Analyzing EPH

By Bruce Schneier: Jameson Quinn and I analyzed the E Pluribus Hugo (EPH) voting system, proposed as a replacement for the current Approval Voting system for the Hugo nominations ballot. (This is an academic paper; the Hugo administrators will be publishing their own analysis, more targeted to the WSFS Business Meeting, in the coming weeks.) We analyzed EPH with both actual and simulated voting data, and this is what we found.

If EPH had been used last year in the 2015 Hugo nominations process, then…

The number of slate nominees would have been reduced by 1 in 6 categories, and by 2  in 2 categories, leaving no category without at least one non-slate nominee.

That doesn’t seem like very much. A reasonable question to ask is why doesn’t it reduce the number more. The answer is simply that the slate was powerful last year.

The data demonstrates the power of the Puppies. The category Best Novelette provides a good example. This category had 1044 voters, distributed over 149 different works with 3 or more votes. Of these voters, around 300 (29%) voted for more Puppy-slate works than non-Puppy ones, and about half of those (14%) voted for only Puppy-slate works. These numbers are also roughly typical. The other 71% of the ballots included under 3% with votes for any Puppy work (this is relatively low, but not anomalously so, compared to other categories).

Despite being a majority, the non-Puppy voters spread their votes more thinly; only 24% of them voted for any of the top 5 non-Puppy works. This meant that 4 of the 5 nominees would have been from the Puppy slate under SDV-LPE or SDV.

(SDV-LPE stands for “Single Divisible Vote – Least Popular Elimination,” the academic name for this voting system. SDV is “Single Divisible Vote,” a long-standing and well-understood voting system.)

To further explore this, we took the actual 2014 Hugo nominations data from Loncon 3 and created a fake slate, then analyzed how it affected the outcome at different percentages of the vote totals:

In Figure 1, we assume perfectly correlated bloc voters. They vote in lockstep (with minimal exceptions to prevent ties), and their five nominations are completely disjoint from the other nominations. As you can see, both SDV-LPE and SDV reduce the power of the bloc voters considerably. Under AV, the voting bloc reliably nominates 3 candidates when they make up 10.5% of the voters, 4 candidates when they make up 12.5%, and 5 when they make up 19%. Under SDV-LPE, they need to be 26% of voters to reliably nominate 3 candidates, 36.5% to reliably nominate 4, and 54% to reliably nominate 5….

Figure 2 simulates a more realistic voting bloc. We sample the actual behavior of the bloc voters in the 2015 Hugo nominations election, and add them to the actual 2014 nominations data. For the purposes of this simulation, we define bloc voters as people who voted for more Puppy candidates than non-Puppy candidates. In this case, the actual bloc voters did not vote in lockstep: some voted for a few members of the slate, and some combined slate nominations with non-slate nominations. For the purposes of the simulation, when they voted for the nth most popular non-Puppy candidate in 2015, we imputed that into a vote for the nth most popular non-Puppy candidate in 2014. In this case, SDV-LPE and SDV reduce the power of those voting blocs even further. Under AV, the voting bloc reliably nominates 3 candidates with 14% of the voters, 4 candidates with 17% of the voters, and 5 with 39%. Under SDV-LPE, they need to make up 27.5% to nominate 3 candidates, 38% to nominate 4, and 69.5% to nominate 5….

The upshot of all this is that EPH cannot save the Hugos from slate voting. It reduces the power of slates by about one candidate. To reduce the power of slates further, it needs to be augmented with increased voting by non-slate voters.

There is one further change in the voting system that we could make, and we discuss it in the paper. This is a modification of EPH, but would — for the slate percentages we’ve been seeing — reduce their power by about one additional candidate. So if a slate would get 5 candidates under the current system and 4 under SDV-LPE (aka EPH), it would get 3 under what we’ve called SDV-LPE-SL. Yes, we know it’s another change that would require another vote and another year to ratify. Yes, we know we should have proposed this last year. But we had to work with the actual data before optimizing that particular parameter.

Basically, we use a system of weighing divisible votes named after the French mathematician André Sainte-Laguë, who introduced it in France in 1910. In EPH, your single vote is divided among the surviving nominees. So if you have two nominees who have not yet been eliminated, each gets half of your vote. If three of your nominees have not yet been eliminated, each gets 1/3 of your vote. And so on. The Sainte-Laguë system has larger divisors. If you have two nominees who have not yet been eliminated, each gets 1/3 of your vote. If three of your nominees have not yet been eliminated, each gets 1/5 of your vote. Each of four get 1/7; each of five get 1/9. This may sound arbitrary, but there’s well over a hundred years of voting theory supporting these weights and the results are still proportional.

