have never had the enthusiasm, while at a Worldcon, to attend a business meeting. All this procedural stuff (and Roberts Rules is also something rather American and alien to me) is a bit offputting¹.

If somebody wants to claim there's a faction of BNFs who want to control the reins of power, I'd find it hard to argue against them. But I don't know enough².

What's the Rule Zero for Worldcon Business Meetings? Right now, it's looking like "not even death shall allow change to happen".³

^{1: And I still a member of the Durham Union Society, so I know about snobbish debating rules being used to diminish the outsider.}

^{2: I admit to being surprised to hear that Ben Yalow is still active. How old is he now? All I can find are undated references to 35 years of fannish experience.}

^{3: And I bet you were expecting a footnote in Small-Caps.}

Chris Battey @12: My two EPH ribbons just arrived in the mail, thanks!

And today, PDX is hazy and ashen with smoke from all the fires, so now I really feel like I'm there at Sasquan.

The Five Percent Solution amendment, which I co-sponsored, eliminating the requirement that finalists appear on 5% of nominating ballots, passed by what seemed to be a very large majority. (I couldn't see hands behind me, so I can't be sure, but the mood of the room seemed to be in favor. And I think my open speech in favor went well and got a laugh at the right time.)

One problem, pointed out by Perrianne Lurie: it was ruled yesterday in connection with the resolution to release nominating data that the final clause of 3.11.4, "but not including any candidate receiving fewer than five votes", modifies only the second half of the previous clause and not the first part. Except our amendment deletes the second half of the clause, so now it appears that it *does* apply to the first part, and that we may have inadvertently modified to the Constitution to *forbid* the Hugo administrators from releasing data on works receiving fewer than five nominations. Unfortunately amendments to fix this problem were ruled non-germane (and my attempt to formulate an impromptu argument in Perrianne's support devolved into incoherent babbling; lesson hopefully learned, don't rise to speak unless I'm sure I can make sense), so we may need to consult with the Nitpicking & Flyspecking Committee to make sure our amendments allows, but does not require, the administrators to release that data.

If there's anyone at the con who has a nearby hotel room or similar space where we could meet to run the newly released 2015 data Saturday night into Sunday morning, please email me (firstname dot lastname at google's email service). Or you can of course buttonhole me here at the con...

257
Mine, also. We don't have smoke, but we did have a nice fogbank to cool the place. (Yesterday was lovely, especially after the three-digits-Fahrenheit temperatures earlier in the week.)

Steven desJardins @258: "we may have inadvertently modified to the Constitution to *forbid* the Hugo administrators from releasing data on works receiving fewer than five nominations"

I don't see that as particularly problematic. Nothing with fewer than five nominations is ever going to be a finalist under any plausible system, and any such works can be anonymised when releasing ballots.

My ribbons arrived today, as well. Thank you! I'm now wearing one while settling in getting ready for the Hugo ceremony in a couple of hours. I understand the livestream starts an hour before the 10pm Pacific time ceremony does and will continue for an hour afterwards.
Livestream here: http://www.ustream.tv/hugoawards
Text livestream (liveblog?) here: http://www.thehugoawards.org/

Eeep! I meant 8 pm Pacific time! Hope I didn't confuse anyone!

Cally @262:

One correction: the livestream address is:
http://www.ustream.tv/hugo-awards (hyphenated).

There: that's enough colons.

All:

Just a reminder that we are going to meet at Shutters Cafe Sunday morning at 8:00. Anyone who wants to support us for the final vote, please be sure to be in room 300A at -10:00-!

I have not yet been given any data, and I'm not entirely sure how they will get it to me. We have been asked that anyone who wants a copy of the data should ask for it from the admins directly, so I don't think I can make copies if/when I get it. It may be at the business meeting tomorrow before I get it, but I'll keep you posted.

See you tomorrow!

Kilo

A pretty decent set of Hugo results under the circumstances, but the voting figures are worrying in some respects. The number of Puppy voters seems to be substantially higher than the number of Puppy nominators, eg 586 voted for VD as their first choice in Best Editor Short Form. If many of these new Puppy recruits nominate according to a slate next year, they could dominate the ballots even more than this year.

Felice@266:

Good observation, I think you're spot on, 586 vs.162. So, if nothing else, the puppies are right that they brought in a significant number of new voters.

All:
It looks like we won't be getting the ballots this evening. I was invited to the Hugo Losers Party, but after a very long (and very inebriated) night last night, I decided I'd better actually get some sleep so I can be sharp for tomorrow. My wife is going, so she'll fill me in on anything interesting that happens.

In other news, I had a very nice conversation with Wendy Delmater (Abyss & Apex editor and one of the Sad Puppies luminaries). She very genuinely wanted to hear about EPH. I explained it to her, along with some of our guiding principles, and she's decided we've got her vote tomorrow. I've been wanting to have a reasonable conversation with the Sad Puppies, so it was nice to finally get that chance.

