2015 Hugo Award Trophy, designed by Matthew Dockrey
In light of this year's Hugo Award shortlist and the attendant furor therein, the discussion on how to "repair" the voting system has once again arisen. Some people are incensed that things haven't been fixed already, because they heard something passed last year, and they either do not know or do not care that all changes to WSFS rules require passage in two consecutive years. (I had someone contact me last year asking for the "Special Secret Emergency SMOF rules for changing things in one year." He was very disappointed to hear that there are no such rules. In fact, our rules are deliberately designed to derail those who would insist that we must Change Everything Right Now because it's an Emergency.) In any event, two proposals are up for ratification at the 2016 WSFS Business Meeting to modify the Hugo Awards nomination process: E Pluribus Hugo (replace the first-five-nominees-past-the-post system currently in use with a system that uses math to reduce the ability of closely correlated ballots to dominate the process so that less than 20% of the electorate can collect nearly 100% of the finalist slots) and 4/6 (limit members to four nominations per category and increase the number of shortlist slots in each category to six). The two proposals are not directly in opposition: either, neither, or both of them could be ratified at this year's Business Meeting. Those proposals that are ratified this year will first take effect with the 2017 Hugo Awards (generally for works first published in 2016) administered by Worldcon 75 in Helsinki.
I do not wish to rehash the arguments for and against either EPH or 4/6 here. However, I would like to set out a plan that I've been thinking about (and have talked about) since this series of bad actors imposed their minority opinion on the majority of the members of WSFS last year. It's a completely different change than either EPH or 4/6, and actually leaves the current nominating and final ballot phases of the Hugo Awards unchanged, but instead inserts an additional stage into the selection process that gives the membership of the current Worldcon an opportunity to weigh in on whether they think any particular finalist deserves to be on the ballot without having to resort to the rather blunt instrument of voting No Award.
The key points of 3-Stage Voting are:
- Nominating Stage Does Not Change: Nominate up to five works per category per member, with members of the previous, current, and following year's Worldcons all eligible to nominate.
- New Semi-Final Round: The top 15 nominees in each category are put up to a yes/no vote on each nominee in a new Semi-Final round, with only the current Worldcon's members eligible to vote.
- Final Ballot Voting Does Not Change: The five semi-finalists from the first round with the most nominations that are not eliminated in the second round (and who don't decline or are found to be ineligible) go on to the final ballot, which is voted by the same Instant Runoff Voting system we have used the 1960s.
Now let's unpack the details of how this would work, because there are a lot of them, and they interact in ways that you might not expect and that I think actually improve the overall process in many ways.
Nominating Stage Does Not Change: Under this proposal, the existing nominating ballot would not change. Members of the previous, current, and following Worldcon would continue to nominate up to five unweighted nominations in each category. However, the time for casting nominating ballots would have to be shortened due to adding an additional round to the system. Currently, you have to be a member of at least one of the three Worldcons in question by the end of January of the current year. I would move that deadline back to the end of December of the previous year (one month earlier) and open nominations on January 1. With the union of three Worldcons' membership eligible to vote, I do not think we need that extra month for people to join. Nominations would be open for two months, closing at the end of February. That should be sufficient time to evaluate the works you want to nominate that (generally) were published in the previous calendar year, even things published in late December.
New Semi-Final Round: The top 15 nominees in each category would be declared "Semi-Finalists" and announced relatively soon after nominations close: I think two weeks should be sufficient. (ETA: This potentially includes more than 15 due to ties; in addition, we might want to make it "Top 15 with ties and anything that gets at least 5%," that being the current post-ceremony report criteria. Everywhere I mention "Top 15" below, assume that there might be more than 15 in the "Top 15.")
These are not the Hugo Award Finalists. This is a very important point, because under current rules, the Hugo Award Administrators are required to attempt to contact potential finalists and give them the opportunity to decline nomination. (Note that "opportunity to decline" is not the same thing as "must obtain positive acceptance;" the latter could potentially delay the final ballot forever, which is why the current rule doesn't require it, and the default state is "accept." This is known as the "Judy-Lynn del Rey Rule" due to the circumstances under which WSFS added it.) In addition, I would not expect the Administrators to do in-depth eligibility checking of all 240 semi-finalists (based on the current number of categories). I'd expect them to do basic checking and to disqualify anything they can obviously catch; however, the Semi-Final Round includes time for catching nominees that aren't eligible, as I explain below.
By the end of March (preferably sooner), the Worldcon would release the Semi-Final Ballot. This ballot would contain each of the semi-finalists in each category. The semi-finalists would not be ordered in any way that would indicate how many nominations each received. Each semi-finalist would have a this question after it: [Edited from feedback]
Do you think this nominee should be allowed to be on the on the Final Hugo Award Ballot?