Implementing SDV-LPE-SL using the actual 2015 Hugo data:

SDV-LPE-SL comes even closer to giving slate voters a proportional share, with 7 fewer slate nominees overall, and only 1 category without a choice between at least 2 non-slate nominees.

For the perfectly correlated voting bloc simulation:

Under SDV-LPE, they need to be 26% of voters to reliably nominate 3 candidates, 36.5% to reliably nominate 4, and 54% to reliably nominate 5. Under SDV-LPE-SL, they need to be 35% for 3, 49% for 4, and 66% for 5.

And for the more realistic voting bloc simulation:

Under SDV-LPE-SL, they need 36% for 3, 49% for 4, and over 70% for 5.

That’s a big difference.

Here’s our paper. It’s academic, so it refers to the voting system by its academic name. It spends a lot of time discussing the motivation behind the new voting system, and puts it in context with other voting systems. Then it describes and analyzes both SDV-LPE and SDV-LPE-SL.

356 thoughts on “Analyzing EPH

  1. @Cassy B

    Indeed, for all the concerns about raising Admin workload when 3SV or DN or any of the other voting systems are discussed, it seems that giving the Admins powers to either vet ballots and discard them would be really jacking up the Admin’s work load. Four thousand ballots is a lot to go through – as is defending there decision to every winterfox or Teddy Beale or future version of either who covers their griefing with a heavy round of concern trolling about the power of the Admins.

    A few years of that, and an intermediate round would seem like a cake-walk. As well as having more democratic legitimacy.

  2. “That said there is still the dog feces problem. I think DN or 3SV do not help there though. Even if Beale and Ted’s Slating Adventure gets down voted or doesn’t make the up vote it’s still out there in the record as a Long List Finalist ™”

    They wouldn’t be finalists as named by the Mark Committee. They wouldn’t be listed on Wikipedia and the Hugo website as finalists. And they wouldn’t be there for the Hugo Award evening to sully it with their presence.

    “Maybe we compromise on a rules change several other folks have already proposed? Strip any work voted below No Award retroactively of it’s finalist status. Find some way the Mark Committee can legally enforce that.”

    I think I had a proposal like that myself a few months ago before the we saw this years nominees. It was shot down then. 😛

    I don’t know. My first reaction is that it isn’t enough, but as everything else, I’ll have to think about it a bit more.

  3. I’ve posted what I think the status of the various proposals is. But instead of breaking it down by solutions, let me try to break things down by problems.

    Problem 1: Non-slate voters do not get enough choices among the finalists.

    In the presence of slates at their current strength, the status quo gives about 0 non-slate choices per (non-best-novel) category (nspc), EPH would give about 1 nspc, EPH+ would give about 2 NSPC, and X/6 would add about 1 NSPC above those numbers. So EPH+,5/6 would be about 3NSPC.

    A+2 would add 2 NSPC when needed. My proposal above using “non-slate satisfaction” would probably make sure that there would be at least 4 NSPC (though I can’t test it with real data under the NDA, so that’s just a guess). Let’s call that proposal S+N, for “satisfaction-based add n”.

    Problem 1a: Many non-slate ballots end up ineffective because they only support “long tail” works.

    A longlist would help with this. In so doing, it would probably add about 0.5 NSPC to any of the combinations without S+N. (Some people seem to think that any votes that were added after a longlist came out would be less legitimate. I disagree. It is entirely possible that a voter would read/view a work because they saw it on the longlist, and judge it worthy. It’s also possible, though probably rarer, that they’d be reminded of a work that they found worthy but had forgotten. Finally, if a longlist had several works by the same author, and told which was more popular between these options, it would allow that author’s fans to consolidate their votes, which I think is legit.)

    Problem 2: Admin burden.

    EPH makes this a tiny amount worse, because you have to be a little bit more careful about canonicalization (the amount of stuff you can more-or-less safely ignore goes down by about 1/4). EPH+ is no different from EPH.

    A longlist makes things worse in some ways (more stages of voting) but better in others (easier to contact authors).

    Problem 3: finalists offensive to community standards.

    A longlist helps a small amount with this, but slate voters could still get at least 1 finalist per category. If you want to really solve this problem, you need either 3SV/omnibus, or No Award. (Perhaps you could strengthen No Award, by passing a resolution that says that any work that finishes below No Award is retroactively redefined from being a finalist to being a “failist” or something).

    Problem 4: New attacks; “gatekeeper” slate

    (see my “anti-Virgo” explanation above)

    This would be a serious problem, but only if there were DN or omnibus without EPH, +N, or (preferably) EPH+.

    Problem 5: Complexity and time at the BM

    For this reason, I think there should be as few proposals as possible for the greatest effect possible. Some examples of combinations I’d consider good for the long term would be:

    EPH+, 5/6, and omnibus
    EPH and S+N
    S+N, DN, and “retroactive failist”

    I think ideally we should settle on one of these or something similar, then offer up the minimum number of proposals necessary to get there.