Wish us luck tomorrow, and if you know anyone who is at the con, please ask them to come by Room 300A tomorrow at 10:00 local to support us! At this point, I think it's too close to call, so it will all come down to the final debate.

Kilo

267
Watch out for those who would claim that No Award winning means we don't need EPH.
What we don't need is a fan revolt every time someone decides to stuff the nominating box.
(There are puppies over at File770 claiming both that this year means they won and that they're going to do it every year until we give them the awards they think they need.)

PJ@268: The quick counter-argument is "do you want to have to hand out a bunch of 'No Awards' every year until the puppies give up?" Particularly since there are several categories where the eventual winner was only on the ballot because other entries withdrew or were disqualified. Looks like Best Novel (where both of the top 2 finishers were not in the original top 5), Best Novelette, Best Dramatic Presentation (Short Form), and Best Fanwriter were all in this category.

In addition, Best Pro Artist, Best Fanzine, Best Fan Artist, and the Campbell all had the eventual winner in the original #5 position, although Fan Artist can't blame that on the slates, who didn't nominate in that category. In short, we came really close to having almost everything that won not appear on the ballot at all.

2015 Hugo summary data appears to be here.

Brandon Kempner's analysis puts the total number of Rabid voters at a bit above 500, all of whome are able to nominate next year.

This is why we need to pass EPH. All the best for tomorrow Kilo!

(Next year's nominations are likely to be at least as bad as this year. As long as we can get EPH passed & ratified, from 2017, the effect of bloc-voting would be greatly lessened.)

EPH passes 186-62. Sunset clause added, will need re-ratification in 2022. Needs to be ratified next year to first go into effect in 2017.

EPH passes!

186 for / 62 against.

http://katsudon.net/?p=4333
http://file770.com/?p=24507&cpage=3#comment-324914

Five year sunset clause added.

Thank you to everyone for all your efforts!

Congratulations everyone who worked on this - you have done a FANTASTIC job and very much deserve that vote!

And 4/6 passes - as 4/6.

I'm sitting on the plane while they take care of a maintenance issue, but I wanted to thank all of you for the congratulations, but most especially for all the hard work you've all put into EPH over the past four months. It's been a long road, but it has truly been a team effort. -None- of this would have happened without you. So I want to thank you all; we have more to come, but we've passed the biggest hurdle. You've saved the Hugo Award. And that is no small thing.

With highest regards,
Kilo

To Kilo and all who worked for EPH -- Thank you!

The four Business Meetings sessions, put together, come to about eleven hours, and I was there for almost all of that. A couple of hundred people (not all of them on our side) did the same. I've seen people sneer at the idea that we should care about awards, and I won't bother arguing should; the evidence shows how much we do care, and I respect the time and effort that everyone there, on either side of the issue, put into defending something that matters to us and to our community.

The business meeting should be up on YouTube by now - search on "2015 wsfs business meeting" there for the segments. We successfully got EPH moved to the beginning of the agenda this morning, so you should be able to view the debate and vote in the first couple of segments from today. If you weren't able to be physically there today, you can still follow along vicariously. I think the final serpentine vote was a really dramatic moment that I was proud to be a part of. Keith did a terrific job as our lead presenter.

One thing that was apparent from the debate was that we should think seriously about how to ease the Hugo administrators' workload in sanitizing the input data. I've got some ideas about this, but it would be good to get input from some real experts in the area, as we did for voting systems. It would be nice if we could reduce the administrators' workload for 2016, even before EPH gets ratified.

280
The work would be to be done without EPH in place. (EPH is a counting method, and doesn't affect data cleanup, which is the biggest part of the workload. Determining the category is the next biggest, I suspect, and also would be there without EPH.)

Thank you for all your hard work!

The claim was made that they don't cleanup the names getting few nominations, though that seems like not a good thing to me. I understand it's unlikely to make a difference, but I would hate not knowing for sure.

I have joined Kate Secor and Jack Foy's open software group. One of the things they are going to be looking at is open versions of all the various Worldcon software, including cleanup (and, of course, EPH). If you're interested in getting involved, let me know, and I'll have Kate send you an invite.

BTW, we're home safe. It was great seeing you all in person and finally putting faces to names!

Kilo

The whole argument of "we only cleanup the ones at the top" is weird. The GOT episode could have been written in a lit of ways, splitting the posts and making that one not one of the top. Same applies for each episode of any show. If they do not clean all the data, how do they know that there is no finalist hiding under the line just because it was written in too many ways. (Which is what I wanted to make a comment on today amongst other things if I had gotten the chance to actually say something.

Kilo, I am very interested in that.

