Liz

The entire theory depends on ignoring the actual ideology on the ground and assuming Palin voters would just as soon vote for a Democrat.

It literally says the opposite, and there's no assumption, it's right there in the voting data. Begich beats Palin and Peltola one-on-one. I'm sorry that you've heard other people talk about this particular election in bad faith, but that's not what I'm doing. We can talk about other particular RCV elections that had spoilers, if you like.

And you haven't mentioned any other type of approval voting until now so yes that's what was assumed.

I mentioned both regular approval and SPAV in my first comment. Maybe that's where some of this confusion is coming from.

STV is also a multiple winner election system. Which is also incompatible with our Constitution.

Can you quote the section that prohibits multi-winner elections? At this point some of the things you've said have me believing you might an inauthentic account, unfortunately. I apologize if you're an earnest American, but I have now have my doubts.

The entire theory depends on ignoring the actual ideology on the ground and assuming Palin voters would just as soon vote for a Democrat.

It literally says the opposite, and there's no assumption, it's right there in the voting data. Begich beats Palin and Peltola one-on-one. I'm sorry that you've heard other people talk about this particular election in bad faith, but that's not what I'm doing. We can talk about other particular RCV elections that had spoilers, if you like.

And you haven't mentioned any other type of approval voting until now so yes that's what was assumed.

I mentioned both regular approval and SPAV in my first comment. Maybe that's where some of this confusion is coming from.

STV is also a multiple winner election system. Which is also incompatible with our Constitution.

Can you quote the section that prohibits multi-winner elections? At this point some of the things you've said have me believing you might an inauthentic account, unfortunately. I apologize if you're an earnest American, but I have now have my doubts.

I dunno why you're bringing back SPAV into this, the discussion has had very little to do with it. There are local races that use STV, which is a bigger change to the voting and representation system than SPAV is.

You should just skip down to the part that explains that yes, Palin was a spoiler. You don't seem to be particularly interested in actually having a discussion, I'm not here to score wins or attack one system or another. I'm here to provide and receive a better understanding of how voting and representation systems work. You don't seem to be particularly interested in that.

I didn't tell you to trust Wikipedia, I told you to follow through to the linked sources in that section. Also, that talk page suffers from the same problem you're having, which is assuming that the RCV results are the same thing as the public opinion. The entire point of analysing the data is to look past the voting system used and try to understand what people's preferences are. Here's another (very long) source that summarizes the full ballot data and explains that, yes, Palin was a spoiler. Justifying this as acceptable by saying that RCV followed its own rules (which it must do, by definition) is the same as saying Ralph Nader spoiling the 2000 election was the correct outcome because those people had Nader as their favorite.

Look I don't hate RCV. It's certainly better than FPTP. I just don't want people to have false ideas about its function. Spoilers can and do happen, they just behave differently than FPTP. And, I will add, they behave in a much more acceptable way, with RCV spoilers being much more likely to be competitive candidates compared to FPTP. Plus, RCV has less center-squeeze than FPTP. Mathematically, Approval doesn't have spoilers nor does it have a center-squeeze effect, and I would argue that it's better than both RCV and FPTP for this and other reasons, but I do want to re-confirm that FPTP is the worst.

Everything alive today is the pinnacle of evolution.

SPAV is specifically constructed to work with proportional representation. It iterates until all seats are filled.

Yes, that's how it works. The first round is functionally identical to regular approval which is why I like using the two. Approval for single-winner, SPAV for multi-winner.

But in the US, by Constitutional law, it's one seat per geographical district.

I'm pretty sure it's just federal law, but I would have to double check. Not like Congress would change it anyway.

A traditional party primary would have nominated Palin, not Begich, and she would have lost anyways.

That's pure speculation. But using the voting data from the general, we Begich was preferred to both Palin and Peltola in head-to-head matchups. Palin pulled enough votes from Begich to eliminate him in the first round and he lost to Peltola in the second. If Palin hadn't run Begich would have won.

You can read more about it from the linked sources here.

Here's the most relevant section:

Some social choice and election scientists criticized the election in published opinion pieces, saying it had several perceived flaws, which they technically term pathologies. They cited Begich's elimination as an example of a center squeeze, a scenario in which the candidate closest to the center of public opinion is eliminated due to failing to receive enough first choice votes. More voters ranked Begich above Peltola, but Palin played the role of spoiler by knocking Begich out of contention in the first round of the run-off. Specialists also said the election was notable as a negative vote weight event, as those who voted for Palin first and Begich second instead helped Peltola win by pushing Palin ahead of Begich in the first round.