  4. @stoic cynic:

    The problem with allowing removal of works that might discredit the Hugos is effects on non-slate works.

    It’s the 1950’s. A book with a non-white protagonist makes the shortlist. Is it discrediting?

    It’s the 1960’s. A book with a clearly homosexual protagonist makes the shortlist. Is it discrediting?

    et cetera, ad nauseum.

    I know we have clear trolling today but granting a power doesn’t ensure it will only be used for today’s problem.

    There is no getting around it: under 3SV, the list of finalists will be no better than 51% of the membership. The situation right now is that, for most categories, the final ballot is no better than 10% of the membership. Under EPH/+, 2/3 – 5/6 of the ballot will be no better than 10% of the membership.

    I would rather take my chances with the 51%. Today, the Griefer campaign is keeping people of color off the ballot in many categories. It’s keeping queer people off the ballot in many categories. It is putting works on the ballot that slander men and fantasize violence against feminists.

    This is not a hypothetical.

    Nor do I think people should comfort themselves that EPH/+ will restore social justice. Let’s say EPH/+ opens up two slots in Best Fan Writer. That means if a trans author, or a POC author, or an outspoken feminist, say, comes in the top 2, they can be nominated despite the Griefers. (Take it as read the current Griefers will keep these people out of their slates.) In a system that eliminates Griefer-slate noms, a trans/POC/feminist author can get on the ballot of they come in the top 5.

    Given the realities of oppression, the marginalized author has a much better chance of coming in the top 5 rather than the top 2. So the EPH/+ world is still problematic from a social-justice perspective.

  5. @Cassy B:

    I have no objection in principle to a 15-work longlist from which Hugo voters can then affirm their five (or ten) favorites for eventual voting. I’m a little leary of downvoting. Upvoting makes me happy. (And if you upvote everything BUT the bad actors (if any), you are, in essence, downvoting the bad actors.)

    Cassy, I hear you on the different emotional valences of upvoting vs. downvoting. But the basic recognition must be that the Hugos are under attack by malign actors. The questions become:

    1) Are people willing to explicitly act (via downvoting etc.) to repel the griefers?
    2) Do they prefer to accept the continuing salience of the griefers as the price of not acting explicitly against the harm being done?

    Everything else is either hoping the problem just kind of goes away (it won’t), or wishing for a way to make it go away without having to acknowledge that that’s what one is doing.

    WSFS threw an open-house party. Drop $40 bucks for snacks and you can come in and do karaoke and play all the party games. Some folks show up who literally hate the guts of everyone else at the party. They despise the hosts. They have contempt for the guests. They act out to make the party as miserable for everyone else as possible. At this point, one thing everyone else can do is change the ground rules of the party to toss the jerks in open recognition that they are getting rid of jerks. Another thing everyone else can do is accept being miserable for the sake of the principle that it’s an open-house party. The other other thing they can do is decide they really want to get rid of the jerks, but they don’t want to make a scene, so they try to come up with a new set of ground rules that will have the effect of expelling the jerks without coming out and saying that’s what they’re doing.

    I would argue only the first response is healthy, but I’d entertain arguments that the second one is too. But the third approach is pure social dysfunction.

  6. Jim’s Last Querulous Comment for the Moment is about 3SV vs. DN. Specifically it’s a question about DN:

    How is this supposed to work?

    Scenario: I read a dozen short stories last year and nominated 3. The longlist comes out with 15 entries on it. Two of them are stories I nominated last year. Three of them are stories I liked okay but didn’t nominate. Five of them are from the Cucks ‘R’ LOL!!!1eleven!! slate and about like you imagine. Five of them I never read.

    What do I double-nominate here? And what does @whitesriteamirite1488 double-nominate?

    I do not have time to read those five non-slate stories I haven’t read, plus the 7 fanzines I haven’t seen, plus the 6 novelettes etc. so I can’t “vote for what I love.” I know darn well I don’t want to read “The Tire Iron That Falls On You From Nowhere” and its four companions. I only have so much time – and interest, frankly – to read last year’s stuff. So how do I “nominate what I love” in this second round in the real world.

    Contrariwise, I know Tropes Against Men and Ancillary Gamma were on a griefer slate, as was The Traitor Alexandra Erin because this is the internet, and those things aren’t hard to know. It would be straightforward for me to – yes – downvote those based on the information I have readily at hand. And those five non-slate stories I’m not qualified to have an opinion on don’t suffer for my lack of familiarity.

  7. Jim H:

    I don’t care about social justice perspectives. I care about keeping griefers out because they are griefers. It doesn’t matter who they keep out, the point is that their slating do keep people out.

