This post is for those who have heard about the controversy over slate voting and the Hugo Awards, but don’t know exactly what that means in the context of the Hugos. I’m going to simplify for the sake of (relative) brevity. ((Disclosure: Although I’ve tried to be accurate, I am not neutral or objective, and I generally disagree with the Puppies on most things.)) . ((For a more detailed summary of events, see Freeping the Hugo Awards. ))
BACKGROUND: HOW HUGO AWARD VOTING WORKS
For any readers who don’t know, the Hugo Awards are an annual award given out for science fiction and fantasy works. The Hugos Awards are voted on in two rounds. The voters are members of Worldcon (anyone with $40 to spare can be a voting member).
In the first round, voters can write up to five works within each category on their voting ballot (categories include “best novel,” “best short story,” “best graphic story,” and so on).
Hundreds of works and creators are written in during the first round by over a thousand voters (iirc), but only five in each category – the five most popular among all voters, using a “first past the post” vote-counting method – get to be “nominees.” This is what people are referring to when they say a work or creator is “Hugo-nominated.”
In the second round, Hugo voters choose from among the five nominees per category, and one winner per category is chosen, using Instant Runoff Voting. (Voters can also vote “no award,” and if no award “wins” a category, then no award is given in that category that year.)
HOW SLATE VOTING WORKS
The current controversy over slate voting is specifically related to the first round of voting. Because the majority of Hugo voters spread their first-round votes among hundreds of different works per category, it takes a relatively small number of votes (40-60, iirc) for a work to be nominated for a Hugo.
Therefore, if a minority of 100 or so voters organizes as a bloc and votes in unison (or near-unison) for the same five works in each category, they alone will determine who gets nominated for a Hugo, while the majority of voters will have no effect on who gets nominated. This form of collective organizing is called “slate voting” or “bloc voting.”
There were two known slates this year, the Rabid Puppy slate and the Sad Puppy slate. The two slates overlapped significantly, and I will refer to them collectively as “Puppies.” In multiple Hugo categories, the Puppies controlled which five works were nominated, locking out the majority of Hugo voters from having an effect on the outcome. In other categories, the Puppies did not control all five nomination slots, but they still had a much greater effect than they would have if they had voted as individuals rather than collectively organizing.
WHY I DISLIKE SLATE VOTING
(This section is less factual and more about my personal opinions.)
Slate voting is antidemocratic, since it is a way for a minority of Hugo voters to control the outcome of the Hugo nomination process.
Slate voting also breaks the longstanding understanding that Hugo voters are supposed to vote based on quality – that is, they’re supposed to vote for the works they as an individual consider the most outstanding work of the year. ((“Doing anything except nominating the works you personally liked best is cheating in my book.” — science fiction author Connie Willis.)) That hundreds of Puppy voters all individually decided to choose almost exactly the same 27 or so works as the most outstanding works of the year, and by a massive coincidence their individual favorite choices matched the works listed on the slates chosen by their leaders, is not a credible claim.
Finally, although the pre-Puppy status quo was not perfect, the Puppy’s slate tactics are exceptionally prone to nepotism and corruption, because the final decisions of which works went on Puppy slates were made by just a few leaders, who operated without any transparency.
As a result, some Hugo nominees this year seem to have been nominated for being pals with Brad Torgensen, who ran the Sad Puppies slate, rather than for producing work that is outstanding either in quality or popularity. And the Rabid Puppy slate strongly favored a previously-obscure Finnish publisher, a company owned by… the organizer of the Rabid Puppy slate.
Slate voting leads to political parties. “What institutional slate voting gets you, no matter how well-intentioned or how much it is aligned with your own views, is political parties. Nothing can get onto the ballot unless it’s part of a slate, so the people who run the slates become the kingmakers; any author who wants any chance at an award has to get in with one of them.”
THREE POPULAR PROPOSALS TO REDUCE THE INFLUENCE OF SLATE VOTING
Many have suggested that all that’s needed to reduce the influence of Slate voting is more voters, that is, for a larger number of people to vote in both rounds of Hugo voting. However, since Slate Voting is a strategy that mathematically allows a collectively organized minority to overcome the preferences of a disorganized majority, I don’t have much confidence in this proposal. (Although it is a nice idea for other reasons.)
Another proposal is the 4/6 proposal, in which individual Hugo voters can only nominate four works per category, and there will be six nominees per category. In this case, rather than a successful slate controlling 100% of nominees in each category, it will only control 66% of nominees in each category. If there are two slates, then the most successful slate will control 66% of nominees, while the next most successful slate will control the remaining 33% of slots. This seems like an insufficient solution, to me.
The proposal I favor is “Least Popular Elimination,” in which voters could still nominate up to five works per category, but the votes are counted in a way that mathematically favors works that appear on the broadest number of voters’ ballots while diluting (but not completely eliminating) the power of slate voting. A detailed explanation of “Least Popular Elimination” voting is available here. While LPE voting is not as intuitive as the other two proposals, I believe it would be more effective. ((This post began life as a comment on Feminist Critics.))
Pingback: The Canine Billion Names of Dog 5/17 | File 770
Why don’t you get each voter to nominate five voting systems, then from all the voting systems proposed chose the five most popular, and then…
Why are they called “puppies”? Is that just a cute name they chose for themselves? I assumed it was a reference to… something.
Tamme: They named themselves “Sad Puppies” after this video, or so I’ve been told.
Daran: Good idea! But first you should write up that proposal so the Worldcon meeting can vote on it.
Ah, but then someone might come up with a slate of voting systems.
On a more serious note, in the FAQ in the post explaining your prefered voting option it says:
In fact the Theorem doesn’t say anything about the possibility of gaming. What it says is that all voting systems which generate a ranking of candidates based upon individual rankings. (These are not the only voting systems possible) are in some sense “flawed” even if all voters are honest. “Flawed” of course is a value judgement, not a mathematical property. What the theorem says is than any such voting system must have at least one from a specific list of properties which most people would regard as flaws, or at least somewhat undesirable.
I’ve just discovered two other theorems, applicable respectively to the first and second rounds of the Hugo voting process, and which more closely fit the description of the “theorem” given in the FAQ.
This really is a fascinating topic in mathematics.
In any case, kudos to the stock photographer who had the foresight to stage a puppy sniffing a trophy.
I can think of a much more appropriate pose for the puppy, which in the interest of good taste I will not describe here.