__ Abstain/No Opinion (Does not count for or against)
Each semi-finalist is a separate question. Your vote on any semi-finalist is unrelated to your vote on any other semi-finalist, and your choices are not ranked in any way. Voting would be open to voting WSFS members (supporting, attending, and whichever others qualify) of the current Worldcon only as of the date of the release of the Semi-Finalist "Long List." ETA: But see comments below; the actual cutoff date is debatable.
ETA: "No Opinion" is an abstention. It has exactly the same effect as if you didn't vote at all. But from feedback I have learned that people would be happier being able to explicitly say "I don't care one way or another," so as long as they are told that their vote for No Opinion counts neither for nor against a semi-finalist, I'm okay with including the option.
Further ETA: There is one case in which an explicit Abstain vote would matter, and that is if you imposed a quorum requirement. Some people appear to be concerned that most members would ignore the semi-final round, and griefers would all turn up and down-vote everything except their own stuff. Unless you're convinced that the griefers represent an absolute majority of the membership (in which case you've lost already, so you might as well give up), the way you deal with this is to require that at least N ballots be cast on a given disqualification vote for that vote to be able to disqualify the semi-finalist. N can be an absolute number or a formula, such as "At least N% of the number of nominating ballots cast in that category." In such a case, explicit abstentions (as opposed to leaving the choice blank or not voting at all) count toward the quorum, but not for or against the semi-finalist.
For example, assume the quorum is 500. On a given semi-finalist, most people ignore it and don't vote at all. Griefers pile on and 400 vote No, while only 50 people bother to vote Yes, and nobody marks Abstain. The semi-finalist is not eliminated, because not enough members voted to make it a legal election on that issue. Conversely, if the vote was 50 No and 400 Yes, with nobody else voting, the semi-finalist is still not eliminated, for exactly the same reason — not because a majority voted Yes, but because there weren't enough votes cast at all. The default state of a semi-finalist is "confirmed." It would take a quorum and a negative majority to disqualify a semi-finalist.
The current Worldcon would need to send that Semi-Final Ballot out fairly quickly. Even though nearly all voting now happens electronically, I still would rather not disenfranchise those members who cannot or will not vote electronically.
Semi-Final voting would be open until the end of April. During this time (about six weeks), the Hugo Award Administrators would contact the Semi-Finalists and ask if they would accept a slot on the Final Ballot should they qualify for it. This should be easier than the current process, because if they can't contact a given semi-finalist, they can put out a public request for assistance in finding the semi-finalist. In addition, the Administrators would be confirming the eligibility of each semi-finalist. Again, this would almost certainly be made easier due to the public nature of the Long List. I think it quite likely that some of the 240 semi-finalists would be ineligible, and people would point this out during this six-week semi-final round.
Another thing Administrators would be doing during the Semi-Final round is checking with semi-finalists as to whether they would consent to include their works in a Hugo Award Voter Packet. (This assumes that the current Worldcon intends to do such a packet. Packets are a lot of work and are not guaranteed, and are dependent upon the goodwill of the rights-holders. I do not criticize any nominee who declines to participate in the Packet, and personally I'm not certain how much longer we'll be seeing such Packets as the pool of potential voters continues to grow.)
At the end of the Semi-Final round, the Administrators would tally the yes/no votes. Any semi-finalist with more no votes than yes votes would be disqualified from the final ballot. The exact percentage is up for debate: you might want 55%, 60%, or two-thirds. You might want a minimum number of votes cast total in order for a disqualification to be valid. I'm open to discussion on this, but the principle is that the members of the current Worldcon can act en masse in a democratic manner to disqualify nominees that they do not think deserve to be on the final ballot. Also out of the running would be any semi-finalist found to be ineligible, or who declined nomination to the Final Ballot.
If a semi-finalist declines or is ruled ineligible, I would not backfill semi-finalists from positions below the top 15. The Finalists will be drawn from the pool of 15 announced semi-finalists, and that's it.
Final Ballot Voting Does Not Change: Relatively shortly after the Semi-Final round (I think it should take about a week because so many pieces of work were pushed into the Semi-Final round), the Administrators would announce the Final Ballot. The Hugo Award Finalists would be the five semi-finalists (possibly more in case of ties, as currently) with the most nominating votes (from the Nominating Round) that were not disqualified for any of the reasons in the Semi-Final Round. I reckon this should happen no later than May 15, and voting would open soon thereafter.
An added side benefit of this system is that Hugo Award Finalists would not be able to leak their status. They wouldn't know themselves whether they made the final ballot until the finalist announcement. This would put an end to the persistent and annoying problem of finalists leaking their shortlist nods, which has been an issue ever since the del Rey Rule went into effect.
