[Editor’s Note: Dublin 2019 has posted the 2019 WSFS Business Meeting Agenda (July 21 update) [PDF file] containing all the business submitted by the July 17 deadline. File 770 will post about some of the proposals and invite discussion.]
While the Sad and Rabid Puppies slates were filling up most of the slots on the 2015 and 2016 Hugo ballots, majorities at the Worldcon business meetings passed and ratified several rules changes that made it much more difficult for that to keep on happening. The success of these majorities has tended to overshadow how many fans did not want any changes made – no matter how often Vox Day dictated what made the ballot – or else did not want these particular changes made. And there are business meeting regulars who evidently feel now is the time to start turning back the clock.
Here’s a matched set of proposals to end the “5 and 6” part of the Hugo nomination reforms. If you are going to the Dublin 2019 business meeting, you will have to decide whether the claims made about convenience and efficiency warrant undoing the protective rules put on the books just a few years ago.
RELATED: The Right Date?
B.4 Short Title: Suspend 5 and 6 for 2020
Moved, to suspend the changes introduced by 5 and 6 for the following year’s Hugo Award nominations (only).
Proposed by: Nicholas Whyte, Kathryn Duval, Marguerite Smith, Steven Mollmann, Ian Stockdale, Tammy Coxen, Hanne Madeleine Gates Paine, Karl-Johan Norén, and Vince Docherty
Commentary: Please see the commentary for Amendment D.7.
D.7 Short Title: Five and Five
Moved, to amend Section 3.8.1 by deleting and adding material as follows:
3.8.1: Except as provided below, the final Award ballots shall list in each category the six five eligible nominees receiving the most nominations as determined by the process described in Section 3.9.
Provided that unless this amendment is re-ratified by the 2022 Business Meeting, the changes to Section 3.8.1 shall be repealed, and
Provided that the question of re-ratification shall be automatically be placed on the agenda of the 2022 Business Meeting with any constitutional amendments awaiting ratification; and
Provided further that any business meeting prior to 2022 may move to suspend the changes introduced by 5 and 6 for the following year’s Hugo Award nominations (only).
Proposed by: Nicholas Whyte, Kathryn Duval, Marguerite Smith, Steven Mollmann, Ian Stockdale, Tammy Coxen, Hanne Madeleine Gates Paine, Karl-Johan Norén, and Vince Docherty
Commentary: “Five and Six” was one of the reforms made in 2015-16 to minimise the future effects of block voting. It already has a 2022 sunset clause and a provision that any business meeting may suspend its operation for the following year’s Hugo Awards.
After three years, we now have enough information to be clear: EPH does make a difference to deter bad actors, “Five and Six” rather less. On the other hand, having 20% more finalists does significantly increase the administrative and financial burden on each year’s Worldcon, as anyone who has been to a recent pre-Hugo reception can testify.
In addition, the burden placed by the Hugo process on diligent readers has also increased in recent years, with the addition of a new category of novels (the Lodestar) and especially of the Best Series category. In 2019 there are 31 categories in the Hugo Awards, a record. It would be a kindness to voters to reduce the required reading from six finalists per category back to five.
Although there is a 2022 sunset clause for “Five and Six”, realistically we already have enough information to repeal it now, and to make life a little easier for Hugo administrators and voters from 2020 onwards.
The Constitution normally takes two years to amend, but in this particular instance the WSFS Business Meeting also has the power to suspend Five and Six for the following year. So we can decide now to do that for 2020 (see Resolution B.3), with the constitutional amendment taking effect in 2021.
The losers will be those who had placed sixth in recent years. There is only one case of a sixth-placed finalist at nominations stage going on to win the Hugo in the last three years (the rather odd situation of Best Fan Artist in 2017, where two finalists were disqualified). On the other hand, a reduced pool of finalists increases the cachet of being among that number.
“After three years, we now have enough information to be clear: EPH does make a difference to deter bad actors, ‘Five and Six’ rather less. On the other hand, having 20% more finalists does significantly increase the administrative and financial burden on each year’s Worldcon, as anyone who has been to a recent pre-Hugo reception can testify.”