Elections scientists were careful to note that such flaws (which in technical terms they call pathologies) likely would have occurred under Alaska's previous primary system as well. In that binary system, winners of each party primary run against each other in the general election. Several suggested alternative systems that could replace either of these systems.

You have to be careful analysing RCV results, because people tend to only look at what the election did, and fail to look at what it didn't do. One of the good things about RCV is that it collects a fair bit of information, but then it usually ignores a fair bit of it. When trying to understand whether a candidate was a spoiler or not, you have to ask what would have happened if they didn't run at all, which requires considering collected information the "unaltered" election didn't take into account. If removing them from the election changes the winner of the race, then they were a spoiler. We know that removing Palin would have resulted in a Begich win over Peltola, so that makes Palin a spoiler. She's a losing candidate that changed the winner of the race simply by entering, assuming voter preferences are stable.

Proportional representation specifically refers to how parties divide the available seats

I apologize for not addressing that, but I didn't think it required expanding on. Yes, that's correct. I feel the preferred proportional method is Sequential Proportional Approval Voting

RCV doesn't eliminate vote splitting, it only mitigates it. If two candidates have similar support in a non-final round, one can act as a spoiler for the other. The problem is that it's harder to understand and FairVote used to lie about it, so a lot of people think it's not a problem. The Alaska special election from a few years back is an example of a spoiler election. If Palin hadn't run (or fewer Palin voters voted) the other Republican would have won. If you want to completely eliminate vote splitting you have to move to a cardinal voting method that satisfies the independence of irrelevant alternatives criterion, which is most of them, including approval voting.

There are lots of voting systems that make third parties less damaging to major parties. Approval, RCV, STAR, Score, to name a few. Approval is a better choice because it's much easier to use and explain to people (RCV disenfranchises minorities, poor people, and under educated voters), while generally agreeing with RCV on the results. Plus, it's much easier to expand to proportional representation, when we get there.

That linked data is collected from local American races. The winner is overwhelmingly the person who won the first round, which is the only round the majority of the time. When people claim RCV will break the two party system they are trying to claim it will change the winners. The evidence largely shows that no voting system can take a single-winner duopoly and break it.

Any new voting system would require only a simple bill from the legislature. "Ballots instructions for every election at every level shall direct voters to select any number of candidates. The candidates with the most votes wins their respective election."

Changing the voting system has nothing to do with the parties in power. Also, it's a referendum campaign. You'd be collecting signatures from the citizens in order to get it on the ballot. Pretty much all you have to do is find some primaries where the winner got like 25% percent of the vote and talk about how unacceptable that is. St. Louis uses approval for their primaries instead of the general. Approval asks what fraction of the population approved of a candidate, so the winner's percentage is practically guaranteed to go up, demonstrating they actually do have broad support.

1. RCV and two-round runoff are very different in practice because the two round system encourages strategic voting, has a higher potential for spoilers (RCV has them too), and has an intermediate time where the advancing candidates have to fight over all the voters who didn't pick them in the first round, which is meaningfully different from when they were a part of the pack.

2. France has some amount of proportional representation at the local level.

3. They're not starting from an entrenched two party system.

4. They're honestly simply one of the big exceptions, it's fairly well-established that single-winner methods tend towards two parties pretty much no matter what you do. Typically when you see more than two parties at the national level, it's because there are regional pockets where only two parties are competitive, but it's not always the same two parties. I'm not familiar with the details about the French political situation, but yeah, they've got a very unusual number of parties for a single-winner dominated structure. Compare them with Australia, who have proportional representation at the national level, and it should be pretty clear they're just plain exceptional. If you need more evidence, Texas, Mississippi, and Georgia already use a two round system for their legislatures but they still have a two party system.

I dunno how much you know about representation and voting systems, but the wiki article on two round systems is pretty good.

You start from the bottom and work your way up. Switch your local elections to approval with a referendum campaign, and by the time you get up to the state level you'll have people in office who have already proven they can win under approval. I'm serious. You should run a referendum campaign.