Since Alison asked: two answers on counting the Hugos - EPH and Best Dramatic Presentation
The latest installment of the Octothorpe podcast has a couple of points on counting Hugo nominations raised by Alison Scott, which I'd like to address here.
First off, at 18:53, Alison expresses the concern that under the counting system (known as "EPH, short for "E Pluribus Hugo"), "if I nominate anything else, it reduces the chance that the thing I love gets on the ballot. And that seems to me to be the big downside of EPH." The podcast is kind enough to link to the explanation of EPH that I wrote in 2017.
It is of course perfectly true that under the current counting system, if you love one thing, and you nominate other things as well, you reduce the chance that your vote will help the one thing you love getting on the ballot. But this was equally true of the old tallying system. It's the principle of monotonicity, as Kenneth Arrow put it in his famous Theorem. Candidate A getting more votes should not lead to her getting a worse result, and Candidate B getting fewer votes should not lead to his getting a better result. If you give an extra vote to A, you are inevitably hurting B's chances, even if you love B more. So it's not uniquely a problem with EPH, but with any system where your vote can go to several candidates simultaneously. (Preferential voting systems, where a lower preference doesn't affect the chances of your highest-placed choice, are a different matter, and in fact are sometimes criticised by their opponents for allegedly violating monotonicity, though I think this is not a reasonable criticism.)
The second point is a question raised by Liz Batty at 21:42, on which Alison calls me out by name at at 22:23. In a situation where a TV series qualifies for Best Dramatic Presentation, Long Form, and one or more episodes of it qualify for Best Dramatic Presentation, Short Form, how do administrators decide which will actually appear on the ballot?
This is of course a recent problem. but my answer is clear: if faced with a choice, administrators should choose whatever alternative gets more voters the thing they want on the vote. To go into the detail:
The Dramatic Presentation categories were only split from 2003. The first series to qualify was season 1 of Heroes, in 2008, and only one of its individual episodes even scraped onto the long list that year. It was in 2012 that the first season of Game of Thrones got 171 votes for the Long Form ballot, and individual episodes got 73 and 60 votes which would have been enough to qualify for the Short Form ballot (and another episode was runner-up with 49 votes); but the administrators decided that the series rather than the individual episodes should qualify. George R.R. Martin had anticipated this outcome several months earlier. I do not know if the showrunners were consulted. The series won the Long Form Hugo.
It's worth noting that the 171 votes for the series in Long Form are more than the sum of the votes for the two episodes that made the top five in Short Form (73+60 = 133), and of course a number of voters will have voted for both of the Short Form episodes, so the total number of voters who wanted to see either of the two on the Short Form ballot would have been much less than 133. The episode which came sixth, with 49 votes, doesn't count here, as it would never have been on the final ballot anyway.
Edited to add: A calculation I made for the other years when I first posted this, but forgot to make for 2012, is the down-ballot impact of the decisions available to the administrators. Excluding the series would have brought a film with 94 votes onto the Long Form ballot. Including the two GoT episodes would have excluded two nominees with 38 and 36 votes which ended up on the Short Form ballot in our timeline. So the administrators' decision was in line with the wishes of up to 171+38+36 = 245 voters, whereas the alternative would have satisfied a maximum of 94+73+60 = 227 voters. There would have been some overlap of supporters in the short form nominations, though more in the latter scenario than the former.
The following year, 2013, the second series of Game of Thrones got 164 votes for Long Form, and an individual episode got 95 votes for Short Form. The showrunners were consulted this time as to which should be on the ballot, and opted for the individual episode, which indeed won the Short Form Hugo. This is the only occasion of the five times the situation has arisen (2012, 2013, twice in 2020 and 2021) where the showrunners were consulted, as far as I am aware. This brought a film with 141 votes onto the Long Form ballot; otherwise an audiobook with 58 votes would have replaced the GoT episode on the Short Form ballot. So arguably more voters were satisfied with the actual ballot (141+95 = 236) than with the alternative (164+58 = 222).
The next TV series to qualify for the Long Form category was the first of Stranger Things, in 2017, my first year as Hugo Administrator. None of the episodes came close to qualifying in Short Form however, so there was no decision to make. NB that from that year on there were six finalists per category, rather than five as previously.
The most complex decision so far was in 2020, when the top six nominees for Long Form included the TV series Good Omens, with 212 votes, and Watchmen, with 81; while the top six nominees in Short Form included an individual episode of Good Omens with 104 votes, and two episodes of Watchmen with 81 and 59. I was Deputy Administrator that year, and we made our thinking pretty clear: more voters supported Good Omens being in Long Form, and more voters supported at least one of the two Watchmen episodes in Short Form, so that was the decision we made, without consulting the showrunners. (We were also, as you may remember, in the middle of a global pandemic.)
