So, You Want to Fix the Hugos…

But you don’t want to “fix” the Hugos.

J. Greely over at .Clue Snagged a comment off of “Making Light” that caught my attention.

“My view is that when we specifically try to change the rules to exclude the Sad Puppies, and we judge how well the changes work by how well they would have excluded the Sad Puppies given historical data, we will have some difficulty explaining to journalists that we are not doing it to exclude the Sad Puppies.”
– J. Thomas, commenting on changing the Hugo rules at Making Light

Those folks have been spinning themselves in circles trying to figure out how to change the rules for Hugo voting to try to prevent Slate voting. They are, of course, ignoring the causes and looking at the effects, and the causes have been known to the WorldCon TruFen for ages, but they never wanted to do anything about them because they benefited them.

Okay, enough sniping, lets look at the REAL problems with Hugo voting. And it’s not something that’s going to be fixed by Weighting and exponential multipliers and block vote detection (Which assumes characteristics of the current vote that are not in evidence until the results are released later this year, and could have unintended consequences in the future).

First of all, if you want to protect against bloc voting, you need to have a Hugo Electorate large enough that assembling a truly effective bloc is impossible. Some have said that lowering the supporting nomination price would increase participation. That may work, but others have said that the costs of processing the Hugos are at break-even at $40. Not having a crack team of Combat Accountants at my disposal, I can’t say if that’s true, or if the increase from circa 500 voters to 2100 has made it more or less expensive per voter to count. (If there are no economies of scale, someone needs to look at the process.) Maybe you could have random prize drawings, or, maybe you could make a bigger deal of the value of the Voter’s packet.

But perhaps you need to look at discouraged voters, people who want to participate, or who have participated in the past but have dropped out, and see why they don’t feel all that great about voting for the Hugos.

One of the typical complaints I’ve seen is that the voters don’t think their vote is effective. Every year they pay their membership, cast their nominations, and feel like they’ve accomplished nothing when nothing they voted for makes it to the ballot. I know that must suck, and the TruFen have certainly gotten a taste of that this year, so I’m sure they understand (Bad Mauser! No Cookie!). They also seem to get disenchanted by the things that do get through that they think are not worthy of the Hugo name. See the previous post for more on that.

The problem is, there are thousands of works published every year, and only five slots per category. And as has been pointed out in many posts about the system, that makes a lot of stuff get lost in the noise. But it also makes the system very vulnerable to small cliques and blocs who can force anything onto the ballot with 10% of the votes (again, a larger electorate makes 10% much harder to achieve). Thus putting factions on the horns of a dilemma, how to keep the other faction out without hurting their own chances. (Okay, I get it, no Cookie, but I’m on a diet!)

Well, forgetting factions, the real problem here is the dilution of the nominating votes. If five hundred things each get a single vote apiece, that’s a significant number of ballots that have no effect on the outcome. No wonder people feel disenfranchised. Some folks are proposing all kinds of complicated formulae for adjusting the cut-off of the asymptotic curve, but they’re trying to fix a dynamic problem with a solution based strictly on their perception of this year’s data. The approach is doomed to fail merely because of its overcomplexity, and it’s only aimed at salving their bruised egos at losing, rather than addressing the basic problems of voters on the fringes.

So, what is my proposal?

The Two Stage nomination process:

Here’s how it works. As before, each supporting member gets a nomination ballot where they can enter five items, unranked, per category. Then, the top twenty-five vote getters are announced as the Nominee Pool. Next the Voters get to vote on which five of those 25 they think should be on the final ballot. And then finally, the ranked, Australian-rules ballot is cast as normal.

Why would this work? First, with regard to slate voting, the best they could take with their five nomination pool slots is five places. Even two competing and opposite slates could only get 2/5 of the slots. Second, with regard to disenfranchised voters, even if all five of their nominees fail to make the pool, they still have a chance to make a difference when it comes to deciding the final five nominees. All the marginal voters are now up for grabs. They are still a part of the process. They may not have gotten the works they want on the ballot, but they’re still involved instead of being shut out. So now, a slate would have to get a majority of the votes to advance itself, rather than the typical 10%.

To be sure, there are fine tuning details to work out. For example, I would lean towards including all the results of a tie in the pool, unless they are single vote getters, or maybe two vote getters, so long as the pool doesn’t get too big. The threshold should be set to keep it between say, 25 and 50.

I’d also suggest making the process electronic only. It allows automating the process, and we’re practically there already, with only three paper votes this year.

The Second Problem

There is one other common Hugo complaint. The best Dramatic Presentation Short Form being all episodes of Doctor Who. I will give the SP campaign props for bringing actual diversity to this category this year. But how can this be accomplished without a slate? Only allow one episode per series? That could get ugly.

My suggestion: Nominate a Series, rather than a specific episode. Not even a Season, since seasons so often span a year end. If a series airs an episode in say, 2016, it is eligible for 2016, regardless of whether it was the last episode of Season 5 or the first episode of Season 6 that tickled someone’s fancy.

What do you think, sirs?

Advertisements

75 thoughts on “So, You Want to Fix the Hugos…

  1. I like nominating a series instead of an episode.

    Regarding the proposal for the two stage nominations, I like it, but it is equivalent to adding an extra round to the Hugo voting, and it has been pointed out that a three-round process would not be acceptable for the already overworked administrators.

    My own impression is that the process is very inefficiently managed. Each convention apparently makes war on its own, when the voting system should be exclusively electronic and the web server should be the same for all conventions. A 3-stage process should not be too much extra work for organizers, and there certainly should be an economy of scale.

    I also think the cost of voting rights should be lowered. I understand that supporting members get physical material sent to them. That should not be done. Everything in electronic format. That way the marginal cost of a new supporting member should be zero.

    • If it goes all-electronic, then it can be automated, and standardized: hell, there are enough Open Source Coders in fandom to develop a standard Worldcon Nominating and Voting Platform. Load it on a Linux Box, load eligible voters, it spits out emails for Initial Noms. Deadline hits, run the Reports, announce the Nomination Pool. Open Nomination Voting, run for announced period, cut-off and run report: formal announcement of Nominations. When Voter packet is ready. email download link to THAT as well. Open voting, get votes, roll credits, cut to commercial.

      Other than running reports, the only REAL work in that is sorting through the initial nominee candidates, for mis-spells, etc. . ..

  2. This was an interesting post.

    This is practically the same as one of the ideas that were thrown around in the Making Light thread, so you’re not suggesting anything completely original.

    The two-stage nominating vote process has it’s advantages, but I’m not sure that a third round of voting would really make it attractive for large numbers of fans to nominate and vote. In this kind of system, the second stage where you get a long list of possible nominees, I guess I would feel sort of uncertain about how many of the works I have to read in order to make an informed decision. Maybe you could explicitly state that you aren’t required to do that, I don’t know. It would be better than the current way things go, though. This could also be combined with the proposed “4 nominations, 6 on the ballot” thingie.