Anonymizing the data should be fairly straightforward. Presumably the electronic ballots come in a standard format and can be stripped of any such data. (If someone gives me the format, I (or lots of people) could write a tool).
The hand ballots were a very small portion of the data and may or may not be worth converting (if they haven't been already).

Well, bother. I've just realised that it is possible-but-unlikely for a ballot to have 7 valid nominations in a category due to transfers under Section 3.8.7 :-(

Tim Illingworth @286

Even more than 7 in some weird circumstances (for example in the categories where there is the Fan/Professional separation). It does not change anything though... - either for 4 and 6 or for EPH.

All:

It's the first week of classes here, so I'm still trying to help panicking freshmen figure out how to handle their first astro course. As such, I'm still getting caught up on emails, but I will be soon!

I think we need to consider finding a permanent home for EPH discussions. Our hosts here have been beyond gracious, and I really feel like this -is- EPH's "home". But I think at some point (soon) we're going to need some threaded discussions, given all the new people who are interested in the system and with all the testing and implementation we're about to do.

I can think of a couple of options:

- Kate Secor and Jack Foy have started a Worldcon open software group aimed at developing open source versions of all of the software needed to run a Worldcon. I've joined the group in the larger capacity, but I think EPH is a logical fit for this. There are already several implementations of EPH "in the wild", and as far as I'm concerned, the more, the better. Kate and Jack have a Google group as well as a couple of online collaboration sites set up already.

- We can set up our own EPH discussion board and just have at, much as we have for the past four months, but with maybe a little bit more structure (our [TAGS] system worked well, but can get unwieldy when there is a lot of discussion).

- Or, of course, we can stay here. The only issue I see with that is that a single linear thread is difficult to follow and eventually is going to be hard to find on the site.

A sort of meta-issue to consider is open access. On the one hand, I feel strongly about having all the deliberations and testing of EPH be done in the open. Currently the open source group is invitation only (though Kate gives invites to anyone who is interested). I feel that whatever we do, we need to have a regular place where just anyone can drop by and visit, participate in the discussions, offer perspectives, etc. On the other hand, I recognize why the open software group has some small measure of controls on it -- it's much easier to get work done. EPH, again, is a logical fit for their group, but I'm not sure that should be the main "home". Then again, we've got a lot of work to do as well...

What are your thoughts?

Thanks,
Kilo

I'm still on the train, about 3 hours out from Boston. It was great to actually meet some of you and I wanted to thank you all. I agree with Kilo that there should be some permanent discussion home. How would people feel about electology.org putting up that space?

Keith@288:That seem's like the next logical step.

How does one get in contact with Kate and Jack?

Steve@290:

The easiest thing is probably for me to just pass your email on to them. You could send that to me, if you like, since googling my handle seems to readily yield my work address.

Incidentally, I can also host a discussion board. If that's what we want to do, as I have my own server.

Kilo

I gotted my t-shirts! In the mail! They are a thing of beauty and a joy forever, thank you Jameson.

How do I send you shipping costs? My email is lady at demesne dot com

Except, of course, that my email is lydy at demesne dot com

damn autocorrect

I got my T-shirt too! Thanks Jameson, I've also replied by E-mail to work out payment.

I got my T-Shirt, too! Thanks!

(Sent you an email)

[DATA REQUEST]

I have a bit of an update on data to run through EPH. Apparently most of the admins combined Worldcon with their vacations, so it may be a couple of weeks before we see any of the 2015 data. I'll keep you posted when I know anything definite.

In other news, I've been in contact with Dave McCarty (Hugo 16 Admin). We will begin working with him in November to prepare for next year's Hugos, so we should have quite a bit of work to do starting then.

And finally, I mentioned this on File770, but one of the things I'm interested in for EPH is the effect of nominee dispersion under the system. For example, if there is literally no overlap on the ballots (i.e., every single voter nominates a unique set of works), then obviously EPH will not give results any different from our current system -- the most nominations will win. We've always known this, and we even put it in the FAQ. If fandom doesn't have a general preference, then any slate will get a larger percentage of the final ballots than they would otherwise (you can't unite what isn't united). My question is: What are the numbers for this? For example, what percentage of "dispersion" -- and we need to define what we mean by that -- will yield what number of slate works on the final ballot for different situations? It's something I'm going to be working on, but I thought it would be useful to have others looking at this as well. Does anyone have any other well-defined questions that we can study?

Thanks,
Kilo

Hmmm… dispersion and EPH…

Perhaps try "M items selected at random from a slate of N items", with different distributions for the "at random"? Like, Gaussian distribution, flat distribution, etc etc etc.

I don't know much about statistics, myself, but I do recall there are various ways to measure the degree of correlation between two variables… so try generating a number of different "M out of N" slates, with each such slate generated to match a different measure of correlation?