That’s it – that’s the argument — the sixth nominees are eating too many canapes at the pre-Hugo reception?
And if “Five and Six” is conceded to have some effect – at the time it was passed people already knew its impact would be “rather less” – then let’s take pleasure that the nominees eating the cheese and crackers were not picked by somebody’s slate.
Grrragh, something ate my infinity symbol in the first sentence.
If I ever formally propose it, it will be titled “To Infinity But Not Beyond”
@JJ, @OGH: so EPH is effective because the Tedster believes it’s effective? I suppose that’s a reason to be grateful he is not a member of the reality-based community, but I wonder what would happen if somebody who doesn’t need a sharp stick and a lot of yelling to make contact with reality were to try his stunt(). I’d love to see a math analysis of this (I thought there was one but don’t find/see a link), or better to run EPH against the actual 2015 ballots (although I can see ethical issues with this, not to mention reopened wounds); instead I see anecdotes about individual effects. Four decades ago, when I taught myself some statistics out of need, I might have tried to mimic the 2015 nominations and run EPH on the results myself (except that back then the only computer I had access to was about the volume of a portable typewriter); now I just don’t have the energy.
() To be clear: I’m speaking of a mass takeover, not the sort of buy-one-nomination that I think @Camestros was describing; the latter isn’t just plausible but has happened (via a third party) at least once.
@Rich Horton: regardless of questions about its effectiveness, ISTM that EPH is not “letting the terrorists win” in any sense; the system is pure math, not depending on some debatably-educated person’s assessment of some other person’s deportment, carriage, complexion, etc. AFAICT it isn’t even a burden on the ordinary traveler [administrator], as the code to execute it is available.
@Xtifr. I’m pretty sure 100% of the categories are voluntary. I thought I had made it clear that I was responding to the claim that the extra item was an imposition on the administrator; thank you for … clarifying.
if the dogs bark in the street instead of shitting in our house we can shut the windows and ignore them. That may not be victory — but as in fighting other diseases, true victory is rare (cf yesterday’s BBC story about a new strain of malaria resisting the latest drugs)
That is a misrepresentation on what EPH accomplishes. EPH rather depends on members nominating more than one object each, and does not water down your nominations. I wrote an analysis of this a while back, showing that is in everyone’s best interests to nominate every work you find worthy.
Chip Hitchcock on July 23, 2019 at 8:13 pm said:
I continue to think that while it would have had an increase at the initial stage, this would have been partially offset by the fact that it would have improved the “crowdsourcing” of eligibility because people could have looked at the semi-final lists and spotted ineligible works or cases where the same work made the list under variant names. Remember, you didn’t have to contact semi-finalists to ask if they wanted to be on the ballot. The several weeks of semi-final voting would have included weeding out ineligible works and also having potential finalists contacting the administrators to preemptively accept or decline a finalist slot. Thus a bunch of work that typically goes on in a very short time (2-3 weeks between close of nominations to announcement of finalists) could be done in public during a longer period.
As far as I can tell, many people seemed to think that (a) the administrators would have had to contact all 15 semi-finalists to get their consent to be on the semi-final list and (b) the administrators would have had to do all of the eligibility checking they do on the current finalist lists before announcing the semi-finalist lists. Neither of these things were required. The deep eligibility checking and consent to be a finalist rules continue to be only at the finalist stage (at that time five finalists per category).
But it doesn’t appear to matter. The members do not want to do their own weeding out. But they also don’t want a Strong Administrator to arbitrary throw people off the ballot. Apparently the members want magic math formulas to protect them. They seem happy enough with the current system, even though I still find it very difficult to explain to people.
I think you understand this, but your joke conflates two measures of effectiveness.
As it happened, the resistance to Vox Day’s vandalism of the Hugos demonstrated by the passage of EPH had a political effect that — while desireable — was independent from what the measure accomplishes technically.
But EPH is not a bluff, it’s a system for trying to conserve some slots in a Hugo category for non-slate finalists. If Vox Day had continued to deliver hundreds of slate voters, it is expected his picks would still have taken the majority (3? 4?) of the 6 spots (under the expanded ballot). No matter that what people would have liked is a purely mathematical system that coincidentally rules all the slate entries off the ballot — they came to understand that’s not how it’s going to look if ordinary voters pick stuff from all over the map like usual, while 200 or 400 vote the straight Dread Ilk ticket. However, if there were still non-slate choices, then No Award would not be the only option for dealing with the situation.