    However, I still prefer a system where there’s one round of voting and some kind of weighted voting system in place to play down the effect of obvious tactical voting (SAV, PAV, RAV or STV).

    Nominating TV series as a whole makes sense.

    • So I started looking at the posts about that. And people think the Hugo Administrators are overworked now! And as it is, people have enough trouble understanding the Australian rules system. Those things are downright ugly.

      Two-stage voting is more understandable, and gives the part of the electorate the satisfaction of feeling they had a substantial role to play in the final results. Which keeps them involved, keeps the electorate large enough to discourage slates, and doesn’t require a math degree to administer (i.e. it’s easier to anyone else to check the results.)

      I mean, they were just talking about a small example with 16 voters and 10 candidates, and had to consider 120 different possible combinations, when the first round of nominations can have a hundred or more titles suggested and over 2000 voters… times 16 categories.

      Maybe if supporting memberships were $100.

  3. That may work, but others have said that the costs of processing the Hugos are at break-even at $40. Not having a crack team of Combat Accountants at my disposal, I can’t say if that’s true, or if the increase from circa 500 voters to 2100 has made it more or less expensive per voter to count. (If there are no economies of scale, someone needs to look at the process.)

    If this is the case, then they’re already screwed, simply if everyone who is actually eligible to nominate and vote does so. Renovation 2011 had 4100 attendees and 5500 members, Chicon 2012 was 4700 and 6200, LoneStarCon 2013 was 4300 and 6000, LonCon was 7950 and 10,800. And Hugo totals for those years were 1006 noms and 2100 votes, 1100 and 1900, 1350 and 1850, and 1900 and 3600, respectively. (Data from http://en.wikipedia.org/wiki/List_of_Worldcons and http://www.thehugoawards.org/hugo-history/ and sometimes the link to a given Hugo year isn’t enough, you have to go to the end of the page and look at the PDF to get both totals.) So, already, only something on the order of 25% of people get involved in the nomination process, and somewhere between 33% and 50% participate in the final voting.

    So if only just everyone who is allowed participates, they’re hosed if they are that inefficient.

    Heh, which might mean that they should rule out “Australian Rules”, since Australia makes voting mandatory. ;) More seriously, if we’re going to get into esoteric voting systems, I prefer a Condorcet compliant system than a Condorcet non-compliant method. (“if a candidate would win a head-to-head competition against every other candidate, then that candidate must win the overall election”.) “Australian Rules” instant-runoff isn’t Condorcet compliant. Of course, I don’t think any of that does much good to reduce the effect of “slates”.

    The simplest solution I can think of that does reduce and/or eliminate the effect of slates is to simply eliminate the 5 candidate limit for the final vote, and put everyone who gets 5% of the vote on the final ballot. If the Gonosypherpalitic Juvenile Canid Slate votes in absolute lockstep, all 300 of them, then yes, their 5 candidates will be on the final ballot. But since they only comprise 14% of the total population (historically speaking), there could be as many as 17 other candidates. And even in that case, we’ve still knocked the pool for the winner of “Best Novel” down from “Everything published this year” to “twenty three potential options”. And administering the final vote has got to be easier than dealing with the nomination process, because with the latter you have to manage to coordinate all the votes of everyone in a dozen gross different forms of the title and author, some of which may be very misspelled, and in the former you just say “Here are your options. Pick.” Any first year CS student could do that, and probably some number of first graders, as well. ;)

    Oi. Well, there ya go. Have a wall of text.

    • I can get behind either one of these solutions, and I’m a Sad puppy. It’s never been about getting rid of the Message fict. some folks like that sort of thing (though the Gods alone know why). It’s been about tearing down the walls and making the gate keeper irrelevant.

  4. Any system can be gamed, and with the ‘prestige’ of the Hugos, well the prestige they used to have, it was inevitable that a group of people would game it (and anyone who thinks it hasn’t been gamed for the last decade is being foolish). What we’ve seen here is fairly typical, one group took them over and did everything they could to hold on to their privilege.
    Now another group has come along, and the old group is all upset and butthurt, even though no rules have been broken, and no ethics compromised. To be honest, the new group is more ethical than the old group, because at least they’re doing it out in the open.

    There are really only two solutions to this problem:
    1) Get rid of the Hugos
    2) Rename the Hugos, to show that they’re just a reward from a small, insular convention.

    People are saying that the value of the Hugos is now diminished, and they’re right. But what many don’t realize is that the value of ALL Hugos going back ten to twenty years is now diminished, because we can all see now that the ‘fix’ was in. As an award of excellence or quality, it no longer is, and hasn’t been for a very long time. Trying to recapture that (as Sad Puppies was), is no longer worth the effort.

    I’m starting to think that the idea of burning them all down just may be right, look at all the damage the awards have done in the last twenty years to SF&F, and look at all the damage they are doing now. And the behavior by the ‘SMOFs’ and ‘Trufen’ involved has really been pretty despicable. Derbyshire was correct of course, and we see once again just what the fallout of his predictions are.

    • Have you SEEN some of the other proposals? They have such obscure math that it would totally kill any public interest in actually voting for the Hugos.

      Some people are so intent on trying to shoot down Slates that they don’t realize they could kill the awards in the process. They could create a system where something that is genuinely popular and gets a lot of votes from all quarters could be excluded.

      • Most of the rule-change proposals I’ve seen consist of complicated schemes that wouldn’t do much of anything but boost the workload of the awards administrators, or attempts to enable the purposeful tossing of BadThink votes. The first is pointless, the second would simply be proof the Sad Puppies are right.

        The closest to workable I’ve seen proposed is “4 to 6,” which would give each nominating ballot four nominations in each category, and expand the final voting ballot to six finalists instead of five.

        The best solution, of course, is simply to increase voter participation among the con members in the nominations process. More democracy.

      • Doc, I like your original idea (I can also get behind an alternative suggested above).
        Re this comment, “They could create a system where something that is genuinely popular and gets a lot of votes from all quarters could be excluded.” You say BUG, they say FEATURE! One side of this is about tearing down walls, the other side is about reinforcing them puppies.
        I disagree with John, “As an award of excellence or quality, it no longer is, and hasn’t been for a very long time. Trying to recapture that (as Sad Puppies was), is no longer worth the effort.” Bullshit. That’s the same sort of defeatist thinking that says “tear it down, let it all burn” regarding the current issues we have in a divided United States. Tearing down is easy. Building on a charred dismantled corpse is HARD… Fixing is almost always easier, and surely less expensive.