What is the effect of increasing the number of finalists, while keeping the number of nominations per nominator constant? (The "5 and 6" question.) My intuition is that in most cases the number of slate finalists remains constant, while the number of non-slate finalists goes up by one (or, in the case of "5 and 5+n", up by n). But occasionally the slate will get an extra finalist, and if we have some estimate for how likely that is, it could have an effect on next year's debate.

298
Right now the rules are that you get an extra finalist if there's a tie (two-way, for 5th, 3-way for 4th). There may be other stuff involved, but that's the basics. You should be able to get an estimate by looking at how often there have been 6 nominees on the ballot.

PJ, that was meant as a response to Keith's question re EPH, "Does anyone have any other well-defined questions that we can study?" Right now "4 and 6", which would reduce the nominations each voter can make from 5 to 4 and increase the number of finalists from 5 to 6, is pending ratification at next year's Business Meeting, and could be ratified alone, ratified along with EPH, or ratified with modifications along with EPH. It would be interesting to see how "4 and 6" interacts with EPH; it would be easier to see how "5 and 6" (and "5 and 7", etc.) interact with EPH. And having good answers to that will be useful at the Business Meeting next year.

Yes, I gathered that. I was looking at the question, 'how often are there more than 5?'
(FWIW, in 1972, there was one category with four nominees, and two with six.)

David W., #280: If something can be done in the way of easing data clean-up and offered to the 2016 committee, that neatly cuts the legs out from under one of the few objections to EPH that has any validity at all. It seems to me that this is an area where the group should concentrate its efforts over the next few months, using the 2015 raw nomination data as a starting point.

My thanks to Kilo, Jameson, and everyone else who put in the effort to develop this proposal and hammer out the kinks. We've demonstrated that a few greedy people with an agenda can't succeed against a genuine grass-roots uprising.

302
Cleanup is something that has to be done anyway (it isn't actually part of EPH) - probably less of a problem with online nominations, because no handwriting, but spelling is always going to be a problem.

Cleanup has to be done anyway, but it has to be done a bit more extensively with EPH. (Under the current system, you can often say, "It's not obvious if these two are the same, but it doesn't matter because they won't be in the top 5 anyway.") At the Business Meeting, the Hugo administrator estimated it would take about 50% longer to do data cleanup under EPH.

304
It will be exactly the same cleanup. (Unless that admin was planning to hand-count the nominations!) If it's clean enough for the regular counting program, it should be clean enough for EPH.

As I mentioned in another thread, data cleanup isn't just fixing spellings. Alexandra Erin writing as Theo Pratt wrote John Scalzi Is Not A Very Popular Author And I Myself Am Quite Popular. Scalzi then read it as an audiobook. Alexandra Erin/Theo Pratt is eligible to nominated for Best Fan Writer on the strength of the book; the book itself might qualify as a Related work (written) or Short Dramatic Presentation (audiobook). In all those cases, the Hugo Administrators will have to know that Alexandra Erin is the same as Theo Pratt, and Scalzi didn't write it (although he read it), and so forth, and so on....

306
Comes with the territory.

In re 302, 303, 304, 305, and 306:

"data cleanup isn't just fixing spellings"

My idea, which is mine, is that the [one, true] Hugo Administrator corral helpers for the clean-up, and each helper would get to do all the cleanup on one particular category, right on out to eligibility checking. There's nothing like full-immersion familiarity to make you an expert.

308
How does that deal with the case of things nominated for the wrong category?
(For that matter, how do they deal with that now?)

There's nothing like full-immersion familiarity to make you an expert.

Don't ask me about how the final ballot-counting works. I've been immersed....

emgrasso @309:

How does that deal with the case of things nominated for the wrong category? (For that matter, how do they deal with that now?)

If it's a length-based category, administrators have rules for it (see Section 3.8 of the WSFS Constitution). For a complete category error, it gets harder. That's why a new rule passed this year to deal with things that could reasonably classed in multiple categories; the category in which the work gets more nominations is where it lands. However, this can lead to "vote splitting" where a work ends up failing to make the ballot in either category because some people thought it was fish and some fowl.

Root problem: It's difficult to write category definitions that are ironclad. People are very good at finding things that break technical definitions. Try writing what you think is an ironclad description and see how many people can break it.

Lee @ 302

Actually a lot can be done as long as someone trusts external software enough to allow the cleanup to be written from outside - both with a community effort to build the closeness maps or with fuzzy logic that does matches and requires minimal human check for validity. There aren't that many entries that really need cleanup once you get the possible automation under way.

I suspect that if cleanup were done once a week on incoming nominations, or once a day in the big rush toward the end, that it wouldn't be that much of a problem. (Especially if you're not dealing with everything being on paper.)

And especially if the cleanup is done with building a map that catches the same misspelling when it happens again.

