4 ms·

This is actually a really interesting real-life instance showing that, in this particular case, RCV didn't achieve an optimal outcome. But what's a better voti

This is actually a really interesting real-life instance showing that, in this particular case, RCV didn't achieve an optimal outcome. But what's a better voting system that would have achieve the presumably optimal outcome here (the less-polarizing GOP candidate Begich getting elected)? If we just compare to simply plurality voting, the outcome seems the same: Peltola got the highest vote of the 3, so in regular FPTP voting, she would have won. Or if we compare to FPTP party-based elections as is normally done in the US, I think again it would have been the same: the GOP vote was split between Palin and Begich, and presumably Palin would have narrowly won that party's primary, and then enough GOP voters hated her enough to list Peltola 2nd-choice in the RCV, so presumably they would have also voted for Peltola in a regular party-based election, so again the outcome would have been the same. If anything, I think this shows that these multi-candidate elections have odd-looking results when there's uneven numbers of candidates for the different parties. If there had been 2 DNC candidates, this might have played out quite differently. But maybe not: in this particular election, one of the candidates (Palin) was extremely polarizing and many voters truly hated her, but she had enough support from the right-leaning voters to get a majority on that side, over the more moderate Begich. But that was unique to the GOP side; there's no reason to think that the left-leaning voters of Alaska would have similarly voted in high numbers for an extremely polarizing candidate like AOC: Alaska just doesn't have many far-left voters the way some other states do. So it seems like the right outcome may have been reached here, because so many GOP voters in Alaska are willing to support a very extreme and polarizing candidate, rather than a more moderate one: they ended up making themselves irrelevant.

AlbertCory 2y ago
I think "begging the question" describes your answer, in the classical sense of "assuming what you need to prove." You are assuming that a multi-party, multi-candidate election is a desirable norm. But that needs to be proven by you, not assumed. In a pre-RCV system, the parties would choose their candidates, however they do it, and then the general election would pit one candidate of each party against each other, plus whatever other parties qualify for the ballot. The public has not accepted that that must be eliminated. It did make things hard for third parties, but so it goes. Prove that that's a bad thing.
- bb88 2y ago
  > The public has not accepted that that must be eliminated. RCV will be on the ballot this fall in Ruby Red Idaho. Let's see what happens. > You are assuming that a multi-party, multi-candidate election is a desirable norm. Easy. Both Trump's and Biden's approval numbers are both in the toilet. Yet they both easily won their primaries.
  - AlbertCory 2y ago
    If that's your idea of "proof" I think we're done here.
    - bb88 2y ago
      So you'll take your arguments and go home, and not even try to convince me otherwise and potentially win me over with your candor and intellect and personality? That seems like the problem with politics as a whole these days.
      - AlbertCory 2y ago
        No, citing one example of the current system as "proof" that the whole system must be thrown out is the problem with "politics as a whole." Also: thinking that a few paragraphs on Hacker News constitutes sufficient wisdom is another problem with "politics as a whole." It is not going to convince you no matter what I or anyone else says. Do some research. That probably means reading some books, hate to warn you.
Izkata 2y ago
> But what's a better voting system that would have achieve the presumably optimal outcome here (the less-polarizing GOP candidate Begich getting elected)? One I don't think gets talked about enough is approval voting. Vote for as many candidates as you like, whoever gets the most votes wins. Voting for all of them is effectively the same as voting for none of them. It also has the massive benefit for the general population that it's simple to explain. Even gives third-party candidates and moderate candidates a real chance at winning, since no one would feel like they're throwing their vote away by including them.
- em-bee 2y ago
  i like this. i think it even works for electing multiple candidates. we need 5 people in a committee. everyone list their preferred 5 or 10 or more choices. add up all the votes and the 5 top get chosen.

AlbertCory 2y ago

I think "begging the question" describes your answer, in the classical sense of "assuming what you need to prove." You are assuming that a multi-party, multi-candidate election is a desirable norm. But that needs to be proven by you, not assumed. In a pre-RCV system, the parties would choose their candidates, however they do it, and then the general election would pit one candidate of each party against each other, plus whatever other parties qualify for the ballot. The public has not accepted that that must be eliminated. It did make things hard for third parties, but so it goes. Prove that that's a bad thing.

bb88 2y ago
> The public has not accepted that that must be eliminated. RCV will be on the ballot this fall in Ruby Red Idaho. Let's see what happens. > You are assuming that a multi-party, multi-candidate election is a desirable norm. Easy. Both Trump's and Biden's approval numbers are both in the toilet. Yet they both easily won their primaries.
- AlbertCory 2y ago
  If that's your idea of "proof" I think we're done here.
  - bb88 2y ago
    So you'll take your arguments and go home, and not even try to convince me otherwise and potentially win me over with your candor and intellect and personality? That seems like the problem with politics as a whole these days.
    - AlbertCory 2y ago
      No, citing one example of the current system as "proof" that the whole system must be thrown out is the problem with "politics as a whole." Also: thinking that a few paragraphs on Hacker News constitutes sufficient wisdom is another problem with "politics as a whole." It is not going to convince you no matter what I or anyone else says. Do some research. That probably means reading some books, hate to warn you.

Izkata 2y ago

> But what's a better voting system that would have achieve the presumably optimal outcome here (the less-polarizing GOP candidate Begich getting elected)? One I don't think gets talked about enough is approval voting. Vote for as many candidates as you like, whoever gets the most votes wins. Voting for all of them is effectively the same as voting for none of them. It also has the massive benefit for the general population that it's simple to explain. Even gives third-party candidates and moderate candidates a real chance at winning, since no one would feel like they're throwing their vote away by including them.

em-bee 2y ago
i like this. i think it even works for electing multiple candidates. we need 5 people in a committee. everyone list their preferred 5 or 10 or more choices. add up all the votes and the 5 top get chosen.