    Jameson Quinn:

    “For this reason, I think there should be as few proposals as possible for the greatest effect possible.”

    And also:

    * Be as simple and easy to explain as possible.
    * Keep as close to current tradition as possible.

    Which means I’m not a believer in introducing new algorithms like S + N. Which means that I’m happy with a longlist of 15 items as that is what is presented today (albeit later in the process). Which means that I don’t want to increase number of nominations in first round if avoidable.

    I think retroactive faillist could be a proposal regardless of system proposed. But I would be happy to have more input from administrators first.

  8. (@Jameson Quinn: Thank you for explaining that EPH+ is not doable this year. That makes perfect sense.)

    ———————-

    I would be in favor of EPH+ being presented as a new proposal. I like that it does even more than EPH, and that it lessens the chances of *any* minority (with a slate or unintentional bloc-voting) from taking over.

    I’m in favor of some form of long list. I suggest using “long list” to refer in general to all variations of 3-stages. 3SV is associated with Kevin Standlee’s original proposal. Omnibus might be mistaken at first glance as a new category proposal by someone who hasn’t been in on this discussion. 🙂

    Right now I’m leaning toward DN with unlimited upvoting. I think that even a year without griefers could benefit from choosing a short list from a long list. But I’m definitely interested in further discussion of any variation. I really like the transparency a long list would add.

    I do not like the idea of admins or an appointed jury adding or removing items. To me, this would be too close to making the “cabal” a reality. I won’t say I wouldn’t consider it at all, but it would take a lot to convince me.

    (Now I’ll go back and read everything I missed while composing this!)

  9. @Hampus:

    I don’t care about social justice perspectives. I care about keeping griefers out because they are griefers. It doesn’t matter who they keep out, the point is that their slating do keep people out.

    Perfectly fair. I was discussing the social-justice angle because it was specifically germane to the concerns stoic cynic raised in the comment I was responding to. I agree otherwise: the problem of the griefers is that they are griefers. Per the party analogy, they literally hate everyone else at the party and want to make the experience miserable for them. The social web gives them the tools to do that.

  10. Camestros Felapton on May 17, 2016 at 2:34 pm said:

    Hampus Eckerman on May 17, 2016 at 2:24 pm said:

    I agree – hence the power rests with a jury rather than the admins.

    I can’t see that as anything other than semantics. In practice, the jury would be administrators in all but name as they would need access to all raw data to be able to remove bad faith block voters.

    It is a tad semantics but not wholly. The difference is the jury have a role different to the admins and would be more the kind of people who can put up with stupid crap.

    If it makes you happier to have group A called “Administrators” and group B called “Jury,” but with both groups having identical authority and access to the data, then okay. Remember, the authority granted by the WSFS Constitution is to The Worldcon Committee. The Worldcon generally delegates it to a Hugo Administration Subcommittee, but the original grant is to the Worldcon Committee. I understand that some people would be happier with having people called Administrators and others called a Jury, but they would all be identically in the same group of people, because they’d all have to have access to the data and they’d all have to be ineligible for a Hugo Award.

    Hampus Eckerman on May 17, 2016 at 2:56 pm said:

    I’m not sure if this would need a rule change by itself. If the administration is given more power (i.e to throw away ballots of griefers), they can by themselves delegate the decision making regarding this to a select group as part of the administration.

    Exactly! We don’t write a set of specific roles for the members of the Hugo Award Administration Subcommittee right now. That is, we don’t say, “You have to have an IT person, and a ballot-counting person, and an eligibility-checking person, and a contact-the-finalists person, and a public relations person, and….” We say, “Worldcon Committee, you have a job to do; go out and do it” and the Worldcon Committee empanels a Hugo Award Administration Subcommittee that includes all of the necessary roles, without having to be told in detail what those roles are. Therefore, if we granted the Committee the right to add additional finalists or throw away specific ballots (shudder), any sensible committee would simply add some more subcommittee members whose task it would be to evaluate whether they should exercise that power, and where.

    Hampus Eckerman on May 17, 2016 at 2:25 pm said:

    Damien Walters is starting to become that person that more or less everyone agrees on being the village idiot.

    He is one of the people who thinks that Daddy Should Beat Up the Bad Man. He doesn’t care about rules or precedent. He just wants someone to Beat Up the Bad Man. Rules are for suckers. Just do whatever you want “because of the emergency.” It infuriates me, because it shows a complete disdain for the rule of law.

    Stoic Cynic on May 17, 2016 at 3:33 pm said:

    That said there is still the dog feces problem. I think DN or 3SV do not help there though. Even if Beale and Ted’s Slating Adventure gets down voted or doesn’t make the up vote it’s still out there in the record as a Long List Finalist ™.