      • Frankly, if you can’t understand simple math like “you get one nomination, divided among the things that are still in the running”, I’m not sure it’s worth reading your ideas on the subject.

        It also leads to your making flat-out false statements like your second paragraph.

        It’s also kindof ironic that your dismissal of “math” leads to your missing the fact that the leading ML proposal (SDV-LPE) expressly deals with issues like “Doctor Who got four nominations” — because that’s exactly the same tabulating flaw that lead to RP getting 3-5 nominations this year. And does so without nominators having to do anything differently than what they do now.

        • And if you can’t learn to read dates, maybe you shouldn’t be engaging in the debate at all. At the time I wrote that, SDV-LPE was not the leading proposal.

          I still think it’s dumb and overly-complicated. And It doesn’t address some of the issues that have been frustrating Hugo voters for a long time, since long before people became aware of slates, or cliques.

          • I totally didn’t notice the age of the post, no (oops).

            At the time you wrote it, though, the leading proposals were SDV (divide your vote X ways) and RAV (discount your vote as things you like make the ballot). Which are both still simple proportional systems, and hardly “obscure math”.

            There are two basic problems to the Hugos:

            1. There aren’t enough people nominating, making it too easy for a dedicated minority to manipulate the nomination process.

            2. The chosen nominations are the 5 things that the largest groups like most, not the things 5 things that are most liked amongst the entire pool. So an organized group with 10% of the nominators can knock out the things liked by 36% of the nominators (in 9% groups). This means that in years with a lot of politics, the nominees may only represent one point of view, whereas in a good system it would represent the top five points of view unless one of them heavily outnumbered the others.

            The first problem has been attacked in numerous ways over the years — from marketing the Hugos, to individual authors mentioning the Hugos to their fans, to changing the WSFS constitution to allow conventions not to have to offer all publications in paper (ratified last year, so Sasquan was the first con affected by this), to changing the rules setting a floor on supporting memberships, etc. And yes, the whole SP/RP controversy may end up helping with this overall.

            The second problem isn’t a social problem–it’s a rules problem, and deserves a rules solution. The simplest is just to let people only nominate one thing per category, but that has other issues, so we ended up testing out proportional voting systems instead, and they seem to work well.

            • The proportional voting systems seemed needlessly complicated, especially after seeing some of the discussions about having to re-calculate everything after someone withdraws, and the possibility that #7 might advance instead of #6. (I remember one of the early posts in one of the threads talking about 16 candidates generating 120 different outcomes to resolve, and in the real world, there are hundreds of candidates and at this point, in excess of 2000 voters. It doesn’t scale well.).

              The biggest problem you’re going to have with getting people to go along with this is the idea of a vote not being an integer. (SDV Single Divisible Vote) if I’ve been reading that correctly. It would be an easier sell if you gave every voter five “Points” they could assign. Although that would run afoul of the tradition that you can’t vote for the same thing five times. But you have that problem with SDV if the person doesn’t divide his vote, his might weigh more than someone who spreads his around.

              And this is just the nomination process! The Australian Rules final vote already confuses the heck out of people. Anything more complex than simply counting votes is just asking for trouble.

              My proposal addresses your two issues this way: By having a “Primary” you have a way of screening down the pool of eligible works from the hundreds of titles published every year, to a manageable number of the most popular ones. Plus, even the most dedicated group could only force five items into the pool of 25. Then, by having the actual nominees chosen in the second round by the entire electorate from the pool of 25, you eliminate the power of the small dedicated group that in the past could win just by being 10% of the electorate. Now people who might have been knocked out of the voting process because they were the only one to nominate their five choices, has a voice again (This also addresses the size of the electorate issue by helping un-discourage a voter).

              I came up with this by looking at the common complaints about the Hugos, such as “Nothing I nominate ever makes it” and the infamous “10% of the voters choose the winner.” (And I’ve seen enough analysis that suggest this has been going on far longer than there have been any puppies.) It doesn’t even have to be a “Conspiracy” just a natural outgrowth of a small cadre of like-minded fen with similar tastes and a dedication to WorldCon. It’s even more obvious in the Fan categories, (“Is it going to be Langford or Glyer this year?” “There are enough people to get Taral Wayne and Steven Styles on the ballot time and again, but never enough to award them a rocket). You can imagine those folks would prefer to keep their little group little.

              So rather than harness math to decide the proper course of the awards, I figured it would be best to put the decision back into the hands of the fans, by giving them a second crack at it. You don’t have to worry about false positives, nor about organized slates. Hell, we’ve already seen that slates can attract a lot more votes, it’s a self-solving problem. (Vroomfrondle and Magicthize could ride THAT gravy train for life!) Even if a Clique could put together 20% of the electorate, AND get them in perfect lock step (Which any honest evaluation of the Sad Puppies shows is downright impossible.) By the time they get through the primary round, they could be shut out entirely by the other 80% of the voters.

              Also, I think they should move to all on-line voting. Have a website, you get access when you get your membership, and you can enter and change your votes right up until the voting closes. Likewise for each round. Thus if you read something in January that you think is good, you can put it on your ballot rather than trying to remember it 10 months later.

              • It could be 60 votes rather than 1, but it doesn’t really make a difference. It’s easier to explain as a single vote, because “one person, one vote” is a useful (and in this case, correct) meme.

                Re recalculating votes, that’s not really needed — it’s a technique that could be used to make replacements more “fair” (maybe), but it’s a needless complication.

                The thing about 120 calculations was a J Thomas invention, iirc–a harebrained idea about using machine learning techniques (in effect) to run an election. Nobody ever took him seriously, because seriously, the idea is nonsense. Now -that’s- an example of too much math.

                Your primary idea was discussed in the ML thread, in a variety of ways. The simple answer is that it’s not a good idea, and it (again) solves a different problem than proportional voting does. A more complicated answer is that having a “primary” step drastically increases the workload on Hugo administrators (already a small, overworked group paid with the satisfaction of a job well done), cuts timelines and deadlnes to the bone, would be profoundly hated by the Worldcon voting pool (who by and large are independent sorts and don’t want to be told who to vote for), and also wouldn’t serve a generally useful purpose, despite appearances. Open nomination works becasue people mostly nominate stuff they love. Voting of the shortlists works because people typically look at -everything- in a category and rank the category; the things they don’t get a chance to look at they don’t vote in. But while a “long list” works for juried awards where the jury commits to reading everything on the list they can manage before the deadline, they don’t work nearly as well for a big award with many nominees, where they’d mostly serve to remind people of their favorites.

                In essence, the primary proposal demands more than is currently demanded of fans, and of the administrator. In comparison, SDV-LP does automatic runoffs (which do pretty much duplicate a lot of the effort of a primary, though not the part where people settle for stuff they don’t like that much but which has a chance of passage) with a week of coding (at most) on the part of the admins, and no change whatsoever for the nominators.