314
The biggest problem is having consistent/standard ways to do things.
I spent way too much time doing data entry and QC thereof. Consistency in procedures was something we had to deal with. The management decided that all changes would be handled by meeting-plus-written-document so as to avoid word-of-mouth inaccuracies. (The pleasure was getting information pulled from the database 20 years after we turned it loose, and having it still be generally good.)

That's where a mapping done properly will help. Pick one spelling as the main one, do mapping for anything else. If something gets mapped wrongly, it is an easy fix.

Not exactly trivial but doable. And if we keep saying that we want more people to nominate, the work load will keep increasing - EPH or not.

Annie, #312/316: Yeah, I kept thinking that we have the technology to automate a lot of this with software that has a "learning curve". Am I right in thinking that a Bayesian approach would be the right sort of starting point, or is my understanding deficient? ISTM that LibraryThing does a similar process with combining different forms of a title into a "work", and while I know that part of that job is crowd-sourced by individuals, I'm quite sure that not all of it is.

And yes, once it's been implemented it will definitely make things easier for the administrators going forward! A classic example of "the curb-cuts effect" -- you put in curb-cuts for people in wheelchairs, and then it turns out that they also help parents with strollers, and people using walkers, and people towing wheeled grocery totes, and...

On data cleanup, I have a vague memory of something called soundex, which was a method of detecting and dealing with words that sounded alike.

I don't know just what gets done with the nominations here and now, but the tools exist, and they're not necessarily going to need to be used more in EPH.

An example that struck me: how might people spell "Toni Weisskopf"? There's a couple of obvious possible errors there, but they'd be picked u by sounething based on the sound of a word.

To toss out the name of another piece of software that might be useful, OpenRefine is used by a lot of librarians (and others) to do data cleanup and normalization.

By analogy with the this-year-passed "A Story By Any Other Name", I would think that the audiobook of John Scalzi Is Not Very Popular... should qualify as a Best Related Work, not Best Dramatic Presentation (Short Form). I.e., if it's primarily prose, then the fact that it's an audiobook rather than printed text shouldn't matter.

I don't think prose v verse is the issue, but rather is is meant to be a performed work.

Ah, fandom, where taxonomy is considered fun, even light entertainment (as well as a contact sport).

I ended up with a couple hundred extra ribbons. I'm going to save them for distribution at MidAmeriCon next year (I might be going, and if not I'll send them to someone who is) - but in the meanwhile, I'm still happy to mail out small batches to anyone who wasn't able to get one.

Kilo/abi: will you be starting a new thread when we have the anonymized nomination data?

Also: while the cleaned-up data will be more useful for analyzing EPH, can I also request that the non-cleaned-up data be made available? I'm hoping to help develop the cleanup workflow, and having a real set of raw data to work from would be really helpful.

Chris,

I somehow managed to miss you in Spokane altogether - had been meaning to look for you every time I was getting to the BM and things got a bit hectic...:( Just sent you a note.

Chris@323:

I'm not opposed to a new thread, though we still have 600 or so messages we could use in this one. I've no idea how long it will be before we get any data, however.

Kilo

Annie @ 324: No worries, I had trouble getting to the BM on time myself some mornings. Con-going with a three-year-old will do that! I saw your form entry, and we'll get it figured out.

Kilo @ 325: I wasn't thinking about the 1000-post limit so much as the fact that this thread has been open since June and having the data will start a new and probably wide-ranging discussion. I'm fine either way, though.

Kilo @325:

The thousand-comment thing is actually just that the blogging software here begins to drag and keen at about a thousand comments per thread. So is a conversation looks to be going past that limit, we open a new thread.

A new topic (like the results of the analysis) totally merits a new thread.

One thing that I think would be helpful is if we could do some formal analysis of EPH that could help the Hugo administrators demonstrate that certain small errors in the data can't affect the outcome. Something like "Well, we didn't catch that 'The Butterfly Effect' by R. L. Smith with 2 nominations was the same work as 'Social Impact of Insignificant Changes' by Larry Smith, with 3 nominations, but because the top ten works all had at least 45 nominations, we know that it couldn't have affected the outcome whether we merged the two works or not."

Jameson may have links to some theorems on the underlying system, but I think we will need to analyze full-blown EPH with the specific tiebreakers that we have now proposed.

To start us off, let me propose Lemma 1: If there are a maximum of n works nominated per ballot in a category, then Work 1 with x nominations can never be eliminated as long as Work 2 with y < x/n nominations has not yet been eliminated.

Proof: Work 1 can never have fewer than x/n points, and Work 2 can never have more than y points. So if Work 1 is selected during the selection phase of a round in which Work 2 remains, Work 2 must also be selected in that same phase, and Work 2 will be eliminated for having fewer nominations. If Work 1 is not selected, it can not be eliminated during that round. Therefore, Work 1 will survive longer than Work 2.