I wouldn’t say ‘waters down’ but yes, an EPH system could work with unlimited nominations per voter. Indeed, it might work better. However, people were uneasy about that at the time and I think people are still a bit suspicious of EPH.
If the number of nominations are capped at a small number (and I don’t see that changing) then 5/6 adds a bit extra to EPH.
Silliness. For the most part, a problem has been repaired. The excuse to rip out part of that repair is thin as tissue. And the act of pulling out this one protection is likely to act as a call to the pups, who still exist and have not changed in their attitudes.
Things are working. It would be stupid to mess them up for the excuse of the cocktails are too expensive.
@Kevin Standlee; I supported that third ballot (weeding out), believing as I still do that works representing bad actors would achieve enough public consensus to get removed, and other internecine sniping would not rise high enough to affect things.
I still do. I also still believe that a note regarding how certain works achieved their “Hugo Finalist” “marketing tool” should be incorporated into the official Hugo Awards materials, as should a strongly worded statement regarding not the letter of the law, but the intent of the law, when it comes to nominating and voting for the Hugos.
Hear, hear, @Camestros, for this:
As to the proposal itself:
I fundamentally agree with this.
It’s pretty easy to intuit: “Five and Six” successfully deterring bad actors would result in a ballot dominated by slate picks, but with lone “good” finalists in each category. EPH works totally differently and keeps any tightly-correlated clique from dominating. The results with EPH look good, and that’s fantastic — for keeping griefers out, and for keeping the finalists themselves from being too concentrated.
So what’s left, indeed, is “what is the ideal number of nominees.” And, I’m in great, great sympathy with “more is better.” But I do also take seriously that this makes voting more laborious — and in a different way than the multiplicity of categories.
Because, here’s the thing: You can definitely say “This year, I’m reading categories X, Y, Z, and not L,M,N,O, or Best Fan-Related P (Long Form)”.
But once you’ve chosen category X, then I think the fair way to judge is to consider all the works in the category. There’s an inherent unfairness if a book or a story you haven’t read is placed “last” just because you didn’t get around to that one.
Which doesn’t argue for 4 or 5 or 6 as the “right” number, but does point to the fact that more finalists can make the same amount of voting substantially harder. A lot of categories require significant time investment, and each finalist can be a pretty big slice. Saying “you’re doing this for fun” or “you don’t have to read everything” is a little beside the point. One of the things I’m doing for fun is comparing a full set of finalists; you do have to read that whole category to do that well.
I’m not a conclusive “for” nor “against” the proposal, but this particular point does seem worth noting.
Just on reading load. As others have pointed out Best Series is bigger by itself than the extra load generated by having 6 finalists. Which takes me to my other observation — I just don’t bother with Best Series! A lingering after affect of the puppy years was the feeling that it was vital to nominate in every category and to read and vote in every category. Remembering that you really don’t have to do that is important, which is why I don’t engage with the Retro Hugos (I read about them and I’m excited to see who wins but they aren’t on my homework list).
Kudos to those who do the whole thing but I’m not one of them 🙂
that’s why, throughout the history of the award, categories have had different total ballots cast numbers associated with them.
If every voter HAD to vote in each and every category, the ballot totals would be the same across the board, AND, we’d have far fewer categories these days.
I give myself permission to skim or just not finish a Hugo finalist I really don’t like. And this year, I didn’t bother with two of the series, because I knew from previous experience that I just don’t like them and am mystified why they are popular. I also ditched a Retro Hugo finalist after a brief reskim of the novel to make sure that I still dislike it and I did not watch The Good Place again, because I couldn’t even finish a single episode last year. But in general, I at least take a serious look at every finalist in a category I intend to vote in (with some dramatic presentations, availability is a problem) and short fiction I read in full, even if I don’t like it. And in one case, I did like the story a lot, once I forced myself to read it, it just started very slowly.