                The existing system of people recomending things basically does that job. Having neutral more central places for such reccomendations (with polls so people could test the temperature of the room) might do an even better job, particularly for the fan and other low-nominator categories, but anything more official would smack of vote tampering (and change the character of the award).

                Re online voting, aside from early nomination (which I agree would be advantageous — I think the primary reason this doesn’t happen is the lack of continuity between committees, so the incoming committee may not even have their Hugo nomination site -ready- much before the new year), we pretty much do all that. Sure, there are paper ballots, but only a tiny fraction of the electorate uses them (and we don’t really want to tell them they can’t use paper any more). It’s only a very small amount of effort, and all other voting is online and can be done and redone any time before the deadline (even without continuity, we do share enough tech to be able to do that much, usually).

                • For online balloting, I would suggest the WorldCon Committee authorize a permanent website (Say, HugoAwards.org or something) that would implement the ballot system (whatever it is) and allow the local Con Committees access to it as their Cons come up. It would also make it easier to keep track of things like the two-year span of people’s voting rights.

                  I think that that would also make a significant reduction of the workload on the committee (regardless of the system decided upon), and the Con could of course enter paper ballots on the behalf of a fan who chooses to go that route. All three of them this year.

                  You say my suggestion solves a “Different Problem”, how so? What problem is SDV solving? Is it straying onto territory covered by J. Thomas’ quote? I still think that the voice of the electorate is a better solution to ANY issues than a system that can’t possibly foresee every possibility.

                  I would have to see a poll of the “Worldcon Voting Pool” to know if they would hate the system, or if that’s just the consensus of the audience at ML. They are not one and the same, as much as I’m sure some wish were the case…. :-)

                  • The J. Thomas quote doesn’t represent anyone except J Thomas.

                    The issue that SDV solves is that currently, you don’t have one vote — you have five. Everyone gets to equally vote for five winners–which means that a highly coordinated group of 10-20% can’t just control 10-20% of the nominees; they can control all of them.

                    This is a bad situation if you’re not in that group. And eventually, everyone will be, since fire is fought with fire.
                    It’s also a bad situation if you’d prefer that the Hugo not be politicized by any party in the long term–since it massively rewards politicization.

                    The 25 nom doesn’t solve this not just because primaries are a bad idea (and the people saying that to a much weaker nomination concept weren’t ML regulars; they were interested parties who were involved in administration, running worldcons, with a vested interest in the awards not being political, etc). It also just doesn’t solve it. If you assume that there’s a group of 20% that loosely blocks (and I’m not assuming here that they strictly block; they don’t have to for this to be the case), then you’ve got, say, 19 votes for each of the 5 “slate” candidates, vs on average 4 votes for each of the non-slate candidates.

                    What SDV (and any other proportional system) does is limit each person to, in effect, picking -one- candidate. As I mentioned, this could also be done by having them, you know, just pick one candidate, but that makes it hard to achieve convergence (this is the problem that the 25 primary helps with, as do suggestion polls, online discussion, etc–but convergence this way (by people taking 6th and seventh choices, etc) is a mixed good, as it can interfere with picking excellence). By concentrating their vote as their candidates are eliminated, voters continue picking a total of one candidate each as long as they still have one candidate in the race–so a demographic that has 40% of the non-wasted ballots (a ballot is considered wasted if they have no viable candidates listed) get on average 2 places on the ballot, etc–no matter what strategy (or not) they use for voting.

                    The point about having a standing website group isn’t a terrible one, but runs into politics. There -is- no “Worldcon commitee” — this organization simply does not exist; the -only- standing organization associated with Worldcon is the Mark Protection Committee, which has very limited powers and responsibilities. It’s not impossible for WSFS to create what amounts to a Worldcon Election Commitee or the like as a standing organization, but it’s not a trivial political process due to people’s resistance to the idea of “WSFS inc.”, historically, as an extreme reluctance to have any group have persistent control of Worldcon.

                    It’s also not an entirely unmixed bad to have different people in charge of Hugo/vote administration every time–it serves as protection against stagnation and special interest capture. But it would be good to have more shared resources, of course. (this is true wrt SDV-LPE, of course, as that ~week of coding only really needs to be once to make a vote tabulator that can then be validated independently).

                    • It seems you’re assuming that in the 25 primary that the non-slate votes would be evenly distributed. Whereas I assumed that if the slate put on something everyone else reviled, they would be outvoted.

                      WRT to picking excellence, isn’t that also a complaint about the Australian Rules final? That it picks the least hated rather than most popular? (One thing I’m not clear on, under SDV-LPE, is that eliminated?)

                    • I’m making no assumptions about the non-slate votes except that if they didn’t form a slate before the primary, it’s doubtful that they’d form a slate after. And since there’s pretty much no time to read 25 works (or even 20 works if they’re discounting the slate based on the authors, cover, first few pages, etc) in 13 categories, the non-slate votes would presumably be a lot less concentrated than the slate votes.

                      Also, no, the SDV-LPE proposal doesn’t touch voting at all, just nomination. I don’t want to get into a discussion of STV here (it’s not really gemane, and doesn’t seem to be broken), but STV as implemented by Worldcon doesn’t pick “the least hated”. It picks the thing that the most people like -best- (remember that something needs to not lose the first place votes; late votes don’t matter at all), which more than 50% of the voters don’t hate. It’s not necessarily the best voting system ever, but it works pretty well for picking a single winner from a group of 3-6. (I’d love to be able to place items tied without leaving them off my ballot entirely, though; that lack of nuance can result in some weird (or incorrect) voting, particularly when there’s something or things people want to put below no award).

                      Looking at the site selection votes for Washington (vs Finland and Florida) is nicely illustrative. Finland had more people who liked it -best- than either Florida or Washington. But the voters who liked Florida liked Washington more than Finland — so because most of the voters liked Washington -more- than Finland, it went to Finland (which makes sense). If, on the other hand, everyone had put Washington second, it would have been eliminated first–this is a quirk of STV, but does make sure you don’t have winners that nobody actually -likes-.

                    • WordPress nesting depth is failing…. Please reply at the root.

                      Short form, I’d like to see some sample ballot sets made up and tested, say 1000 ballots apportioned like votes from the past, and a test case of 100 (10%) slate voters, both in lockstep (impossible) and 50% falloff, and see if the outcomes resemble what is being sold.

      • Now there’s a thought! It really shouldn’t be so hard with today’s technology to tabulate votes. A good website should make a three stage vote process simple, fast an efficient. Drop the added complexity caused by the Australian voting process, and you might have a viable way ahead.