David@328:

Your analysis looks good to me. Very well done. Does it tell us anything about the 2015 nominations? I'm thinking it may be time to try to do the sort of computer-generated ballots we did for 2013...

Kilo

I see that Glazer and Lorentz are refusing to release the data because they say it can't be sufficiently anonymized. Given that they didn't want to do it in the first place, this raises the question of what their standard is, and why they can't come up with one that's workably anonymous.

They're using the AOL and Netflix studies as reasons, but without more information on what those two studies did in the way of anonymizing, and how hard the researchers tried to identify the people behind them, that's not a really solid basis for their response.

I also got the message from Glenn. I did not get the follow-up from John, but recall that we are already set up to work with Dave McCarty starting in October or so. I presume that's the committee he's referring to. If so, we should still eventually have data for testing, though it may have to be me running whatever tests the group needs done. We'll cross that bridge when we come to it though.

For now, I think the type of analysis Dave has proposed is a good way to go. For example, what are the repercussions of 4/6 in light of Dave's analysis and the analysis I made in the other thread? I'm thinking that 4/6 may actually empower slates in the short fiction categories since any given work now only has to make it to fifth place to be safe instead of fourth. I haven't thought enough about it to be sure, though.

Kilo

EPH generates a list of all the nominees in order of elimination, so it's easy to tell the effect increasing the number of finalists has: you just go a bit further down the list.

It seems intuitively obvious that eliminating a slate work results in the remaining slate works getting a big jump in their point total, while eliminating a non-slate work causes the remaining non-slate works to get a small jump in points. The implication is that, after one slate work is eliminated, several non-slate works will be eliminated before another slate elimination.

So, except at the very top of the list, we expect a wide gap between slate eliminations, which means if we choose a random point on the list, the odds are many-to-one against it being a slate choice. In general, I'm pretty sure the implication is that increasing the number of finalists from five to six will most often increase the number of non-slate finalists by one.

(I'm less certain what effect reducing the nominations to four would have. I think it might introduce coordination problems for the slates—they won't be able to divvy up their votes in the most efficient manner possible—but that will be balanced by lowering the number of nominations the top non-slate works get by anything from a tiny amount (assuming that when voters cut one, the ones that would have ended up with the most nominations on a five-vote nominating ballot are the ones least likely to be cut when going to a four-vote ballot), up to 20% (assuming no correlation between how many nominations a work would have had, and how likely it is to be cut)..)

I waded into the swamp which are Beale's comments (I have yet to find those Grant praises as being "worth reading") and found this gem:

What I suspect it will show is that the SP/RP voters had more variety, while the TORlings voted the same.

I keep trying to parse it: almost as meditation while cleaning the stove, etc., and it's gibberish.

The smaller number of Puppies voted with more variation, yet took over the ballot; and their greater variety happened to almost perfectly map the Beale Ballot. If so they must have had overwhelming numbers, but they lost every section of the ballot to No Award.

So... how many of them are there? Why didn't this greater pool of voters (if they manage to have more variety, and still overcome the lock-steppers from TOR), turn out to vote for the awards?, etc.

333
Terry, they also seem to have trouble in their minds, telling the nominations from the final ballots. (And they don't seem to understand IRV, from what they write, so even when they're writing about the final ballots, they're still wrong.)

Terry Karney @ 333

Let me try to help here. You know the logic that you learned in school and/or college? And you had heard of common sense, correct. Now - forget all of that. Read that again. Accept it as a lemma. Now look again. See - it makes sense now :)

Terry Karney #333: Not gibberish, but bullshit. As he has long promised, he is claiming victory regardless of what actually happened. He may even expect his followers to buy his claim.

Dave: If it were Teddy saying it, I'd know it was active mendacity (his lips were moving), but this is one of his acolytes (and the "we are all individuals, tell us what to do is amusing, in a bitter, cynical, acceptance of how willing we are to lie to ourselves).

It's like reading manosphere nonsense (no surprise, given the known overlap in the Venn diagram). I guess I still figure people run their written thoughts past some sort of internal process of proofreading as they set them down.

The Six Impossible Things Before Breakfast may be a wonderful intellectual exercise, but it doesn't make them anything close to plausible.

P J: Yeah, the equivocation of "ballot" in that comment stream is painful. As is the idea that anyone could call Glenn and "drop a word in his ear" that made him do anything he didn't want to do, just because he'd been told.

I have to confess the scene that person painted made me laugh. It was textbook melodrama of "the Evil Overlord" from a thirties serial.

It's amusing, reading the comments over there. They think we are using the same "rules" they are. That the issue is "winning" at getting "our" books on the ballot.

I.e. they think putting a book of the "SJW Approved" type will make us "lose", in that we either vote for it (showing the latent hypocrisy of the "sjw"), will No Award it (showing we don't care about "quality", whatever that is) or will somehow manage to make it possible for a Teddy Beale Selection to slip through because we "split our vote" between No Award, and "the SJW Tome".

