![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Entry tags:
STV + multiple seats can get weird
Worldbuilding moderator election: four positions, 19 candidates (ten proceeding to final election), single transferable vote (AKA "Australian ballot", like the Hugos). With the Hugos you're choosing one winner; applying the scheme to a race with multiple winners can get a little odd.
I don't understand all the math, but with multiple seats the algorithm sets a "threshold", a number of votes a candidate needs to win a seat, and excess votes are then transferred away to other candidates to try to determine the next round. ("Meek STV", for those to whom that means something.) If that doesn't advance things then the candidate with the fewest votes drops out and those are transferred to other candidates. Iterate until done.
The first three winners were determined in the first two rounds of vote processing. It took another twenty rounds to determine the fourth. Some of those vote transfers are tiny, as in "I'll take your word for it that there are changes there".
Each voter got three picks but there were four seats. We all expected a lot of jockeying for spot #4 because of that. I wonder how the dynamics would have been different if voters had as many picks as open seats.
The (anonymized) ballot data is available for download. I'm curious about patterns but not quite sure how to look for them.
I don't understand all the math, but with multiple seats the algorithm sets a "threshold", a number of votes a candidate needs to win a seat, and excess votes are then transferred away to other candidates to try to determine the next round. ("Meek STV", for those to whom that means something.) If that doesn't advance things then the candidate with the fewest votes drops out and those are transferred to other candidates. Iterate until done.
The first three winners were determined in the first two rounds of vote processing. It took another twenty rounds to determine the fourth. Some of those vote transfers are tiny, as in "I'll take your word for it that there are changes there".
Each voter got three picks but there were four seats. We all expected a lot of jockeying for spot #4 because of that. I wonder how the dynamics would have been different if voters had as many picks as open seats.
The (anonymized) ballot data is available for download. I'm curious about patterns but not quite sure how to look for them.