So, if you are one of the lucky people who doesn’t follow the twitter-sphere, here’s the story so far. The Hugo Awards, SFFs most influential honors, announced their list of nominees on Saturday. A group of writers on the conservative side of the political spectrum had campaigned for a slate of nominees, the Sad Puppies, and were very successful in getting their picks onto the ballot, completely taking over some of the categories. Many people were very upset by this, which was probably the point. I’ve been participating in multiple simultaneous discussions, so I thought I’d write up some thoughts here where I can conveniently link them.

Why This Happened

Various people have accused the SPs of cheating, which is almost certainly not the case. On the SP side, people have been crowing that this shows they’re really in the majority, which is also almost certainly not the case. Given the way the Hugo nomination process is structured, organized slate voting is a dominant strategy, and a minority being able to completely dominate the ballot should be an expected result.

If you’re not familiar with it, the process is pretty simple. In each of the various categories (Best Novel, Best Novella, and so on) each voter can nominate up to five works. The votes are all added up, and the top five in each category become the nominees, to be voted on at WorldCon to determine the winner. The SP strategy was to pick five works for each category and encourage their voters to choose those five.

Let’s consider a hypothetical election between Green and Purple voters. There are 800 Greens in the voting pool, and 200 Purples. The Greens mostly prefer Green works, of which there are, say, 10 in serious contention — we’ll call those G1, G2, etc. The Purples similarly prefer Purple works, P1, P2, etc.

The Greens have no organization. Each Green picks the five works out of the ten that he or she personally likes best. Assuming each work has its fans, this will lead to a vote distribution that is reasonably even — say 95 for G3, 93 for G5, 89 for G8, down to 56 for G1.

[EDIT — It has been pointed out that I suck at math here. The actual vote totals would be closer to 400 for each (since each person gets five votes), so in order to make this example work, we need more Green works to spread across, and similarly for Purple. The numbers are just illustrative anyway, although this actually suggests an important point — the SPs had less success in the Best Novel category, where there were some clear front-runners, then in a category like Best Short Story where the vote was more widely scattered.]

If the Purples voted similarly, they would get a similar distribution: 34 for P2, 30 for P10, and so on. In this case, the ballot would be all Green, since the fifth-most popular Green work is more popular than all the Purples.

Instead, Purple Leader says, “Hey, lets all vote for P1, P2, P3, P4, and P5.” The Purples all go along with this. So those five works receive 200 votes each, and the others zero. Now the final ballot will be entirely Purple! The minority, by being more organized, runs the table. The Purples don’t cheat; neither have they suddenly become a majority. They simply have a more effective strategy, considered solely in terms of getting Purple on the ballot.

Why This Is A Problem

Back in the real world, why should we be concerned about this? John Scalzi suggests we should not be. The final Hugo voting includes a “No Award” option, for almost exactly this reason, and the voters can make use of it if they are sufficiently pissed off. Justin Landon broke down last year’s Hugo voting, in which there was a similar, if less successful, SP campaign. The upshot is that the SP candidates were completely defeated in the final voting, which uses a very different voting system — ranking, with instant runoffs, mean there’s no vote-splitting “spoiler” effect. Again, not an unexpected result, and I predict we’ll see something similar this year. (No Award will almost certainly win a few categories.)

But, for me, that’s not good enough. It’s a bit like saying it’s okay someone came over and kicked down your sandcastle, because they weren’t able to build their own. The problem is not that a bunch of conservative-leaning writers got on the ballot; as Scalzi says, that’s not a big deal. The problem is that it is now blindingly obvious that “slate” voting, if widely used, will dominate the nominations.

Suppose next year, when Sad Puppies IV is announced, a liberal-leaning writer counter-organizes a Happy Puppies slate. He or she would probably get a lot of support. Given the composition of the voting pool, Happy Puppies would probably win; let’s say they shut out the Sad Puppies completely. Is that better? Now we have an award in which the organizers of the two slates decide who gets to be on the ballot, independent of what works people really think are worthy. In the above example, note who loses out — Purple works 6 through 10, who didn’t get picked for the slate and were thus completely removed from consideration. In real life, too, there’s crossover between the Green and Purple sides, but slate voting eliminates that entirely. Someone who might attract Green votes, but gets picked for the Purple slate, is going to be screwed.

It’s tricky to talk about the “true spirit of the Hugo awards”, because they mean different things to different people. But I like to think that the scenario where each person chooses the works that they personally found to be best is closest to the ideal. If that scenario is unachievable (and it is) then we can at least try to get as close as possible.

Can We Fix It?

Voting systems are hard. They are the realm of unintended consequences and unforeseen strategies. A whole branch of academic game theory studies them, and there is no system that is clearly best in all circumstances, even give ideal voters; once you add human foibles into the mix, things get even more complicated. So beware of anyone saying, “Oh, it’s easy, we just have to X.”

Here are some things that will probably not work:

Just ban slate voting and campaigning. Impossible to enforce. You can never prove that a voter didn’t independently happen to choose the same works that are on the slate, and it would put the judging authorities in an impossible position, ripe with possibilities for abuse.

Expand the electorate. A good idea in general, but not a silver bullet here. The Hugo electorate is small (~2,000 voters, members of WorldCon) but that isn’t actually the problem. The SPs didn’t win by signing up enough of their people to be a majority (usually what you worry about with small voting pools) they won with a minority by being better organized.

What we need, ideally, is a change in the voting rules that aligns the result we want (everyone picking what they think is best) with the optimal strategy, as much as is possible. We also need the rules to not be excessively complicated and cumbersome, or else no one will vote.

A Modest Proposal: Anti-Votes

I’ve heard several proposals that might work. The most obvious is to change the number of votes allowed so it’s smaller than the number of final nominations, making it much harder to coordinate a slate takeover. (That is, everyone gets to pick 3, and the final 6 top are chosen, or similar.) This might help, but I don’t know that it gets to the root of the problem — it would be harder to coordinate a slate, but far from impossible. I’m interested in thinking about it, though! Here is my suggestion, which seemed a bit odd to me at first, but which I think gets closer to addressing the underlying issue.

The ballot stays as is, except that each category gets a section for five anti-votes. Each voter can both vote and anti-vote, for a total of ten choices. Anti-votes are subtracted in the final voting tally, and the top five results get on the ballot, even if their vote totals are negative or zero.

Why would this help? Because it turns the structural advantage of an organized slate into a disadvantage. Imagine what would happen to the Sad Puppies under this system. People who don’t like them, or who don’t approve of slate voting, can anti-vote their whole slate, just as easily as their supporters can vote for it. The more widely known the slate is, the more anti-votes it will attract.

Of course, the SPs would get anti-votes too, and could easily publish an anti-vote slate. But as long as their opposition isn’t pushing an organized slate of their own, the anti-votes will be split among many possible candidates, just as the Green votes are above. Organization and campaigning would become a liability instead of an asset.

With any voting system, we have to think about possible consequences. Who would be unfairly hurt by this system? The obvious answer is “people who the slate voters very much dislike”. It’s quite possible that, for example, John Scalzi would attract a disproportionate share of the anti-votes from the SPs and their allies, thus crippling his chances compared to a writer who has a similar position but is not as politically active. It might have a chilling effect, where writers think harder about taking certain political positions, lest they get on the anti-vote list of an opposing group.

These are real problems, and it may be that they outweigh the benefits. But it’s at least worth thinking about. As it stands, the Hugo nomination process is badly broken, and unless something is done the system is not going to produce useful results, only political football.