As P J says, they don't understand anything about the people they are "fighting". Sun Tzu would laugh all the way to the Triumphal Parade.

That said, there is obviously chicanery being planned; they are boasting of it.

339
They will, of course, try another slate. Apparently they figure if it has more than five items, we won't be able to tell, especially if they order it by 'popularity' (quotes because I don't know how they measure it). They may also figure they can send people to the business meeting to vote down EPH ... but I doubt that most of them realize how much that's going to cost. Some of them have figured out that it will 'help' them because it will limit the (non-existent) slates they think we have.

They aren't in our cognitive universe. They seem to figure saying something makes it true, and that no one remembers what they said last hour, or last week, or can find it where they've said it before.

Looking at the notes from Glazer and Lorentz on File770, it looks like their main concern is with nomination ballots that include unique nominees for one or a handful of people.
At a guess (and simplified), there are probably nominations like:

Best Fan Writer:
Crusty Fan1
Best Related Work:
Thoughts From a Crusty Fan1
...

where the nominator is also the aforementioned Crusty Fan1. Thus, simply removing the ID of Crusty Fan1 from the nominating data doesn't really anonymize it very much.

For EPH purposes, it would seem that simply removing ballots with unique nominations or even nominations that amount to less than the 5% rule wouldn't effect the EPH testing.
An alternate method would be to replace the relatively unique strings so, for example, the above nominating ballot would become something like:

Best Fan Writer:
AnonString212
Best Related Work:
AnonString143

Anonymizing the nominees wouldn't affect EPH in the slightest, though one of the big things people would want to know is who would have been nominated under EPH. Of course, identifying the top 15 works would be trivial since their titles and number of nominations are published, so I'm not sure anonymizing the nominees does as much as we might think.

Kilo

Keith@342:Right--I think (am guessing) the concern of the data holders is only down at the low end of the nominee pool. If a work is in the top 15, then it is easily identifiable, but who nominated it (other than probable puppy association) is already effectively anonymized by the number of ballots.

I would be willing to be that a lot fewer people are interested in the under-5-nominations section that Glazer and Lorentz may think. Replacing those with random or other strings should cover it.

The other problem is people who made some of their nominations public, and figure that this makes all of them identifiable. I wouldn't put it past the canids to try doing that kind of back-tracing - they want every advantage they can get - but I expect that there's a lot less interest in that than might be expected.
As a fan, I don't really care who nominated: it's what that matters.

PJ Evans, It's difficult to know how puppies rate popularity (I almost wrote "pupularity" there), given that I've heard reports (disclaimer: I have not tried to verify this empirically) that nominations made to the puppy site by known SJWs have never made it out of moderation.

345
I've seen those also.
They seem to have three conflicting measures: book sales, Amazon ratings, and Goodreads ratings.
I'm wondering how much any of them actually read, or if they just look at the first few pages and the last few, and call it a 'read' book.

Steve@341: You mean they were thinking of giving out the data as one file with the nominees not replaced by numbers? Yes, that could be broken.

What I was expecting was a file per category consisting of number pairs:
0001,001
0001,002
and so on, for ballot number and nominee number. Moreover, I wouldn't expect ballot 1 in Fan Writer to be the same ballot as 1 in Fanzine, and so on. I would be very surprised if anything were identifiable out of that, and EPH could run on it just fine.

The top 15 nominees in each category would (probably) be identifiable, but with no links across categories it should not be de-anonymisable.

I cannot say what pattern they've identified, but trying to think of something based on my past experience (not in the current year), I thought of situations where if even if you didn't track across categories (and there was no intention of doing so), you could make a very good guess about ballots in small runs of identical nominations. The more times you see the same group of nominees grouped together, the harder it would be to make an educated guess about who cast them. Furthermore, such "mini-slates" are unlikely to have any effect on the overall election, because the nominated works probably don't have a chance to make the ballot.

Assuming that what's happening is what I postulate here (and I don't know whether or not it is), you could, I think, get rid of the effect by replacing any nominee that got less than N nomination votes with an anonymous tag. (Work 0001, Work 0002, etc.)

Regrettably, while I'm pretty sure this meets the requirement to protect voters' privacy, I doubt this will satisfy those people who are convinced that there is a Massive Conspiracy of Tor (or whichever demonized group name they want to use right now). But OTOH, I don't think there is anything that would satisfy them. I'm sure that even if you gave them the complete raw data including the name of the individuals and the way they voted on all categories so you could do cross-category correlation (and that's never going to happen!), if the data didn't confirm their paranoid fantasies, they'd start screaming that it was all faked.