If we had instead kept the Good Omens episode on the Short Form ballot (where it actually came top) we'd have got a film with 74 votes on the Long Form ballot and lost a TV episode with 36 votes on the Short Form ballot, thus satisfying 104+74 = 178 voters rather than 212+36 = 248 voters. On the other hand, if we'd kept Watchmen on Long Form and dropped the two episodes from Short Form, the category ballots would have lost a film with 75 votes and gained two TV episodes with 34 and 35, which also clearly satisfies fewer voters, especially if the two TV episodes had supporters in common.
We followed that precedent again in 2021 (official results sheet, tidier version which I supplied too late to the convention), when I was WSFS Division Head and a member of the Hugo sub-committee at the time nominations closed; the second series of The Mandalorian got 67 votes for Best Dramatic Presentation, Long Form, and two individual episodes got 67 and 45 for Short Fom. Clearly more voters supported having at least one of the episodes in Short Form, so again without consulting the showrunners, that was the decision that we made. That brought a film onto the Long Form ballot that had got 63 votes; if we had gone the other way, two TV episodes with 28 and 25 votes would have been on the Short Form ballot. I don't need to work out the arthmetic in detail, but you can if you want; it was pretty clear what outcome better reflected the wishes of the voters.
So, in summary (I bet you're glad there's a summary), the practice of administrators has been that the wishes of voters should be given priority, if they are clear, and whichever alternative gives more voters a thing they want on the ballot is the one that should be followed. 2013 may look like an exception at first glance, but if you look at the down-ballot consequences of the decision, it's also defensible in those terms.
I would resist any move to formally throw the decision to showrunners rather than the Hugo administrators. I'm uneasy with the idea that studio execs rather than WSFS voters should get to decide what is on the Hugo ballot, and it should be added that in practice, many showrunners are not very responsive to communications from Hugo administrators (Game of Thrones was very much an honourable exception there).
I'm also opposed to further codifying existing practice in the rules. Let's concentrate on fixing the things that need fixing, in particular the Best Artist categories, and not waste time on the things that already work.
But thank you for asking, Alison!
First off, at 18:53, Alison expresses the concern that under the counting system (known as "EPH, short for "E Pluribus Hugo"), "if I nominate anything else, it reduces the chance that the thing I love gets on the ballot. And that seems to me to be the big downside of EPH." The podcast is kind enough to link to the explanation of EPH that I wrote in 2017.
It is of course perfectly true that under the current counting system, if you love one thing, and you nominate other things as well, you reduce the chance that your vote will help the one thing you love getting on the ballot. But this was equally true of the old tallying system. It's the principle of monotonicity, as Kenneth Arrow put it in his famous Theorem. Candidate A getting more votes should not lead to her getting a worse result, and Candidate B getting fewer votes should not lead to his getting a better result. If you give an extra vote to A, you are inevitably hurting B's chances, even if you love B more. So it's not uniquely a problem with EPH, but with any system where your vote can go to several candidates simultaneously. (Preferential voting systems, where a lower preference doesn't affect the chances of your highest-placed choice, are a different matter, and in fact are sometimes criticised by their opponents for allegedly violating monotonicity, though I think this is not a reasonable criticism.)
The second point is a question raised by Liz Batty at 21:42, on which Alison calls me out by name at at 22:23. In a situation where a TV series qualifies for Best Dramatic Presentation, Long Form, and one or more episodes of it qualify for Best Dramatic Presentation, Short Form, how do administrators decide which will actually appear on the ballot?
This is of course a recent problem. but my answer is clear: if faced with a choice, administrators should choose whatever alternative gets more voters the thing they want on the vote. To go into the detail:
The Dramatic Presentation categories were only split from 2003. The first series to qualify was season 1 of Heroes, in 2008, and only one of its individual episodes even scraped onto the long list that year. It was in 2012 that the first season of Game of Thrones got 171 votes for the Long Form ballot, and individual episodes got 73 and 60 votes which would have been enough to qualify for the Short Form ballot (and another episode was runner-up with 49 votes); but the administrators decided that the series rather than the individual episodes should qualify. George R.R. Martin had anticipated this outcome several months earlier. I do not know if the showrunners were consulted. The series won the Long Form Hugo.