If I had only read four finalists out of a category, and one of them was far and away better than anything else I had read that year, I’d have no problem voting for it without going back and reading the remaining material.
I figure Hugos are to recognize excellent science fiction, and I can do that without exhaustively ruling out all other contenders.
The very fact that there is a short list means that one round of winnowing has already taken place, and it was done by people who did not read all possible novels/novellas/short stories/ etc. If it is fair to make a judgement in the nominating round based on incomplete knowledge, it can’t be unfair to do so in the last round.
You used the word “last”. I don’t like that term, because the work with the fewest votes isn’t “last” in any sense — it is in 6th place. Five works were better, and dozens/hundreds weren’t as good.
While I respect the effort of those that do go ahead and try to read all nominees, my own sense of what I as a voter should put into the process doesn’t require me to do so.
bill: I figure Hugos are to recognize excellent science fiction, and I can do that without exhaustively ruling out all other contenders… You used the word “last”. I don’t like that term, because the work with the fewest votes isn’t “last” in any sense — it is in 6th place. Five works were better, and dozens/hundreds weren’t as good.
I used to feel as Standback does, that Hugo voters needed to read everything in a category in order to be able to properly and fairly vote on it.
But over time, I have come to share your opinion on this. Though I do make an attempt to read everything, it’s okay for me to not finish/read everything in a category, if something really has no appeal for me. It’s okay for me to just vote for things I loved. It’s okay for me to place the works I hated under No Award.
The Hugo Awards are about a convergence of consensus: in other words, there’s no way for such a large and diverse group of people to come to a consensus on which are the best works, in what order of preference. All that’s possible is to get a general convergence of opinions. In any given year, I only think about 1/3 of what’s on the Hugo ballot deserves to be there. But other people think those other things deserve to be there, and their opinions are important, too.
And people just voting on some of the things, the things that they liked, contribute to achieving that convergence. It’s okay for people to not read/watch/listen to everything in a category and still vote for the things they enjoyed.
I haven’t been voting in categories where I’m unfamiliar with any nominees. But I figure I can vote for the ones I’ve read. AFAIK, that’s the way it’s always been.
Yup, I agree with these last comments about not needing to be “completest” to vote in a category. If for whatever reason I don’t read/view/listen to something, then I simply leave it off rather than ranking it. If I’ve read enough of series to get a sense of how I feel about it compared to the others, then that’s enough for me to rank it.
Rich Horton: It’s clear that this is a case of “letting the terrorists win”, in a way. We let them force us to do unsavory things, just to keep them in line. This may be unavoidable, because of the threat they post, but it’s not a good thing.
In my opinion, the unsavory thing we were forced to do was not give out Hugos in five categories in one year. I think both EPH and 5&6 make a better ballot regardless of whether there are deliberate saboteurs involved in nominations or not. But even if I didn’t feel that way, it would be better than resorting to No Award for an entire category.
Also, the changes weren’t (as the Puppies complained) just to keep them out. It was to improve the nomination process — make it harder in general to skew things against a broader consensus. Rolling the changes back now that Puppies seem to be gone would make it seem that all we did want to do was keep them out. When it actually allows a convergence like theirs on the ballot, but doesn’t let them dominate.
Dear me, I do think that’s the most heated I’ve seen @Kevin Standlee in the last 24 years. I don’t remember hearing that the 15 pre-nominees would not have been vetted*; I wonder how much of the work of vetting is simply finding the authors and confirming their willingness to be on the ballot, and whether that part could reasonably be skipped. (I’m split on this; filling the shortlist shouldn’t be hard even if one or two authors pull out, but ISTM letting them do so before the longlist is circulated would be due courtesy.)
*OTOH, I was not tightly focused on this issue even when it was hot.
Interesting: the system takes parentheses around an asterisk as both delimiters (no ‘/’ required) of a block to be italicized; this accounts for some strange fonting in recent comments, and when I pull the comment back into the editor the ‘*’ have become “em” and “/em”. Is this part of the recent upheaval, or did I just not trip it before?
oh, and the italicization does not show up in the preview. I miss a lot of things (just call me George), but that would have caught my eye.