Tim@347:They don't give details on what they were thinking, but seem to imply that a first approach of just removing the ID's was their thought (as well as that of others). But, luckily, they then fairly quickly realized that there was quite a bit of information in the ballots and a good scrubbing of the data would take some time and thought.

Your idea would work nicely for EPH purposes, I believe and seems like it would remove any identifiable traits.

You know, I have a genealogy program that can generate reference strings that would work; they aren't based on names, but on relationships to a given person in the database where they're being generated. The strings are a mix of digits, letters, and symbols that look pretty random. (Ex: 103.F7*1/2.012*24)

Kilo@342

As nice as it will be to know who would have been nominated, it is actually more important to see how many of the current nominees are staying nominated - the idea is to see if slates are loosing power, not to mourn the authors that got pushed this year. And you are absolutely right - the top 15 will be a child's game from entirely replaced names. However - the problem with this approach will be normalization - I am not sure how much of a difference the non-normalized ballots will make...

Just saying :)

Cassy B.@345: Both I and another clearly recognizeable file770 user have had recommendations get published at the SP4 site; the one comment of mine that didn't had two links in it. I suspect multiple links go to moderation, but so far, no censorship.

Greg M, very good to hear; I withdraw my allegation. I must have conflated the nominating site with another site; I apologize for my error.

Annie@351: It might be good to know how well 'would be nominated under EPH' correlates with the (more or less already known) 'would have been nominated in the absence of slates'.

Andrew M @ 354

Except that we won't know. We can guess eventually (the next 5 in the list) - but just removing the slates entries won't show us where these nominations would have gone otherwise - some of these guys would have nominated anyway.

Keith "Kilo" Watt:Is there some way we could get the format of the files in which the ballots are recorded? If we had that, we could produce an anonymizer (or a couple of them) that the Hugo Admins could try out.

Steve@356

At this point, it doesn't look like it. That was actually proposed earlier, but the admins voted not to release that. I don't know if they would change their minds now that the Hugos are over or not. I'll send a message to Dave and see what he says, though.

Kilo

Keith@357:Thanks. I can see where it would have been some work they didn't need prior to the Hugos.
Hopefully, now will be a better time.

GregM @ 352, I just ran across this message over at File770.

It appears that at least some non-Puppy nominations are not making it out of moderation on the Puppy list, according to this report.

Steve@358: I sent Dave a message, so we'll see what happens.

Kilo

Some thoughts about various issues involved with anonymizing nominating ballots:

First, I appreciate that the Hugo administrators are taking the time to think through privacy implications of what they may release, and ways in which information can inadvertently leak.

Second, it's probably worth thinking about what kinds of information leakage would be particularly damaging, and taking special care to prevent them. The most obvious one that occurs to me is if editorial nominations could be identified from fiction choices (due to consistent random IDs across all categories): an author who had published some of his or her fiction nominations on a blog might be especially concerned about editors taking offense about who he or she had or had not nominated in the editor categories. Other risks may be apparent to people who have been more heavily involved in the industry.

While EPH can certainly be tested with data that has userIDs in every category anonymized separately, and that has individual nominees with few votes anonymized, the accuracy of the results depends in part on how raw the released nominating data is. If a work should be consolidated with another form of the name in the same category, having the work's name anonymized will prevent that. Likewise, if a work should be moved to a different category and consolidated there, having inconsistent userIDs between the categories will prevent this, since correct treatment partly depends on how many nominations the same user made in the new category.

So what I suggest is that if anonymized raw data is to be released, that consistent userIDs be used between categories where the likelihood of misclassification is high and the danger of inadvertent information leakage between the categories is deemed low, while using differently randomized userIDs between categories where either of these is false. My first suggestion for groups where consistent UserIDs should be used is: {Novel, Novella, Novelette, Short Story}, {SemiProzine, Fanzine}, {Pro Artist, Fan Artist}, but I'm sure the administrators have more experience to judge where these boundaries should be. My other suggestion is that work names not be anonymized within these groups, again to facilitate moving works between categories if necessary.

If the data being released has already been cleaned up by the administrators, then neither of these measures is necessary. I think there is some value in letting us see the raw data and crowdsource the cleanup, to better appreciate what the administrators have to do and come up with solutions to help them in the future. But I don't know what the state of data that could be released is, and I certainly don't want to require the administrators to put in a lot of extra work to get it in shape to release. I trust that there is some balance between safety and effort required that could get a reasonable anonymized data set release, and I hope it can be found.

Me @359, just following up to say that the non-Puppy nomination referenced earlier is now out of nomination.

Out of moderation, Cassy B. @362? Took me a minute to realize that was probably what you meant.

Tom Whitmore @363, yes, I mean that the non-Puppy recommendation on the puppy site had made it out of moderation, a few days after submission. Sorry if I was unclear. In the interests of fairness to the pups, I figured I should mention it. According to the submitter, it appeared as submitted.