        Also, your opening quote is gold. Hard to explain that all the convolutions they are going through are not to exclude, when that seems to be the only end result they care about.

  5. perlhaqr, someone has already given a plausible non-partisan reason for not increasing the size of the final ballot: all the nominees are invited to the award ceremony. Time and space requirements for the event would make a significantly larger nominee pool unwieldy.

    You could make other changes to deal with that, but it may not be the only legitimate concern, and if you can’t address all such, your entire proposal could be shot down by someone who just wants to go back to a World Without Puppies. (because some people think killing puppies will eliminate puppy-related sadness)

    -j

  6. Reblogged this on The Improbable Author and commented:
    Yet more Hugo nomination alteration suggestions (because it will and perhaps should change in order to address the Sad Pups complaints as well as their opponents on the side of the WorldCon status quo). Perhaps we could mix this with the category changes I recommended for the “Erics”?

  7. “Not having a crack team of Combat Accountants at my disposal, I can’t say if that’s true, or if the increase from circa 500 voters to 2100 has made it more or less expensive per voter to count.”

    If only there were someone trained in accounting, familiar with the science fiction world and business….

    I wonder if such a person exists……

    • He’d have to be a stalwart fellow, able to ignore the slings and arrows of outrageous insult and take up arms against a sea of discontent.

      Familiar with both accounting and the science fiction world? Why such a fellow would have to be a veritable mountain… A Mountain Who Writes…

      I like the cut of your jib, sir, and endorse a plan to have a science fiction accountant look to the efficiencies of the process.

  8. I keep hearing that Supporting Members are a break-even for the Worldcon.

    Now, I have not seen a Worldcon budget, nor have I investigated printing or mailing costs.

    But if cost IS an issue. . . send PRs and Programs as PDFs. Send out the actual Program after Worldcon. Cuts the costs massively.

    And, at a much lower cost per supporting member, chances are, you’ll get a lot MORE supporting memberships. . . .

      • I’m suggesting that the option be removed for physical for anything BUT the actual Worldcon Program. . .Likely cut the cost of a supporting in half, if not more. .

        Lower cost = more voters. . .

    • I for one am frankly somewhat annoyed to be informed that supporting memberships are break-even. Based on the name and description, I was under the impression that they were designed to provide, you know, support.

      • I’m not sure how true it is. Supposedly it includes things like sending out all the supporting materials (program book, etc) but I’ve heard that this year getting them electronically is an option. It could just be a response to people who want to lower the cost to encourage more participation in voting, which isn’t really wanted by those who control the awards now.

        • This is almost a month old, but since our comments have gotten a little less hostile, I figured I should address it.

          In sum, the high cost of supporting a supporting membership comes from three sources:

          1. Historically, the iniitial price of an attending membership was capped (in order to guarantee site selection voters a low conversion rate) at two times the voting (which grants a supporting membership) fee. This meant that to make sure they could set attending membership high enough to not lose money, that cons needed to set supporting memberships ludicrously high compared to what they cost the con at the time. We’ve been chipping away at this; I know we increased the cap to 4x voting; I don’t think it’s changed since then, but that’s already a pretty substantial change.

          2. When WSFS passed an amendment allowing electronic publication support, there was still enough non-digital population that it made sense (theoretically) to require that cons continue to offer paper publications at no extra cost. This requirement was only dropped as of this year, but we still aren’t consistently gathering enough information on the site selection survey to be able to offer paper publications at an additional fee at the point where people register (which is what would really help fees reflect costs). This will likely stablize in the next few years.

          3. Shipping costs have gone up substantially over the last few years. This has a big effect on the costs needed to service a paper-enabled supporting membership.

  9. I think your 3 stage plan is an excellent one however there are 3 big IFs that get in the way. It would work if…

    1) WorldCon could find someone(s) with the skill and desire to set-up and run a web site to handle the process. They cannot pay for it due to the silly nature of how the Con is run.

    2) The goal is actually to have a fair and open process. I do not believe the majority of the other side really wants that.

    3) A larger pool of voters is desired. Again those at the core of the Con want to keep things tiny and controlled. I do not believe they want to see large numbers of Fans involved.

    I think the quote you posted says it quite clearly. They want to take control back from the Puppies. And with that as goal any really fair open system will never be approved.

  10. Two stage voting sounds like a pretty good idea. Regarding nominees attending the banquet, the first round picks would only be semi-finalists, so only those on the ballot get the banquet invite – no change in numbers from the way it is now.

    Just curious – why’d you pick 25 for the semi-finalist level? Was there a numerical reason, or could it just as easily be 20, 15 or 10?

    Regardless of the number, I like it. :)

    • Seemed like a good idea. Five times the current number. Enough to cover should there be multiple competing cliques. Although in some categories you’ll probably run out of nominees before you hit 25, or you’ll be down to authors who only got a couple of nominations. Although the larger the electorate, the healthier the numbers will be.

      The sad thing is, I won’t be at the Business Meeting, so there’s no way I can propose this. I’d love it if someone who WAS going to be there could. Or if a LOT more people would link/trackback, whatever to this post.

      • what would be truly Evil is to restrict nominees in all categories to those that get a certain percentage of ALL ballots. So if only a handful of people nominate for a category compared to the total, the category automatically becomes No Award. :-D

        • Nah, because honest voters have a legitimate reason to not vote in categories they know nothing about. (Forex, if they don’t read the magazines, there’s a good chance they haven’t seen any shorts.)

  11. RE: Supporting Price/Costs

    A supporting membership isn’t just a Hugo voter membership. It specifically entitles the member to all publications, most notably the updates and the Con Book. Printing and mailing these is the “break even” mark mentioned (probably).

    To truly reduce the price, they’d need either a Hugo Only membership, which would only need two ballots mailed out, or an Electronic Only Supporting, which would be PDFs and e-mails and website log-ins.

    Of course, that would cause a bit of a split between Hugo/World Con identity maybe…

  12. If the business mtg is Fri. Afternoon, or anytime Saturday or sun. I’ll be there. Shoot me a copy of your proposals and I’ll make them. Carbonelle is my user name at Juno.com. Subject Hugo rules so it doesn’t get eaten by the spam filter.

      • Here you go:
        How do I submit a proposal? Any two or more attending and/or supporting members (including discounted attending members, as above) can submit business to the meeting. To submit a proposal, send it to wsfs-business@sasquan.org. You can contact us at that address in advance if you need help crafting your proposal into the correct format.

        Since my husband and I are both attending (2 or more) if you send me the proposal we can submit it ahead of time. Deadline is August 6th.

        We’re also looking at me getting a extra day off and going Thus – Sunday, so we’d be able to attend all the business meetings (Th, Fri. & Sat).

  13. “WordPress nesting depth is failing…. Please reply at the root.”