Wordpress comments allow Markdown as well as some html.
Yeah, I think the asterisk–>italics thing is new since the last upheaval, which also resulted in the bold and italics buttons being covered up on the comment interface.
asterisk – *
italics – <em>text</em>
(“i” instead of “em” works on some but not all browsers)
bold – <strong>text</strong>
(“b” instead of “strong” works on some but not all browsers)
Ugh, they should stick to one form of markup. Mixing HTML and Markdown is bizarre.
Anyway, I don’t feel strongly about 5 & 6 and wasn’t (IIRC) a big fan of the idea initially, but given the 2022 sunset clause (unless voted to be retained, it’s out, right?), the rush seems odd to me. The reasons are completely uncompelling, for reasons mentioned upthread. Let it stand, then in 2022, ask me again. 😉
@Soon Lee: Totally! I don’t get the anti-EPH.
@Rich Horton: First past the post is hardly a perfect method of tabulating nominations. So as a different way of tabulating results, saying EPH deprived someone of a nom makes no sense to me. To me, EPH is a better way to tabulate nominations. (@Camestros Felapton & @Xtifr gave great details on what’s better about EPH.) I find calling it “unsavory” pretty bizarre. (shrug)
Interestingly, the best (FSVO “best” that is approximately “be as sure as possible that what you propose to get to the final gets to the final”) way of gaming EPH is to simply get a whole bunch of people to nominate exactly one thing, the thing you want to get on.
And while I think that’s a bit sad, I also think that if enough people nominate a single thing, that single thing should probably be passed on to the final, nominated in good faith or not.
@Ingvar: Yes, true. But you’d need a lot of people to pull that off, with a very high level of coordination. If we were facing that scope of opposition, we’d be in pretty serious trouble.
And, yes, I also agree with your conclusion: I don’t think there’s any reasonable procedural way to determine that this batch of bullet-voters is insincere and Bad, whereas that batch of bullet-voters is fannish and Good. Except let the voters vet the finalists, like 3SV proposed (which IMO is a deeply problematic solution which, in its proposed form, might have done some serious unintentional damage).
Chip Hitchcock on July 24, 2019 at 7:04 pm said:
Then you’ve not been paying very close attention. 🙂
My frustration is over people — including people who have been around longer than I have been and should know better — who seem to think that everything we do today is the way it has always ever been done going back to the dawn of time.
We didn’t used to contact potential finalists (then called “nominees” until one-vote-wonders ruined the term) to ask whether they wanted to decline being on the ballot. When we added that requirement, we added an extra administrative step that is actually pretty difficult in many cases and slows down the entire process. In many cases, we don’t know how to contact potential finalists! Am I the only person who remembers that Hugo Award finalists (nominees) found out that they made the ballot at the same time as everyone else?
Having an intermediate stage with the semi-finalists would have put Paul Cornell on the ballot in Montreal, as one of his works was nominated with three different title variations that the administrator didn’t realize were actually the same work; the sum of the nominations would have put the work on the ballot. This didn’t come to light until the “top 15” list appeared after the ceremony. When it was brought to my attention the day after the ceremony, I personally apologized to him as the WSFS division manager for Anticipation, and he was very gracious about it, recognizing that there really wasn’t anything that could be done about it. A “top 15” list that had included all three variant titles would surely have quickly led to a correction and the merger of the three works into a single semi-finalist. This is what I mean about “crowdsourcing” eligibility.
We put a whole lot of pressure on Administrators to do a lot of work in a very short period, and we expect them to be perfect. The day after nominations close, people start clamoring to know who the finalists are. Spending weeks chasing down finalists and checking eligibility gets people antsy. Having to ask finalists to respect news embargoes is hard on the finalists and leads to lots of leaks. I would be much happier with an intermediate announcement of semi-finalists without having to do either the notifications or deep checking of eligibility, if we knew that the final five (or six) would be drawn from the semi-final fifteen.
I would love it if we could see the long list between nomination closing and the finalist announcement. Even without 3-stage voting. Possibly just sent to nominators or the current year’s membership. No need for further publicity until the finalists are officially announced. And obviously no actual stats until after the ceremony.
