paging JMS

[calimac]

In the days of Babylon 5, its proprietor, J. Michael Straczynski, was mildly infamous among Hugo administrators for an interesting habit. If more than one episode of the show received enough nominations to be a Hugo finalist, he would withdraw all but the episode he judged most popular. His reason was that he didn't want to risk losing the Hugo by competition with himself.

Several of us, kevin_standlee in particular, tried to explain to him that the Hugo final ballot doesn't work that way, but JMS was insistent.

That's why I'd like to show him this year's Dramatic Presentation Short Form result. Three episodes of Dr. Who (actually four, but two were a two-parter and treated as one), one episode of Battlestar Galactica, and three miscellaneous others. On this theory, the three Doctors should have cancelled each other out. But that's not what happened.

Dr. Who: Empty Child/Doctor Dances  112  112  114  116  148  163  247
Battlestar Galatica: Pegasus        137  138  141  154  167  184  195
Dr. Who: Dalek                       70   70   70   75   95  107
Jack-Jack Attack                     67   68   72   85   89
Dr. Who: Father's Day                64   64   65   72
Prix Victor Hugo Awards              42   43   61
Lucas Back in Anger                  34   34
No Award                             29

BG had the plurality on the first count, but with only 25%. The three Doctors between them had 44%. And watch what happens on the fifth count when one of them gets eliminated. A whopping 72% of its voters choose one of the other two over any of the three other options, including dropping out. The effect is even greater when the other one is eliminated. Now united behind their favorite episode - without a creator having to guess which one that would be - the Whovians take the majority, and thus the Hugo.

Whatever you think of the relative merits of the two shows (I'd be in the BG camp myself), it's clear that BG could not have pulled ahead against all those Dr. Who episodes, and once they were whittled down to one, it still had the majority. Hugo finalists need not worry about competing with themselves.

Comments

[voidampersand.livejournal.com]

I'm with JMS. Preferential voting is a good idea, but there is no guarantee that someone who particularly likes one B5 episode is going to highly rate the other nominated B5 episodes. Of course, in practice, most B5 fans would give their top votes for all the B5 episodes, but there is at least a chance in theory that vote splitting could affect the outcome. You can see how the three Dr. Who episodes are interspersed with other shows. I think this year Dr. Who was helped by the fact that two of the entries, while wonderful, didn't have a prayer of winning. I'm glad at least that Sorensen came in ahead of No Award.

[kevin-standlee.livejournal.com]

And I think that anyone who withdraws works due to perceived vote-splitting is a fool. Withdrawing additional works makes you less likely to win.

Look at it this way: Assume there are no nomination ties, and therefore only five nominees. If vote-splitting lessens your chances of winning, then if your works had all five nominations, you'd have zero chance of winning. I know you're going to brush that aside as a trivial case, so let's move on. If you have 4/5 nominations, then anyone who believes in vote-splitting must think you have less of a chance of winning than if you only have 1/5 nominations. That's just silly, and is symptomatic of people whose only understanding of elections is first-past-the-post, where vote-splitting can ruin you.

[voidampersand.livejournal.com]

What if your fans are perverse? They really like your show, but they feel it would be unfair to vote for all your nominated episodes when the other nominees are good too. Let's say this is the 4/5 case and that 70% of the voters are your fans and prefer your work. Each of your works gets 17.5% of the #1 votes, and 7.5% of the #2 votes. Your lone competitor gets 30% of the #1 votes and 70% of the #2 votes. It wins on the third ballot.

Please remember that I was saying that there is at least a chance in theory that vote splitting could affect the outcome on a preferential ballot. You're saying that's just silly, but many things that are silly in practice are still possible in theory. Also, one could argue that voters would never be so perverse as to vote against their own interests, but it's been proven many times that they often do.

[kevin-standlee.livejournal.com]

What if your fans are perverse? They really like your show, but they feel it would be unfair to vote for all your nominated episodes when the other nominees are good too. Let's say this is the 4/5 case and that 70% of the voters are your fans and prefer your work. Each of your works gets 17.5% of the #1 votes, and 7.5% of the #2 votes. Your lone competitor gets 30% of the #1 votes and 70% of the #2 votes. It wins on the third ballot.

Then your fans are idiots. Do you really want people that stupid voting for you? I also think this example is even less likely that one show having all five nominations. (And there was one historical case where all five nominees were from the same series.)

Please remember that I was saying that there is at least a chance in theory...

It's a fannish fallacy to assume that because perverse behavior is theoretically possible, it is highly likely to happen. This is not true. Oh, it could happen, but it's highly unlikely.

I'm reminded of the people who argued against L.A.con IV's "taster" memberships because they were convinced that single-day members would come in, vote in site selection, and cash in their taster. My response was, "Who in their right mind would do that?" If even one day member to later cashed his/her "taster" tried to vote in site selection, I'd be astounded.

Not all possible outcomes happen. Usually only the likely ones do. I try to concentrate on cases that are most likely to happen; the others are only of academic interest. Sometimes, they're even intellectually amusing, such as prospects of a large proportion of the 6,000 or so attending members of a Worldcon wanting to attend the Business Meeting and its chairman having to go to Programming and say, "Well, I know you really need the Arena for Hugo Award rehearsals and setup, but we've got 4,000 fans here demanding to have a debate all at once, so you'll have to get out of our way."

[voidampersand.livejournal.com]

Then your fans are idiots.

Not that I actually have any fans. But if I did, they would be idiots...

Do you really want people that stupid voting for you?

Yes.

[rwl.livejournal.com]

I'm in favor of not even allowing nominated works or people to be withdrawn, unless there is a technicality which results in their ineligibility.

[kevin-standlee.livejournal.com]

Well, you know why the current rule (effectively the Judy-Lynn Del Rey Amendment) is there. It wasn't meant to allow for "strategic withdrawal." There's another withdrawal rule that talks about a version of a work "not being representative," but again, strategic withdrawal was not contemplated as part of that. It does go to show that rules with good intentions can be misused.