It's worth noting that the 171 votes for the series in Long Form are more than the sum of the votes for the two episodes that made the top five in Short Form (73+60 = 133), and of course a number of voters will have voted for both of the Short Form episodes, so the total number of voters who wanted to see either of the two on the Short Form ballot would have been much less than 133. The episode which came sixth, with 49 votes, doesn't count here, as it would never have been on the final ballot anyway.
Edited to add: A calculation I made for the other years when I first posted this, but forgot to make for 2012, is the down-ballot impact of the decisions available to the administrators. Excluding the series would have brought a film with 94 votes onto the Long Form ballot. Including the two GoT episodes would have excluded two nominees with 38 and 36 votes which ended up on the Short Form ballot in our timeline. So the administrators' decision was in line with the wishes of up to 171+38+36 = 245 voters, whereas the alternative would have satisfied a maximum of 94+73+60 = 227 voters. There would have been some overlap of supporters in the short form nominations, though more in the latter scenario than the former.
The following year, 2013, the second series of Game of Thrones got 164 votes for Long Form, and an individual episode got 95 votes for Short Form. The showrunners were consulted this time as to which should be on the ballot, and opted for the individual episode, which indeed won the Short Form Hugo. This is the only occasion of the five times the situation has arisen (2012, 2013, twice in 2020 and 2021) where the showrunners were consulted, as far as I am aware. This brought a film with 141 votes onto the Long Form ballot; otherwise an audiobook with 58 votes would have replaced the GoT episode on the Short Form ballot. So arguably more voters were satisfied with the actual ballot (141+95 = 236) than with the alternative (164+58 = 222).
The next TV series to qualify for the Long Form category was the first of Stranger Things, in 2017, my first year as Hugo Administrator. None of the episodes came close to qualifying in Short Form however, so there was no decision to make. NB that from that year on there were six finalists per category, rather than five as previously.
The most complex decision so far was in 2020, when the top six nominees for Long Form included the TV series Good Omens, with 212 votes, and Watchmen, with 81; while the top six nominees in Short Form included an individual episode of Good Omens with 104 votes, and two episodes of Watchmen with 81 and 59. I was Deputy Administrator that year, and we made our thinking pretty clear: more voters supported Good Omens being in Long Form, and more voters supported at least one of the two Watchmen episodes in Short Form, so that was the decision we made, without consulting the showrunners. (We were also, as you may remember, in the middle of a global pandemic.)
If we had instead kept the Good Omens episode on the Short Form ballot (where it actually came top) we'd have got a film with 74 votes on the Long Form ballot and lost a TV episode with 36 votes on the Short Form ballot, thus satisfying 104+74 = 178 voters rather than 212+36 = 248 voters. On the other hand, if we'd kept Watchmen on Long Form and dropped the two episodes from Short Form, the category ballots would have lost a film with 75 votes and gained two TV episodes with 34 and 35, which also clearly satisfies fewer voters, especially if the two TV episodes had supporters in common.
We followed that precedent again in 2021 (official results sheet, tidier version which I supplied too late to the convention), when I was WSFS Division Head and a member of the Hugo sub-committee at the time nominations closed; the second series of The Mandalorian got 67 votes for Best Dramatic Presentation, Long Form, and two individual episodes got 67 and 45 for Short Fom. Clearly more voters supported having at least one of the episodes in Short Form, so again without consulting the showrunners, that was the decision that we made. That brought a film onto the Long Form ballot that had got 63 votes; if we had gone the other way, two TV episodes with 28 and 25 votes would have been on the Short Form ballot. I don't need to work out the arthmetic in detail, but you can if you want; it was pretty clear what outcome better reflected the wishes of the voters.
So, in summary (I bet you're glad there's a summary), the practice of administrators has been that the wishes of voters should be given priority, if they are clear, and whichever alternative gives more voters a thing they want on the ballot is the one that should be followed. 2013 may look like an exception at first glance, but if you look at the down-ballot consequences of the decision, it's also defensible in those terms.
I would resist any move to formally throw the decision to showrunners rather than the Hugo administrators. I'm uneasy with the idea that studio execs rather than WSFS voters should get to decide what is on the Hugo ballot, and it should be added that in practice, many showrunners are not very responsive to communications from Hugo administrators (Game of Thrones was very much an honourable exception there).
I'm also opposed to further codifying existing practice in the rules. Let's concentrate on fixing the things that need fixing, in particular the Best Artist categories, and not waste time on the things that already work.
But thank you for asking, Alison!
Last of the Hugo nominees for Best Graphic Story (well, the last that I will read; I will skip Schlock Mercenary as I did last year).