@Standback: I would be interested to hear your details about the 3SV proposal being “a deeply problematic solution which, in its proposed form, might have done some serious unintentional damage” — and what changes would be necessary to avoid that damage. wrt bullet voting, ISTM that all bullet voting is suspect to some degree — but the response to any attempts against it would include nominating utterly random works.
@Kevin Standlee: I wonder whether any of the failure-to-remember-history comes from the people who claim that the Retro-Hugos are necessary to remember/revive history — no matter that RH for general achievements rather than single works have gotten bizarre results pretty much from the beginning (e.g., Silverberg (fan writer) and Freas (pro artist) for 1950). I suppose not having to get permission from people to put them on the longlist would be reasonable as an unaltered longlist may be published.
wrt Markdown: I share @Kendall’s dislike of merging pieces of two systems (especially when one of the systems is already a random mixture of HTML and homegrown imitations); I’ll just have to remember that asterisk-in-parends gets post-processed rather than shown in preview. (Is that how Markdown is supposed to work? ISTM that’s reverting to the pre-WYSIWYG days of roff and its successors.) wrt ‘b’ and ‘i’: I find that ‘i’ works in the live preview as well as posting, but ‘b’ does not. (I’m using Firefox.) What browsers support ‘b’ and/or don’t support ‘i’?
The Hugo Administrators do a lot of work, but the fact is that fans do have to keep an eye on their results. Think how often we have seen eligibility verification in effect crowdsourced over the past few years (Examples below.)
From 2015: Sasquan Replaces Two Ineligible Nominees on Hugo Ballot
In 2017, when Nicholas Whyte was Hugo Administrator, five ineligible finalists were removed from the ballot in the weeks after it was published. (See summary in
Measuring the Rabid Puppies Effect on the 2017 Hugo Ballot.)
@Chip: Sure 😀
I wrote about it in a previous thread, over here.
I agree we shouldn’t drop 5/6 now. It has an important role in preventing slates, both as a backup in cases where EPH fails to work, and as a corrective to cases where EPH malfunctions, favouring a fan club or single-issue campaign over a work with wider support. (And if you say we don’t have to worry about slates an more, the answer is ‘see how effective my powder is’.)
But I’m worried by people saying we don’t have to read everything (in the categories we vote in). Certainly we don’t have to read everything from beginning to end – if we find something deadly dull after a hundred pages, it’s fair to put it down. After all, nobody thinks we have to read the complete works of every fan writer, or the whole year’s run of every semiprozine. And with series, it’s fair to sample them, especially given that incomplete series are allowed anyway, which can’t be appreciated at a whole.
But it does seem to me that the Hugos are built around the idea of comparison; the voting system is designed for that, asking us to judge not just that something is good, but that it is better than the other things on offer. The Hugo packet, and the long period between announcement of the ballot and close of voting, are designed to help us actually read and compare things. And if the amount of material becomes unmanageable even to sample – as it’s threatening to do, what with the Lodestar and the series award and 5/6 happening all at once – then it’s in danger of turning into a ‘vote for what you are a fan of’ exercise, which I think fundamentally changes the nature of the awards.
It changes them back to something more like they probably were in the early 1950s, when the awards started. The sense I get from looking at fanzines and convention publications of the era is that the awards were answering the question “What’s the best thing you read last year?”; not “What was the best thing published last year?”
@bill: another reason not to want to return to the 1950s….
@Standback: I see nobody responded to that specific set of arguments (rather than some of your later discussion) in that thread; I guess I was asleep at the switch, because ISTM that your bad-case scenario is in fact a perfect-storm scenario: in the very unlikely case of the Hugos being under attack from that many directions at the same time, they have bigger problems than the nominations being hacked. (OTOH, for the Hugos to get attention from that many groups at once would suggest to me that their profile is a lot higher than it is now.) I think you also don’t account for the effect of the supermajority requirement; more than half of the voters would have to decide that the work is so bad, or its nomination so blatantly faked, that it deserves to be punted. Yes, some of the punted griefers would scream in their little circles — but if their circles weren’t so little they’d be able to get works on the shortlist legitimately. (And if they scream about the Hugos, they’ll still be raising the profile of the Hugos, to some extent among people who hear the pitch of the screams and dismiss their content.) At this point it’s academic — ISTM that the simplify-the-process types are shouting loudly enough from inside that retrying is a waste of spoons — but the issue may come up again down the road.