    1. So what? We’ve been publishing the longlist after every Hugo ceremony for years now. Who cares? I personally placed (as I recall) eleventh in 1995 for Best Fan Writer. I don’t think I’ll start billing myself as a “Hugo Award longlist semi-finalist,” because nobody would take the claim seriously, nor should they.
    2. The term you used there is not recognized, and the people on the longlist are not Hugo Award Finalists. They are semi-finalists. Only the works/people on the shortlist are Finalists. Attempting to claim otherwise would be a violation of WSFS’s service marks.

    Strip any work voted below No Award retroactively of it’s finalist status. Find some way the Mark Committee can legally enforce that.

    I really don’t want to have to try and do that, and currently I’m the person who leads the committee responsible for doing so. People could call themselves Hugo Award Semi-Finalists. How much publicity traction that is worth is debatable. The one-vote wonders have done their best to destroy the value of Hugo Award Nominee, just like Nobel Prize Nominee is a laughable term. Let ’em yap, I say.

  11. And I can’t say anything about Omnibus as I clearly don’t understand the system and only get confused when it is brought up.

    If it is going to be part of possible proposals, it has to be explained first.

  12. @Jim: I’d want to divide the second nomination set between “No Award” and “I’m willing to consider it”.

    The limited time makes me unhappy with the idea of trying to nominate a new set out of the longlist.

  13. ” I’d want to divide the second nomination set between “No Award” and “I’m willing to consider it”.”

    Actually a nice way for a second nomination round. Be able to choose:

    * Nominate
    * Willing to consider
    * <nothing>

    And if there aren’t enough people willing to consider it, it is out, even if some people nominated. In practice it is No Award, but in feeling, it is different.

  14. @Jim Henley: WSFS threw an open-house party. Drop $40 bucks for snacks and you can come in and do karaoke and play all the party games. Some folks show up who literally hate the guts of everyone else at the party. They despise the hosts. They have contempt for the guests. They act out to make the party as miserable for everyone else as possible.

    WSFS’s amendment process is slow, but it accounts for malicious internal behavior. In that spirit, voting to remove someone from future consideration for an award could be done in two steps. That would make it harder to abuse the vote, while leaving expulsion open as a possibility. Yes, anyone tagged for that sort of attention would make a lot of noise for two years. But, if they do that anyway–

    In the long run, it may be easier to just grind things out with a series of nominations and no awards. It’s not as cathartic, but a big NO WAY from the electorate has its merits.

  15. This sort of data analysis really isn’t in my wheelhouse, and I feel like I’m missing something obvious. We’ve been told all along that EPH works best with large voter turnout. The authors here re-emphasize that. Last year did not have large voter turnout, especially in comparison to a 150-300 person slate. Most categories had fewer than a thousand ballots cast, many of them quite a bit fewer. EPH didn’t work well there, which I’d imagine should have been expected, so I’m not clear on the reason behind the general wailing and gnashing of teeth I keep seeing about this. EPH would have kept the slates from sweeping any category, which actually seems pretty great to me, considering the relatively low nominating turnout.

    I understand that the point is to make the process more representative of actual voting percentages, and EPH needs work to achieve that. But when it comes to the current problem, I’m not sure why people seem to be treating EPH’s effects on last year’s nominations as some devastating reveal that means future ballots will still be slate-dominated? This year had more than twice as many nominating votes cast in each individual category as last year. Yes, votes will still scatter, but just how would EPH have worked with these higher numbers? Are the percentages stated for last year (those necessary for a slate to triumph) expected to hold true for larger turnouts? Are people assuming that last year’s slate vote percentage necessarily held true for this year, and will for years to come?

    I guess my main question is this: isn’t it a bit premature to declare EPH something that will, every year, only be capable of knocking one or two slated works off a ballot? I’m not saying that it’s the best, or that the other systems proposed wouldn’t work better. I’m just wondering how certain it is that the results of last year are applicable to much larger turnouts like this year.

  16. Hampus Eckerman on May 17, 2016 at 5:25 pm said:

    Actually a nice way for a second nomination round. Be able to choose:

    * Nominate
    * Willing to consider
    * nothing

    And if there aren’t enough people willing to consider it, it is out, even if some people nominated. In practice it is No Award, but in feeling, it is different.

    “Willing to consider” would include “nominate.” I would also like to re-order the long list based on upvotes in this round.

  17. Emma;

    ” I’m just wondering how certain it is that the results of last year are applicable to much larger turnouts like this year.”

    We need to have a system that will work any year, not only those with record turnout.

  18. @Emma

    Good point. Not only did we get double the nominators this year, but I bet we also got many more nominations per ballot. I know I made the extra effort to fill as many slots as I could. Discussion about potential nominees was increased and started earlier.

    When might it be possible to see EPH run on this year’s nomination data?