The Final Ballot announcement would not include the number of yes/no votes each semi-finalist received, nor the number of nominations each finalist received in the Nominating Round, nor a statement regarding which semi-finalists were disqualified by the vote of the members. This information would be included in the post-Hugo-Ceremony information already released.
Whether to announce semi-finalists declining or being ruled ineligible would be up to the semi-finalists and the Administrators; however, I personally think that eligibility announcements are wise, in order to close off speculation. Decisions to decline are traditionally left up to the nominee, and if that nominee chose not to publicize the choice to decline, I would respect that decision and include it in the post-ceremony figures.
This system leaves open the possibility that more than ten of the fifteen semi-finalists will be eliminated. I would propose that if fewer than three nominees are left after the Semi-Final Round, the category be dropped for that year. Hugo Award Administrators already have the right, on their own subjective assessment and without debate or appeal, to cancel a category if in their opinion there is insufficient interest in the category to justify presenting it. This has generally only applied to some Special Categories; however, it does show that there is precedent for dropping a category before the Final Ballot. (I was, along with kalimac, part of the Hugo Administration Subcommittee that ruled that the Best Translator category that the 1993 Worldcon put on the Nominating Ballot had drawn insufficient interest to justify including on the Final Ballot.) Also, several of the Retrospective Hugo Award categories were dropped this year (and were in the past) due to lack of nominations. (Part of that is because the rules require including all current categories, including some that in practice didn't exist as works 50/75/100 years ago.) I choose three as the minimum because the existing 5% rule (also up for removal this year) sets three finalists as a floor for the minimum number in a category. I'm not wedded to this number, and would listen to debate on what the best minimum number might be.
I would not permit finalists to withdraw after the Final Ballot was announced. They had at least six weeks to think about it during the Semi-Final Round, and that should be plenty of time. If someone announced after the Final Ballot came out that they didn't want people to vote for them, or that they would decline the trophy if they won, that's up to them, but the Administrators should in my opinion under a 3-Stage Voting system ignore any such declarations.
The Final Ballot voting system remains unchanged. We would continue to use Instant Runoff Voting and No Award would continue to be a candidate, as we currently do. It is still possible that a nominee could get through the Semi-Final Round onto the final ballot and finish below No Award. It is even possible that No Award might win outright as it has done ten times in the history of the Hugo Awards (five of those times in 2015); however, I think it a lot less likely to happen with the members of the current Worldcon being given the opportunity to weigh in at the Semi-Final stage.
There would be new information included in the post-Ceremony report, including the yes/no Semi-Final votes on each nominee. Because of the rules about the Hugo statistics, there might be more nominees listed than just the top 15.
Benefits of 3-Stage Voting:
- Relatively easy to explain (vote yes/no on each semi-finalist)
- Final Ballot will reflect the majority opinion of the members of the current Worldcon that the finalists deserve to be on the ballot
- Administrators will have the help of the public in checking eligibility and contacting potential finalists
- Finalists will be unable to leak results prior to the announcement of the Final Ballot
- Relatively small minorities will not be able to dominate the shortlist
- No Award is unlikely to win a category
- Does not require a Strong Administrator to subjectively decide whether a nominee is "legitimate" because the decision can be submitted to the membership as a whole
Drawbacks of 3-Stage Voting:
- Increased workload for the Hugo Award Administrator (3x Semi-Finalists versus today's Finalists)
- Increased expense and workload for Current Worldcon (an extra ballot mailing and work to administer voting)
- Potential for campaigning against Semi-Finalists
- Need to explain to members that the yes/no decision is about whether the semi-finalist is worthy of the honor, not just "Did I personally like it."
- Potential for categories being dropped due to >12 semi-finalists being disqualified
- Shorter period for Final Ballot voting, and a tighter schedule overall
I think the benefits of 3-Stage Voting system outweigh the disadvantages. I also think that it's more practical to conduct a 3-Stage election now that all but a handful of members are voting electronically.
Will I propose such a system this year? I do not currently expect to do so. There is already a crowded agenda with the Business Passed On from last year, including contentious and complicated proposals such as E Pluribus Hugo. Introducing this proposal would muddy the waters by leading people to vote against proposals that may improve the current system and would not require an additional round of voting. Should anything get ratified this year, we really need to let it run for a year or three just to see how it works in practice. And furthermore, I'm Chair of the 2017 WSFS Business Meeting in Helsinki. Were this proposal to be introduced and pass this year, I would be honor-bound to recuse myself from the Chair again next year when it came up for a ratification vote (as I did with Popular Ratification in Spokane), and I'd prefer to not have to do so. So at the moment, the earliest I could see introducing 3-Stage Voting would be 2018, and I'm not even sure about that.
I reserve the right to modify this proposal or change my mind about introducing it based on developments as time goes by. Changing my mind is not "flip-flopping," but is sensibly re-evaluating a situation in light of new information.