    Ah, WordPress. No love, me. :) Done.

    “Short form, I’d like to see some sample ballot sets made up and tested, say 1000 ballots apportioned like votes from the past, and a test case of 100 (10%) slate voters, both in lockstep (impossible) and 50% falloff, and see if the outcomes resemble what is being sold.”

    If you look on, IIRC, the second of the three long posts on ML, people coded and did -extensive- tests with RAV, SDV, and SDV-LPE, and posted the results (and in some cases, code, IIRC).

    The tests included al lot of slate tests — as well as, from one of the testers who was a former Hugo admin (and therefore had legitimate access to ~1984 Hugo data he could use for testing, but not share), tests against live nomination data.

    In summary, all of the systems under test performed well. RAV was the most slate-punishing (too much, for exponential weights, IMO, as it encourages slates to game by coordinating their votes very artificially). SDV works well, but had the potential issue of symmetry–there is too much incentive to leave off your second favorite work if you think that will help your favourite work and that your second favorite has a lot of uncorrelated supporters. That resulted in SDV-LP — which is SDV where the actual eliminations are by popularity, not votes, giving a work that was listed by a largish segment with other works a huge advantage over a work that was bullet-voted by a much smaller group, while still not letting a slate (or otherwise highly correlated group of nominators) get more than their proportional share of the nominations.

    • How about false positives? Edge cases? As a former programmer, it’s easy to make things work properly when your incoming data is exactly what you were expecting, it’s when things deviate a bit that you can handle it, fail gracefully, or crash and burn.

      • I’m a current programmer, and I haven’t been able to come up with anything where this system is less than fair (except for the tie case, where there’s a lot to recommend having -some- tiebreaker even if it’s more or less random, just to avoid throwing out what’s left of a slate or otherwise paired vote). But while there have been a lot of tests done, independent verification is good; that’s why there’s code posted.

        It’s worth noting that voting theory seems to indicate that -no- voting system is free of artifacts. But some are still better than others.

        • What’s the definition of Fair, in this case? I’m getting a hint that there’s some form of “Slate Detection” and vote invalidation/reduction there. What is a “Paired vote”?

          That does sound like it would knock out the Dr. Who fans if they all nominate the same five episodes….

          • It’s more a question of how you break ties (that don’t currently include last place).

            The classic way to do so (with Hugo-style approval nomination) is to just eliminate all the tied nominees.

            I retained this for my initial draft of SDV, treating the unlikely slate penalty as a feature (I might have been annoyed by the Hugo nominees this year at the time).

            However, it’s a random enough that it can affect pretty much -any- correlated vote; Dr Who, Buffy, Puppy, whatever. As said, it’s an ugly edge case.

            So I’ve switched my preference to want ties that aren’t for last place to be broken–as long as -any- fair method (random is fine here; so is looking at ballot order even if that’s not used for any other purpose; there are plenty of good intermediate tiebreakers, but they won’t help for a too-early tie) is used to break them — because if you break a tie and you don’t immediately eliminate the winner right afterwards, this means that the same voters were voting for both tied candidates, so it’s obviously better to retain one of their preferences than to throw out both.

          • Sorry for the double response, but you asked about fairness.

            A fair voting scheme is one where the results map reasonably on the preferences of the populace.

            I’m hardly a voting expert (I’ve spent some attention on it, both because it interests me and because it’s a basic matter of civics, particularly as a member of the fannish community as well as a US citizen), but I believe proportionality is one of the big considerations here — that in general, representation should map to the preferences — in proportion to their numbers — of the voters. If 10% of the population voted for each of 7 books (70% total), and 30% voted for 5 books in varying numbers, the fairest result would be to have 10 nominees — the 7 books that each got 10%, and three of the five books voted for by the 30% block. Of course, if you’re picking 5 winners, that’s hard to achieve–but any system that didn’t give any of the groups -more- than that share out of the 5 would probably be pretty reasonable (for that difficult case).

            • If the goal is to pick the best nominees, trying to be “fair” by finding some way to give a boost to works that didn’t get as many votes seems counterproductive. What you’re saying is fair reads more like taking the top member of each Quintile when it comes down to it.

              Aside from the fact that it never breaks down that evenly. With something like a hundred different titles getting at least one nom, the graph of number of votes looks geometric, if not asymptotic.

              With the limited information we have, even the Puppies had an incredible rate of defection from the slate (with top slate candidates getting something like twice the nom’s of the bottom – understandable since many of them said they only nominated the things they’d actually read, or had their own substitutions) and there’s no way to distinguish those from independent voters who came to the same conclusions.

              That we’re so in the dark, and that this is the first time something like this has happened (at least openly), this is really not a very good basis for a radical change in policy.

              I heard a really good analogy though. Let’s say you’re in charge of buying the Ice Cream of the company picnic. You decide to ask everyone what they’d like. Of course, a lot of folks say they’d like Chocolate, or Vanilla, or Strawberry, but then there’s the few who say Rum Raisin, or Butter Brickle. Unfortunately, you can’t accommodate all of them, And being Fair as in your example, you’d want to include at least one of the “Minority” flavors. Unfortunately, Mr. Butter Brickle isn’t going to be too happy when Rum Raisin gets the nod, just because two people wanted that. And it’s a bad idea anyway, since you’re going to be buying 5 gallon tubs.

              Now, applied to Two pass, after you ask everyone, and pick the top few flavors, and then ask people to pick from THAT list, you’re more likely to get a selection that everyone can live with. There may not be Butter Brickle OR Rum Raisin, but from a more limited menu, they can at least have input on what will finally be served.

              Heck, even the Australian Rules final is like that. While everyone is evenly split into Chocolate, Vanilla and Strawberry camps, if enough listed Vanilla as their second choice, Vanilla wins.

              And then there are the No Award people, who say “If I can’t have Spumoni, and you tasteless guys are going to waste money on Vanilla, then Fuck it, I don’t think we should have Ice Cream AT ALL!”

              • It’s not a great analogy, but I can improve it.

                Lets say you’re picking 5 flavors for the company picnic. And everyone can choose 5 flavors.

                Now, most of the company isn’t that interested in chocolate. I mean, they like it, but most of them like something else more. And there are a fair number of people allergic to chocolate, so it hasn’t made the picnic for a while. But people like lots of different flavors, so the flavors at the picnic tend to change a lot…but no chocolate.

                However, there’s a group that really likes chocolate. Enough that a bunch of them have been avoiding the picnic because everybody knows that they never have chocolate ice cream. So one year, the group gets together, decides on 5 chocolate flavors, and pushes them — and bang, all the flavors selected for the picnic are chocolate. Even though most people at the picnic don’t actually want chocolate. Even though some of the people are -alergic- to chocolate and can’t or won’t eat it (it’s not a perfect analogy here, but hey, chocolate is a thing some people are physically allergic to, so I’m going with it).