Bear in mind that Hornet’s Nest is one criticism amongst several substantial criticisms. Neither “Significant reduction of shortlist reading time” nor “small-ineffectual-but-loud campaigns to 3SV various finalists will make the award discussion way more toxic” should be discounted; they’re both big.
I think you misunderstand my idea of the Hornet’s Nest attack — the entire point is for an attacker to create that perfect storm. Which really isn’t hard. You just join a bunch of likely communities, and start egging them on.
I’ve personally seen an instigator coming into Israeli SF circles urging people to buy memberships and bullet-vote an Israeli worm “so Israel gets the recognition it deserves” (happily, the Israeli community has lots of people who know that’s a garbage argument). We’ve also seen a very similar issue with a particular group of indie writers for the Nebula. This isn’t hard to do, and it becomes much easier the moment that you get recognized for coming in 15th.
And, this is crucial: it doesn’t even matter if nobody actually gets 3SV’d, ever. Or if exactly the right finalists do, and only them. (Don’t forget, most of us will never know what’s been 3SV’d and what wasn’t.)
We’ve just created this huge new vulnerable flank that pretty much any concerted effort can get onto; and created an endless outrage cycle of which 3SV campaigns are or aren’t “good.”
That’s not the conversation I want around the Hugos. No thanks.
(Also posted to the discussion about 5/6 over at Scalzi’s Whatever):
One of the things I often say about both the Hugo nominating reforms we voted in starting with the 2017 Hugos (5/6 and EPH) is that they are slate-resistant, not slate-proof. Just like my wristwatch is water-resistant, not water-proof, which means I can wear it out in the rain and maybe even take it swimming, but it would be a bad idea to go diving with it, the reforms help keep a moderately-sized slate group from completely taking over the ballot, but they won’t stop a sufficiently large and well-organized group from doing so. They just make the process harder. The two reforms complement each other, providing an extra degree of slate-resistance than we would get from either reform acting alone. As such, I think we should be careful about weakening our protection by getting rid of 5/6 too quickly.
It’s true that the last organized slating attack took place in 2017. We’ve now had two years of Hugo nominations without such slates, and have enjoyed a rich and diverse group of nominees in those years. But that doesn’t mean that we no longer need the protections. Just as we shouldn’t throw away our umbrella after a couple weeks of dry weather, we should be aware that slates could return in the future if we leave ourselves vulnerable. EPH alone will give us some protection going forward, but EPH + 5/6 is stronger.
To understand why the combination is stronger, we should take a moment to look at what EPH alone can do. EPH is based on some theory about proportional voting systems. It tries to assign finalist slots in a way that is roughly proportional to the support various subgroups have in the nominating electorate – the idea is that if you had 6 finalist slots to allocate (under 5/6), and two thirds of your nominees love stories about spaceships but hate dragons, while the other third really love dragons but are tepid about spaceships, the ideal allocation of finalists on the ballot would be around four spaceship stories and two dragon stories. EPH gets closer to that ideal than the old nominating system did, but it’s not perfect.
The best academic analysis of EPH based on historic Hugo voting data is Jamison and Schneier: https://www.schneier.com/academic/paperfiles/Proportional_Voting_System.pdf . In that paper, AV is the old nominating system, and SDV-LPE is EPH (the aqua-colored line in Figures 1 and 2). With 5 finalist slots available (under the old 5/5 system), the old nominating system lets an organized slate take over all 5 slots in a category with only around 15-20% of the electorate (pretty close to what actually happened in 2015 and 2016 for several categories). With EPH in place, the slate needs more like 40-45% of the electorate to get the same result with reasonable confidence. That’s not perfect proportionality, but it helps.