  19. Hampus Eckerman on May 17, 2016 at 5:58 pm said:

    We need to have a system that will work any year, not only those with record turnout.

    This is true. However, knowing the difference a higher turnout would have made with EPH might encourage continued participation.

  20. “I would also like to re-order the long list based on upvotes in this round.”

    I’m not sure what you mean here. Do you think nominators should be aware of how other nominators have voted?

  21. Hampus makes the key point, and it’s an ethical one. I’m pretty sure last year was not a “low turnout” year for nominations historically. And even if it had been, so what? The large majority of nominators who were not looking to kick the rest of fandom in the taint don’t deserve to be told, “Look, you just weren’t an overwhelming enough majority.”

    And there is reason to think this year’s jump was driven by anti-Pup excitement and may not be sustainable over the long haul. People console themselves that the griefers may lose interest over time. But so may a lot of non-griefers. If the nominating pool shrinks back down by a third, the remaining fans still won’t deserve to have their nominations nullified by a hostile fringe.

  22. @Hampus

    I mean after the second round. Re-order based on the second round upvotes, and take the top 5 of the second round. Not just take out the least upvoted, and base the top 5 on the 1st round nominations.

    Good night all! I look forward to more discussion tomorrow.

  23. @Emma

    The concern is: EPH, as tested against last year’s data, wasn’t as effective as a lot of us hoped. EPH was designed to allow slaters a proportion of the ballots equal to their actual voting numbers. The tests are showing they still receive a disproportionate share.

    Additionally, the slates have shifted from nepotism / ‘look at me!’ to more of a destructive ‘burn it down’ dynamic. Even if EPH limited the proportions as intended it’s not weak works anymore but works designed to embarrass the awards. Some of them are edging into slander and libel territory depending on your jurisdiction.

    This is all combined with an intentional slowness to making changes with the WSFS Constitution. Usually this is a feature not a bug. No one wants one year’s WorldCon able to change the rules and deliberation is a usually a good thing. Except a lot of folks see this as an a existential emergency. Mandated deliberation periods make effective action difficult when speed is demanded. If EPH fails folks want to have a second arrow nocked and ready to go.

  24. @Mokoto: Hi! To extend my party analogy a little bit further, some people have noticed that there’s a small group of guests who hate everybody else and have dedicated themselves to making everyone else miserable. One guest, let’s call him “Damien,” demands that The People Who Take Care of Such Things throw the bums out immediately! Not so much because the jerks are dedicated to making everyone else miserable but because the jerks have terrible opinions, but whatever. Another guest, “Kevin” in our story, points out that, in fact, the party doesn’t actually have any rules for getting rid of jerks attempting to make everyone else miserable, which is a big problem when it comes to getting rid of jerks attempting to make everyone else miserable. But hey, here are a couple of rules we could put in place that will make all of us collectively The People Who Take Care Of Such Things.

    It’s fair to say the difference between me and “Damien” is that I recognize Kevin has a point.

    Now, the analogy breaks down in the sense that the issue isn’t throwing people out as such. We can tweak it: the party has a karaoke machine and the jerks – who, I remind you, hate everyone else at the party and want them to be miserable – have figured a way to hog the machine using the existing signup sheet system. And to crank it to maximum volume, so that it’s unpleasantly loud throughout the entire house. And all they play is the Barney theme song, except they make up their own lyrics about the sexual inadequacies and imagined crimes of other party members.

    In this analogy, Plus Two is a new rule that after every five venomous filks of the Barney song, two people get to sing something else. Then five more malicious Barney filks, and so on.

    In the long run, it may be easier to just grind things out with a series of nominations and no awards. It’s not as cathartic, but a big NO WAY from the electorate has its merits.

    No, I’m sorry, it does not.

  25. Laura;

    “I mean after the second round. Re-order based on the second round upvotes, and take the top 5 of the second round. Not just take out the least upvoted, and base the top 5 on the 1st round nominations.”

    Yes, exactly so! Thay was the original proposal for Double Nomination.

  26. I’ve been discussing this with various people IRL, and I’m coming around to the idea that the ideal solution (starting two years from now) is:

    EPH+ (which can be written as a one-paragraph proposal)
    3SV (with a few modifications, discussed below)
    Retroactive fix (also explained below)

    3SV should be run on as tight a schedule as possible; it is only a chance for an “immune reaction” and should not be an issue in most years. Also, voters should be able to add one nomination per category to their existing ballot after the long list comes out; this allows for some last-minute reading and consolidation of consensus, without giving enough scope for griefers to veto anything off the longlist. EPH+ should be used for both phases of 3SV

    By “retroactive fix”, I mean a proposal that states three things:
    -Works that come below “no award” will not be considered as having been finalists. Possibly make up a new name for them.
    -For any “no award” category in the last 5 years, we have a special “retro hugo” re-vote. The ballot shall consist of the 5 works which came closest to nomination, plus any works which were withdrawn from consideration. Any withdrawals from these ballots will not be replaced.
    -In the future, any “no award” will trigger a vote during the following year’s nominating phase as to whether or not to do such a “retro hugo” round during the voting phase.