                Of course, not having any chocolate ice cream wasn’t optimal. But not having any -non- chocolate ice cream is even worse–not to mention, people were fond of having a process that resulted in a lot of churn every year, and aren’t looking forward to having to organize to defeat the chocolate lobby next year.

                What you actually want is a process that will pick some chocolate ice cream (the most popular one or two flavors among the chocolate lovers), and also the most popular flavors that aren’t chocolate. That way, almost everyone has something they can and want to eat — and people who will eat anything have a good basis to get a meal with the most popular chocolate -and- the most popular vanilla -and- the most popular fruit flavors. It won’t have everyone’s favorite–but it will have a broad enough base that almost everyone has something they -like-, and you’ve got a good chance of including the best.

                That’s what proportionality is.

                And again, what you said about STV is flat-out false. Something that is everyone’s second choice will -always lose- in STV. A work simply cannot win in the Hugo voting system if it has no first place votes, because the works with the fewest first place votes are eliminated -first-, before anything else happens.

                • My poor analogy! You have tortured it nigh unto Waterboarding!

                  My point in picking ice cream is about TASTE, as in everyone has his own preferences, and no one flavor is necessarily “better” than any other – at least not objectively.

                  And now you have taken this idea of a matter of taste and turned it into something physically harmful (allergies) and made the folks who like chocolate into an evil conspiracy out to inflict their choices on an unwilling company picnic. (I suppose that would turn the No Award crowd into the Lactose Intolerant).

                  Besides, if you REALLY looked at Brad’s nominees, if anything he was advocating for Neopolitan, Tutti Frutti, and even the aforementioned Spumoni. And even there, in your analogy you put the cart before the horse, assuming the conspiracy was pre-existing and the slate came later. Really, not everything is as sinister as you seem to think. (Especially the Right :-)

                  And the unwillingness of some people to even look at creators they are unfamiliar with is rather distressing. “No! NEW people! Not whom I was EXPECTING to be able to vote for! We must Destroy the Hugos in order to save them!” Hell, even GRRM is saying people should read and evaluate and vote normally, and condemning Guilt by Association.

                  Try the Tutti-Frutti.

                  (And in any case, when you call proportionality is more like affirmative action for less popular works, which only guarantees they’ll get creamed in the final popular vote for the actual rocket. Kinda cruel, actually, setting them up for that kind of failure.)

                  (And you misread my point about the finals. None of the flavors had NO first place votes. If I really have to lay it out, let’s say on 1/3 +1 of the ballots Strawberry was 1st, and Vanilla second. on 1/3 of the ballots, Vanilla was first and 2nd was split between Chocolate and Vanilla, and finally on the last 1/3 -1 Ballots, Chocolate, followed by Vanilla. Chocolate is eliminated in the first round, and Vanilla wins in the second (with 2/3 of the second place votes, and 1/6 each for chocolate and strawberry. But really, this is just a side point)

                  • The problem is, any system that is designed to ‘prevent’ a certain type of behavior WILL be gamed and has ALWAYS been gamed. It’s human nature.
                    That is why we have FIFO, because ALL other algorithms were heavily gamed.
                    I don’t like the current system, because it is fairly easy to game, (as the anti-SP people are showing us with their NA vote plan) but at least it’s not as game-able as everything else that has been suggested, and gaming it doesn’t (usually) have the major consequences as in other systems.

                    As for the people bitching about ‘slate voting’ they really need to just get their heads out of their asses.

                    ‘Slate voting’ has ALWAYS gone on in the Hugos and it ALWAYS will. And just where the hell in the rules does it say it’s illegal? Oh right NOWHERE. This is just a bunch of people who were already ‘slate voting’ and ‘log rolling’ in private, being pissed off, because they got beat by someone being HONEST. Yeah, it always burns the people doing stuff in secret when they get trumped by honesty. That’s why they scream so loud.

                    As has been said before, either the people bitching shut up, or they redefine the award to be that of a small convention and a small group of people (in short, take their ball and go home). Of course the problem is it isn’t their ball. They have, in effect, stolen it from everyone else.

                    • True, they were perfectly fine with the nominees being determined by 10% of the electorate when it was the same 50 or so “core” voters, who were in PNH and TNH’s social circle. Now that it’s not, things MUST be changed.

                • Oh, and another take on the analogy came to mind.

                  What Brad did was propose Chocolate Ice cream, and hundreds of employees who were basically skipping the company picnic because they didn’t like what was being offered all said, “Chocolate? Sure, Get Chocolate and I’m there!” and attendance soared.

  14. See, the thing is, Mauser, that “50 or so core voters” is a lie. It never happened. This has been confirmed multiple times by people who actually got to see the ballots, not to mention even a casual analysis of the nominations and ballots. There -was- no slate voting before this year’s Puppy campaign. (except for disorganized buffy/dr who campaigns, which we -did- consider a problem, and still do, it was just that there was no obvious solution).

    What was true was that the nomination counts were such that if someone -really- cared about a work, they could lobby a bunch of their friends and get -one- thing on the ballot. Which is abusable, but not in a way that hurts much; sometimes one thing that probably shouldn’t be there gets on the ballot, but it’s ok. Very different from 10% controlling the -whole- ballot.

    I’m not quite sure why some people on “your” side think that when we choose to change the rules to remove a hole, rather than fight fire with fire, that this is a problem. A hole is a hole; it’s only worth filling once someone starts putting things into it consistently.

    And I’ll note that the result of our “fighting fire with fire” would be worse for -everyone- (except people who expressly want to destroy the Hugos as a serious award). Some years, the ballot would be full of interesting works. Sometimes, it would be full of liberal polemics. Sometimes, it would be full of conservative polemics. In no years would it actually be diverse and interesting.

    • No, someone put up a statistical analysis of all the votes, showing how many votes won each of the major categories, going back quite a few years. Might have been Vox Day, might have been someone else, I forget. And it was pretty obvious that for the most part, voting was pretty well organized for those awards. Only once was there a split in the group, but that was about it.
      And again, even Martin has admitted to the log rolling and such that went on. The very fact that the people at TOR knew EXACTLY who was on the ballot before it was announced is proof that they have been keeping a fairly tight leash on the nominations. Unless of course you believe that there are people at Worldcon who are reporting the results to Hayden under the table?

    • Which conveniently ignores the other campaigns over the years to get Fred Smith’s book nominated, and then winning a Hugo.

      No, I won’t say what writer’s benefited. I have no reason to believe that they had any input into the campaigns, and the books were good enough to win.