How does 5/6 add to this? Well, if we have a small slate at around 10% of the electorate, they may expect to get one slot on the ballot, maybe two if they are lucky. This is similar to what we faced in 2017. EPH may give them a small diversity boost to get that one slot, as it did in at least one category in 2017. But the rest of us can just ignore that one slot if it is bad, treating it like the sacrificial piece of meat we put at the far end of the picnic table to occupy the yellowjackets and keep them away from the rest of our food. We will in most cases get five good nominees in a category, and can treat it as a normal Hugo election, similar to a pre-2015 ballot. That, I argue, is a satisfactory result given the presence of the slate in the first place. Conversely, a large slate of around 45% of the electorate may grab five slots for their nominees in spite of EPH. In that case, 5/6 acts as a safety valve, ensuring that we get at least one nominee on the ballot who wasn’t slate controlled, so we don’t have to No Award the entire category. It would take a much bigger (and better-coordinated) slate to claim all 6 slots under the combination of EPH and 5/6. For intermediate values of slate support, we will generally get one additional non-slate nominee on the ballot with 5/6 in place vs. the older 5/5 system. That is also good, as it helps ensure the quality of the winner by having more than one competitive entry (2015 and 2016 had several categories where the only non-slate nominee won by default; it would have been better if another non-slate nominee could have joined them on the ballot.)
So that’s the positive case for keeping both EPH and 5/6 to work together. Now I should note that there are two different proposals on the Business Meeting agenda addressing 5/6. There is a proposed constitutional amendment, D.7 (Five and Five), which would get rid of 5/6 permanently, but which would even if it passes need to be ratified by the 2020 Business Meeting, to first go into effect in 2021. Then there is an ordinary resolution, B.4 (Suspend 5 and 6 for 2020) which would suspend 5/6 for next year’s Hugos (only). Normally it takes two years to make any changes to the Hugo voting procedures, but both 5/6 and EPH had a special suspension provision that I wrote, which allows any Business Meeting prior to the sunset meeting in 2022 to suspend the operation of that measure for the following year. This was part of the negotiations to get support for ratification in the first place, and was something I proposed to head off a worse (from my perspective) amendment that otherwise might have passed. Part of the idea was that we were in a fluid situation against an intelligent adversary who was changing his slate strategy every year in response to what we did, and there were multiple proposals on what to do about it that might not all play well together. So I wanted to make it easy for the Business Meeting to temporarily get rid of provisions that might not fit well with other proposals, while making it easy to bring them back the following year if getting rid of them turned out to be a mistake, rather than having to wait two years for a fresh constitutional amendment. Hence the suspension amendments.
I should also note that the sunset clauses don’t mean that either EPH or 5/6 is automatically going away after 2022. Rather, they mean that the 2022 Business Meeting will have to vote on whether to make each of them permanent from 2023 on. Only if that vote fails does the related provision go away. So combining this with the existence of the suspension amendments, the only practical effect of passing amendment D.7 is that if ratified in 2020, it would avoid the need for the 2020 and 2021 Business Meetings to pass suspension resolutions for 5/6 affecting the 2021 and 2022 Hugos, respectively, and eliminate the sunset ratification vote for 5/6 in 2022. Even if D.7 passes this year, opponents of 5/6 would presumably still need to propose a suspension resolution for the 2020 Business Meeting just in case ratification of D.7 fails. So basically D.7 is a “save a little time in the 2021 and 2022 Business Meetings” proposal, plus whatever symbolic value it has for its supporters.
I will acknowledge that I do feel a certain amount of fatigue at having extra finalists on the ballot to read, especially in a year that also has a lot of retro-Hugo nominees on the ballot such as this one (but also joy, because some of those sixth finalists are really good). I would have a fair amount of sympathy for B.4 as a stand-alone resolution if the primary argument in favor was “2020 is also going to have a lot of extra reading on the ballot due to the retro-Hugos that year, and it’s unlikely that slaters will attack in 2020 specifically, so why don’t we just suspend 5/6 for 2020 only?” My concern is that supporters of D.7 would take the passage of B.4 as a signal that we don’t need 5/6 any more, period, and use it to rally support for ratification of D.7 next year. If D.7 is defeated first, I will be happy to consider B.4 on its own merits.
Pingback: Dublin 2019 WSFS Business Meeting Day 1 | File 770