4 ms·

I think when people talk about RCV (in the context of single-winner elections), they're almost always talking about the same thing as IRV. "Ranked voting" is t

by elihu 6y ago

I think when people talk about RCV (in the context of single-winner elections), they're almost always talking about the same thing as IRV. "Ranked voting" is the more general term. > Ranked voting is any election voting system in which voters use a ranked (or preferential) ballot to rank choices in a sequence on the ordinal scale: 1st, 2nd, 3rd, etc. There are multiple ways in which the rankings can be counted to determine which candidate (or candidates) is (or are) elected (and different methods may choose different winners from the same set of ballots). The other major branch of voting systems is cardinal voting, where candidates are independently rated, rather than ranked.[1] > The similar term "Ranked Choice Voting" (RCV) is used by the US organization FairVote to refer to the use of ranked ballots with specific counting methods: either instant-runoff voting for single-winner elections or single transferable vote for multi-winner elections. In some locations, the term "preferential voting" is used to refer to this combination of ballot type and counting method, while in other locations this term has various more-specialized meanings. https://en.wikipedia.org/wiki/Ranked_voting https://en.wikipedia.org/wiki/Ranked_voting I share your frustration with FairVote and CGP Grey (and recently an episode of Patriot Act) painting an unrealistically rosy picture of RCV/IRV as the solution to our voting problems. There are some other systems that aren't ranked voting at all that I think would be significantly better than IRV, such as approval voting, range voting, and STAR voting (which is basically range voting with an immediate runoff between the top two). For some reason, FairVote persists in claiming that in approval voting (which is like traditional first-past-the-post voting except that you can vote for more than one candidate), you somehow maximize your voting influence by voting for only one candidate. They call this bullet voting. I don't know what possible reason one could have for believing this. This manifests in their comparison chart: https://www.fairvote.org/alternatives https://www.fairvote.org/alternatives They show "resistance to strategic voting" as "high" under RCV and "low" under approval voting. This is somewhat subjective, but I think they've got it backwards. Under approval voting, strategic voting and honest voting are basically the same: you vote for as many of the candidates as you can tolerate, and maybe if there's a candidate you really don't want to win you vote for everyone but them. Under RCV, it's only safe to put you first choice first if they are either clearly in the lead or so far behind they have no hope of winning. If the outcome is in doubt, it's possible to cause your first place candidate to lose by putting them first. (That's a weird problem for any voting system to have.) For this reason, I don't trust FairVote; I don't think they're presenting their preferred voting system or the alternatives honestly. They're more of an advocacy/PR organization that's pushing their chosen solution out of a sense of inertia or something. The Center for Election Science is a smaller, less-well funded group that advocates for approval voting (and similar methods) that I donate to from time to time. They had a successful campaign to convert Fargo ND to approval voting back in 2018.