    • It’s not just my side who thinks that you’re changing the rules merely to try to further entrench your side, (and language like “When WE choose to change the rules” sure doesn’t help dispell that image). Even George R.R. Martin is pissed at you guys.

      Over at Making Light, and on several other sites, various rules changes are being proposed to prevent this from happening ever again. There are so many different proposals they make my head spin. More nominating slots, less nominating slots, weighted voting, eliminating the supporting memberships, outlawing slates, limiting nominees to a single nomination, juried nominations… on and on and on. The worldcon business meeting is never exactly a funfest, but if the proponents of half these proposals show up at Sasquan, this year’s will be a nightmare. And will probably still be going on when MidAmericon II convenes.

      I am against all these proposals. If indeed I am at Spokane, and if I can get myself up in time for the business meeting, I will vote against every one of them.

      Most of them, frankly, suck. And the mere fact that so many people are discussing them makes me think that the Puppies won. They started this whole thing by saying the Hugo Awards were rigged to exclude them. That is completely untrue, as I believe I demonstrated conclusively in my last post. So what is happening now? The people on MY SIDE, the trufans and SMOFs and good guys, are having an endless circle jerk trying to come up with a foolproof way to RIG THE HUGOS AND EXCLUDE THEM. God DAMN, people. You are proving them right.

      The thing is, 50 core fans doesn’t have to be an organized slate, they just have to be enough like-minded individuals in social contact with each other enough that common voting patterns are inevitable. Say, The Neilson Hadens and their assorted clingers and hangers on, whether they be colleagues, clients, or wannabees who think that brown-nosing will help them get published (or Employees – the rumor being that TOR buys WorldCon memberships for all its employees. Although the numbers don’t support the idea that all 160 of them form a Bloc.)

      The thing is, 50 core fans could have been a force when the electorate was 500 fen. Now for the noms alone it was 2100, and the numbers are skyrocketing for the final. 50 core fans is a powerless group now, and I’m sure they don’t like that.

  15. Also? “No Award” isn’t an abuse. If 51% of the voters don’t like your nominations, they’re supposed to win. That’s–you know–how elections work.

    • But you see, there’s a problem with it when it’s wielded like a political weapon. Remember Sad Puppies 2? That one, run by Corriea instead of Torgerson, was about the political bent of the Hugo electorate at the time.

      And frankly, it was a CLASSIC Xanatos Gambit. He proposed that the electorate was biased against conservative authors, considering how few of them won, or even were on the ballot in the first place. This was excused as “They can’t win if they’re not on the Ballot”. So he set out to get some on the ballot, only one or two (Something SP3 opponents say would be alright, but this history shows is not true) and the SJW contingent set out to prove him right by organizing a campaign to put every single one of his candidates below No Award. he lost the award, sure, but proved his point.

      If there hadn’t been a campaign against him, and he’d merely lost normally, that would have been the end of it.

      And they’re doing it again.

  16. Hi Dr. Mauser,

    Great post! Very refreshing ideas.

    But you will have nominated five items that you have actually read and loved. At the second stage, you won’t have time to read all 25. You might not have read anything else by a certain author. On what basis will you evaluate them?

    It might be more fun to have a long list of a modest size (10 or a maybe little more), especially if publishers are usually willing to make some decent excerpts (or whole short stories) available, with a reasonable period for reading and discussing them before making your picks for a shortlist. I doubt that would feel like a political primary at all. It would feel like fans being excited to have a chance to read and discuss genre fiction, and accepting the responsibility to do a good job of it.

    • Actually, that problem exists at the nominating stage in the current system, there are hundreds of new releases every year, and when choosing their five candidates to nominate, they again couldn’t have possibly read a thousand new books. But what I’m betting here is that those who nominated the more obscure works might have also read the more popular works.

      Nobody has read everything, and even in the current system, people are either leaving off the things they haven’t read from their ranking of the five finalists, or abstaining from voting in some categories (the turnout in the fan categories is notoriously low).

      The advantage of having a huge electorate, as opposed to having a Juried award where the entire jury reads every work is that you can have those gaps in the voter’s knowledge of the works involved, and still get a decent collective picture of the group’s opinion. The larger the electorate, the better that picture is.

      • It is interesting to think about. I would still think that in the second round, if the number of works is that high, you’d essentially have to rely on your preconceptions or what other people say on the internet about them. Sounds like it might be a recipe for disaster. Better limit the long list to whatever number helps provide a kind of “sampler” of the current state of the genre, so you can at least try to get a handle on what they are all like in the time available. What that number is I don’t know but it is apparently not five.

        • At that stage in the game, it’s less important that people have read everything. The importance of the first round is twofold, to prevent a slate (or two) from dominating the nominations without actually having broad support, and to give the folks who were voting for obscure works that didn’t make the cut another chance to have their voices heard. It’s the final vote where it really matters that all the voters have read all five nominees.

          One other suggested system that touches on this idea in a more timid way is called 4/6, which if I recall has people only get 4 nominating slots, but takes the top 6. But it’s still easier to game than this one.

          But the main point of this system is to protect against a surprise slate domination. The main reason a slate, or a clique can control the nomination process (or could, this may be a thing of the past with unprecedented participation this year) is because a great many voters are “wasted” or “spoiled” voting for singularly unpopular, or possibly ineligible works. The finalists might have been nominated by only 15-20% of the electorate. That means that 80-85% of the people voting may have had no say in who the finalists are. (Now the split might not be that clean but bear with it) This idea harnesses those wasted votes and says “Okay, so you didn’t get your brother’s fantasy space knight novel nominated, but who would you pick out of these 25?”

          • Well, “give the folks who were voting for obscure works that didn’t make the cut another chance to have their voices heard” is a good idea. The issue I’d see is that if it is tempting to run a slate campaign now, imagine how tempting it would be to get everybody together to bloc vote the five things out of those 25 that match your group’s biases most closely, without necessarily reading them all.

            • The thing is, currently, 20% could control the ballot completely, and nobody would know until boom, they’re presented with 5 finalists to choose from (This is what the “trufen” are accusing the puppies of, but they’ve been doing it by word of mouth for a decade). With this system, it would be blatantly obvious before any serious damage had been done, and if 80% of the voters only pick the top five of the remaining 20 non-slate items, it doesn’t matter what the gang does, they can be shut out.

              Although with the puppies, virtually every voter defected from the “slate” in one regard or another, only voting for the items they had read being one of the most common, or substituting some items they personally preferred. And they were quite open about it. And what numbers we have so far show this, where the top 5 novel nominees differed by nearly half the number of nominations between the top and bottom entries.

              There will be more data after the con, and people will be all over it, I’m sure.

    • Thanks, I’m trying, but I’ve been a bit buried lately. I was going to write something about the Antoinelli affair, but that’s fading out faster than I could write about it.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s