    As to 4/6 (or 5/6) and !5%: with all of the above in place, I think these hardly matter.

    What convinced me that 3SV is a good idea? The realization that it could be done on a tight schedule, without wasting too much time.

  27. Can it be run on short schedule when snailmail is still supported?

  28. @Jim Henley

    Working under the assumption EPH is combined with +2 then you end up with 3 to 4 untainted noms not 2 as in your analogy. Mixing threads: I know you see +2 as a geek social fallacy. I’m assuming you mean trying to apply an engineering solution to a social problem. Any of the nomination proposals fit that criteria though. +2 doesn’t solve the problem but opens a bypass. It may be less a geek social fallacy than the others in that it doesn’t promise a silver bullet.

  29. @stoic cynic: Hi. I think I’m more thinking of the “We include everyone” social fallacy. Plus 2 is “Yeah, Rachel Swirsky has to deal with a fantasia about violence against her person being a ‘Hugo Finalist’, and for the rest of her life and beyond it’ll be on the historical roll of finalists, but we’ll also have a couple of nice things on the ballot too.”

    DN and especially 3SV aren’t themselves mechanical solutions to a social problem as giving fandom the tools to address the social problem. It doesn’t absolve fandom of responsibility for the social order, but rather makes fandom responsible for it. This is precisely what leads to the concern – was it you who expressed it? – that fandom could use downvoting provisions in an unworthy manner. Fandom could! Just as we could toss someone from our weekly RPG group for bad reasons as well as good. (I was trying to find my original comment of late last month likening EPH-type proposals to the classic attempt in roleplaying groups to solve social problems at the mechanics level, but my search fu was weak.)

  30. Can it [3SV] be run on short schedule when snailmail is still supported?

    Possibly not, at least not without rush mail. But I’ve heard that at least some of the few remaining snail mail voters are OK with that. The purpose of the negative votes in 3SV is not to help order the list from which the finalists will be drawn, but to allow fandom to rise up against potential finalists that are clearly offensive to community standards (however the majority defines that). I’d expect most of those cases to be more or less clear-cut, so it is reasonable to hope that a few snailmail votes more or less will not make a difference.

  31. It’s possible that EPH+ could be adopted this year; the Chair of the Business Meeting has explicitly not said that he’d rule for or against it being enough of a change to require its own two year process. In any case, I don’t see how it can hurt to raise the question at the meeting to get an official ruling.

  32. As far as I can see, the simplest longlist option is to just publish the current top 15 in each category before the nomination period closes. Those who wish to change their nominations as a result can do so, while those who don’t want to change anything don’t have to. If that results in a couple of hundred extra nominations each for the top five non-slate works in each category, the slate works would be pushed off the ballot entirely.

    @Jim Henley: In your scenario, I’d be inclined to drop your nomination for the story that didn’t make the top 15, and add the three you liked okay. With any luck your two favourites will make the final ballot, and you can rank them above the other three then. If the okays make it but your favourites don’t, well, that’s a pity, but it means more people nominated the okays, and your individual nomination is unlikely to have made the difference.

  33. @the little mole:

    In your scenario, I’d be inclined to drop your nomination for the story that didn’t make the top 15, and add the three you liked okay.

    Thanks. Can we all agree that this is already not “voting for what I love?” (And it’s very much less so if I kind of disliked two of the three stories.)

  34. @Jim Henley

    Who knows? There’s been a lot of threads between wherever and here I’m sure 🙂 Certainly potential gaming or misuse of a new long list round are concerns I have. One of the rules of computer security is the more complex a system is the more prone it is to vulnerabilities. It feels like we’re layering additional complexity with almost any of these solutions. I have no specific vulnerabilities to point to but I suspect they’re lurking in the weeds. I favor + 2 as much for simplicity as anything. It’s not a silver bullet but at least has predictable failure modes and limitations.

  35. the little mole: Except that, if what I remember reading about the voting stats for the Hugos is correct, a vast number of the Hugo votes are last-minute. Or at least last-couple-of-days. I know that I was adding to my ballot as late as the final day of voting eligibility, and I know that I’m not alone in that. I don’t like the idea of publishing a longlist before nominations close because, honestly, it’s giving a disproportionate bullhorn to people who nominate first.