godelski 6y ago
I'm basically agreeing with literally everything you are saying, I just want to expand on certain things. > RV vs RCV These terms are too similar and really just confuse people, which is why I push back against it. I try to use "ordinal" a lot, but it is best for places like HN but not when I'm discussing with family. RCV makes people think that this is the only way you can rank people. > STAR Anyone who is advocating for STAR has my approval. It is the preferred system in my mind. I want to plug my longer post that has a lot of links and expands on all the topics you described https://news.ycombinator.com/item?id=23705413 https://news.ycombinator.com/item?id=23705413 > Fair Vote frustration I think this is a style of argument that is increasingly becoming more common and is destructive. Pushing the your argument past the point of validity in an effort to make it stronger. Their use of broad classes is a gross mischaracterization. When they discuss Condorcet methods they discuss several types, lumping together the worst features from different ones. But they ignore the good methods like RP and Schulze. The comparisons are unfair. But then again, they aren't trying to appeal to people that are informed on the subject, they are trying to appeal to the masses which are completely unfamiliar with the subjects. But I do not think this justifies the outright lying and mischaracterizations. > Spoilers and Strategy In my other post I actually link an example of where IRV fails, and it is the specific type of spoiling that we are about: when similar candidates spoil (e.g. Bernie and Biden, not Stein and Biden). IRV doesn't solve this (but FV claims otherwise). For strategy, this is the argument that I get into with the Condorcet camp (though we're clearly on the same side and these are nuanced friendly arguments not hostile). To me STAR is good because its range of VSE is more compact (i.e. resistant to strategies). The Condorcet camp claims that there is enough dis-incentivization to not vote strategically therefore just maximize VSE (VSE is only slightly different between STAR05 and RP). So for anyone listening on the sidelines, this is really what we're arguing, minutia. I'm certain everyone in the Condorcet camp would vote for STAR and everyone in the STAR/Score/Approval camp would vote for Condorcet (specifically RP) if given the chance. > Center For Election Science I'd also like to mention Equal Vote, which advocates for STAR. https://www.equal.vote/starvoting https://www.equal.vote/starvoting
- AnthonyMouse 6y ago
  > I'm certain everyone in the Condorcet camp would vote for STAR and everyone in the STAR/Score/Approval camp would vote for Condorcet (specifically RP) if given the chance. Eh... I'm in the STAR/Score/Approval camp and I would be somewhat inclined to vote strategically against Condorcet, because it's worse than what I want and once it was implemented it would be even harder to dislodge in favor of STAR/Score/Approval than the status quo. It's also a false compromise because the opponents aren't generally people who benefit from the worse voting method, only people who haven't yet understood why it's worse. Of course, if we were using STAR/Score/Approval to vote on which voting system to use then I could express my preferences more accurately.

by elihu 6y ago

godelski 6y ago

I'm basically agreeing with literally everything you are saying, I just want to expand on certain things. > RV vs RCV These terms are too similar and really just confuse people, which is why I push back against it. I try to use "ordinal" a lot, but it is best for places like HN but not when I'm discussing with family. RCV makes people think that this is the only way you can rank people. > STAR Anyone who is advocating for STAR has my approval. It is the preferred system in my mind. I want to plug my longer post that has a lot of links and expands on all the topics you described https://news.ycombinator.com/item?id=23705413 https://news.ycombinator.com/item?id=23705413 > Fair Vote frustration I think this is a style of argument that is increasingly becoming more common and is destructive. Pushing the your argument past the point of validity in an effort to make it stronger. Their use of broad classes is a gross mischaracterization. When they discuss Condorcet methods they discuss several types, lumping together the worst features from different ones. But they ignore the good methods like RP and Schulze. The comparisons are unfair. But then again, they aren't trying to appeal to people that are informed on the subject, they are trying to appeal to the masses which are completely unfamiliar with the subjects. But I do not think this justifies the outright lying and mischaracterizations. > Spoilers and Strategy In my other post I actually link an example of where IRV fails, and it is the specific type of spoiling that we are about: when similar candidates spoil (e.g. Bernie and Biden, not Stein and Biden). IRV doesn't solve this (but FV claims otherwise). For strategy, this is the argument that I get into with the Condorcet camp (though we're clearly on the same side and these are nuanced friendly arguments not hostile). To me STAR is good because its range of VSE is more compact (i.e. resistant to strategies). The Condorcet camp claims that there is enough dis-incentivization to not vote strategically therefore just maximize VSE (VSE is only slightly different between STAR05 and RP). So for anyone listening on the sidelines, this is really what we're arguing, minutia. I'm certain everyone in the Condorcet camp would vote for STAR and everyone in the STAR/Score/Approval camp would vote for Condorcet (specifically RP) if given the chance. > Center For Election Science I'd also like to mention Equal Vote, which advocates for STAR. https://www.equal.vote/starvoting https://www.equal.vote/starvoting

AnthonyMouse 6y ago
> I'm certain everyone in the Condorcet camp would vote for STAR and everyone in the STAR/Score/Approval camp would vote for Condorcet (specifically RP) if given the chance. Eh... I'm in the STAR/Score/Approval camp and I would be somewhat inclined to vote strategically against Condorcet, because it's worse than what I want and once it was implemented it would be even harder to dislodge in favor of STAR/Score/Approval than the status quo. It's also a false compromise because the opponents aren't generally people who benefit from the worse voting method, only people who haven't yet understood why it's worse. Of course, if we were using STAR/Score/Approval to vote on which voting system to use then I could express my preferences more accurately.