  36. @Stoic Cynic:I think a major asset of 3SV is its relative lack of complexity. I think under 3SV you could even sunset EPH after its initial three-year run. EPH+ and (EPH+)^D*RX and (EPH+)^D*RX MOD(Q) or whatever are way past the levels of complexity I even want to read about in detail, though.

    My burgeoning hatred of Plus 2 does not need another rehearsal, so I will exercise unusual restraint on that topic now. 🙂

  37. EPH is not actually all that complex.

    Look: with first-past-the-post, you just work top down because that’s quicker. But it would be exactly equivalent to work from the bottom up, right? Count up the nominations, toss out the one with fewest – repeat until you’re down to five.

    EPH works like that. It goes two at a time, instead of one: it uses a measure of popularity to find the two least popular, and then compares those two on a different measure, to eliminate the one that loses. It then reweights the selection mechanism, so that people who lost that one from their ballots have their other works better protected; in this way it tries to increase the chance that everyone gets something they nominated on the shortlist. Repeat until you’re down to five.

    Is that really so hard?

  38. @Jim Henley: if you disliked two of the “okay” stories then only nominate the one you didn’t think was that bad. And you did “nominating what you love” before the longlist was announced; this optional step is picking your preferences out of what other people love, just like the final voting. Don’t nominate anything you’d rank below No Award.

    @Cally: the longlist publication would be the new last-minute deadline, and most people would get their nominations in before then.

  39. @the little mole: So, the griefers will have no such compunctions in the middle stage of DN. So if non-griefers stick to “nominating what they love,” the griefer impact is magnified here too. But if non-griefers need to vote strategically then you might as well go whole hog (3SV).

  40. Pingback: The Hugo Vote Vote Ramble | Randall P. Fitzgerald

  41. the little mole on May 17, 2016 at 7:55 pm said:

    As far as I can see, the simplest longlist option is to just publish the current top 15 in each category before the nomination period closes.

    Which means that Griefers will simply wait until the last day to dump their nominations into the system.

    I’m amazed at how many people think that nominations are spread evenly throughout the nominating period. They aren’t. Most nominations come in during the last week or even the last day. Interim results are meaningless.

  42. @Jim Henley: except 3SV is significantly more complicated. The griefer impact isn’t magnified, because they already know what they want to nominate and publishing the longlist doesn’t change that. And only a small proportion of non-griefers need to nominate strategically to overwhelm the griefers. Even if nobody changes their nominations, a longlist could encourage enough additional nominators to make a difference, at least in some categories – reading 10-15 short stories in a couple of weeks isn’t an excessive burden for many people.

    @Kevin Standlee: why would it matter whether or not the griefers wait? If they wait, then there’ll be 15 decent works on the longlist instead of 10, and those 15 will still get a bunch of extra nominations as a result of being longlisted. Maybe the nominations would be slightly more dispersed that way, but non-griefers massively outnumber griefers, so that’s not a problem.

    People nominate at the last minute because it’s the last minute. If nominations had been announced as closing on March 24, people would have nominated a week earlier, and it they’d closed on April 7, people would have nominated a week later. If a deadline for being counted in the interim longlist is announced, most people will submit their nominations in the week before then.

  43. I would just like to add a quick “Amen” here to basically everything Jim Henley’s written in this thread.

    Carry on.

  44. the little mole: As far as I can see, the simplest longlist option is to just publish the current top 15 in each category before the nomination period closes.

    Thus disenfranchising everyone who hasn’t submitted their nominations yet — which will likely be way more genuine nominators than griefers.

    In other words, what you’re proposing is that the Hugo Admins say to all the Worldcon members, “It doesn’t matter that the deadline hasn’t arrived yet, we’ve decided what the longlist will be, so those of you who haven’t submitted your nominations are S.O.L., and we’re declaring the deadline moved to right now.”

    Seriously, did you even think about the implications of this before you proposed it?

  45. @JJ: or the Admins could announce the deadline for inclusion on the longlist well in advance, and anyone who cares can make sure they get their nominations in before then.

  46. The little mole:

    “@Jim Henley: except 3SV is significantly more complicated.”

    Honestly, I can’t in any way see how someone could say that 3SV was complicated.

  47. the little mole: or the Admins could announce the deadline for inclusion on the longlist well in advance, and anyone who cares can make sure they get their nominations in before then.

    Tell me, please, because I’m dying to know: why would anyone bother to nominate after the “longlist inclusion deadline”, when their nominations will simply go off into la-la-land, and not count for anything?

    The “longlist inclusion deadline”, whatever it is, would have to be the nominating deadline.

  48. @JJ: their nominations would count if they nominated works that are on the longlist. That’s the whole point – getting more nominations for the most popular works instead of the long tail.

  49. “That’s the whole point – getting more nominations for the most popular works instead of the long tail.”

    I’m not fond of this idea to create a slate in the middle of nominations.

Comments are closed.