What Is Voter Satisfaction Efficiency?

> IRV can be implemented as either requiring complete ordering or allow partial ordering (you have n candidates, you can assign each one level 1..n, but you may use the same level for more candidates to describe partial ordering).

Technically, that’s a complete ordering with unforced preferences, not a partial ordering.

A ballot with a partial ordering would be won for which for some pair of candidates A, B, none of A> B, AEqual preference and absence of preference are different.

A ballot that worked this way would be a meaningful extension in existing ranked preference methods that aggregate pairwise preferences from individual ballots, lile ranked pairs. (Whether the effect would be desirable is not something I’ve really analyzed, but that kind of method would naturally make distinct use of the information.)

What's kinda interesting about Election Science is how they've changed strategies. They used to push STAR till about a year ago and have since moved to approval voting. It seems they have found that approval is easier to sell to the average person and is still good enough.

> Valuing clear and honest discussion about member preferences in voting systems

This is a difficult thing tbh. A lot of people learn about voting systems from CGP Grey or Hasan Minhaj and get really passionate, so you get an armchair expertise. Then I think many learn about VSE and see Condorcet methods as the obvious winners (I used to be in this camp). But it often takes awhile to internalize all the nuances in voting. About how to balance VSE, resiliency, simplicity, transparency, computational cost, and more. It is one of those problems that looks easier than it is. A lot of very smart people get into it because it is an interesting problem but it is also easy to convince yourself that your understanding is far better than it is. Every time these threads come up I learn something new and I've been interested in the subject for almost a decade. I think the biggest thing I've learned is to look at the people who have been studying the problem for a long time and see why they are making their decisions. It is difficult to separate the signal from the noise.

In addition to the discord link I also suggest following Clay Shentrup[0]. He's active with Election Science and the co-inventor of STAR (he also typically joins voting threads on HN. Username is his full name). The reason I follow him is because reading his comments and posts have led me to a lot more sources and brought up a lot of the above nuances I didn't understand when I first started getting into the subject.

[0] https://clayshentrup.medium.com/

> What is the simple way to pitch its advantage to IRV without using technical terms?

Good question. It's probably that the tabulation works the way a voter thinks it would.

Neither IRV/RCV nor approval has a way to tell how much more you prefer your favorite to your #2 or #3 (something STAR/Score has, and kudos to it). What's that mean? They both have to assume. Most voters don't realize that IRV assumes massive gulfs between their #1, #2, and #3.

The example I give is Seattle's 2021 mayoral race, which has 15 candidates, at least 8 of whom are running serious races. With 8+ serious candidates, all sharing the left-of-center spectrum, there's tons of overlap. So, imagine you and a friend or spouse have similar views and support the same candidates, but rank them subtly differently. You rank A, B, C, your friend ranks B, C, A. In each runoff, your and your friend's opinions will end up competing with one another, even though both of you share the same opinions and support the same candidates.

(For anyone else reading, here's a longer version that I didn't write: https://psephomancy.medium.com/how-ranked-choice-voting-elec.... Ignore the title - the problem isn't extreme or not extreme, it's that the result doesn't reflect the opinion of the electorate.)

And the reverse is also true. IRV unintentionally rewards candidates who have their own dedicated constituency who always put their candidate #1. While it's not my role to moralize about that, choosing a candidate because a small group of voters vehemently support a candidate is not likely to make the most voters happy.

Since you brought up simple explanations.. a related simple benefit is that essentially everyone can explain how winners are chosen in approval voting. Very few people, even many active proponents of IRV, can thoroughly explain tabulation of IRV ballots. In an era when election processes are becoming politicized, a system that the average person understands well enough to explain to someone else is valuable.

I'm ignoring the most obvious easy-to-explain reason, that it does a better job of "reading voters minds." That's part of the original submission, though.

(I've now talked or exchanged emails with a fair number of people who thought they were RCV/IRV supports, and I learned that STAR delivers the benefits that many or most assumed that RCV provides. STAR lets you score every candidate on a 1-5 star scale, so you can rate candidates the same, rate them better/worse than one another, and critically, can decide for yourself how big each gap is. The vast majority of people who think they're strongly supportive of RCV/IRV would switch their support to STAR if they heard of it, and when I talk with them, they often do switch.)

> My belief is IRV opens the door to approval voting.

I support IRV as an improvement, but I'd need more evidence that IRV opens the door for approval - or that it doesn't. If you have evidence or at least a hypothesis how that would happen, I'd appreciate it.

For example, would IRV fail (what's that mean?), yet voters would not recoil, they'd decide they like the idea and just chose the wrong thing? Or would it succeed and people would look for something better? Something else?

My baseline is that it took us something like 200 years to get here (and 30 years since FairVote was created). If history is a guide, whatever we pick seems like it'll stick around for a while. That's a weakly held belief, though.

> When approval voting advocates oppose IRV in jurisdictions with plurality voting, however, it just splits the effort and leaves FPTP in place.

I'm probably the wrong person to ask about this, since I don't oppose IRV - I just don't support it anywhere near as much as I do approval or STAR. My take is simpler: Seattle doesn't have either right now, and if we're going to switch, getting something great on the ballot (an initiative to change to AV) doesn't require any more lobbying or fundraising effort than something good or okay.

A skeptic could have asked "But will it pass?" until a year or two ago. St. Louis just switched to approval last year (https://www.stltoday.com/news/local/govt-and-politics/overha...) and 68% of voters were in favor (!), and Fargo, ND switched so in 2019, so that's been affirmed. The Center for Election Science was only created in 2011, so for 9 years (3-5 years of which was discovery), that's a fast adoption curve.

Also, approval is the only one that I think delivers enough benefit to be worth a significant amount of my own time. That's an entirely personal judgment though.

(I'd have a different answer if AV fans were running an explicit campaign against an IRV/RCV initiative that was on the ballot, but again, that's not me. At least in Seattle, I don't think either one is more or less difficult than the other.)

Maybe, but as I understand it, within voting system reform there's at least one big practical conflict: the tradeoff between 'the best' vs. the simplest voting system that you can convince people is marginally better. The average voter has no clue what's wrong with the current system.

Proponents of one system over the other worry that voters will get confused, or worse, that established groups that disagree with the reform could attack a new voting system as too complicated.

Personally I think this is part of the reason approval has gotten more traction than score or STAR systems, but who's to say. The Center for Election Science has better PR and web design than https://www.rangevoting.org/, which largely advocates for Score voting.

But it's a multifaceted issue. Academics argue over different methods, and even as a casual observer it seems pretty clear that no system is flawless.

For anyone who hasn't read "Gaming the Vote" by Poundstone, it's a useful introduction that discusses both the academic study and practical use of voting systems.

Not necessarily; I think it's usually more of a prisoner's dilemma situation: if you optimize for getting what you want, then the overall outcome is worse.

For example, in the results Borda count is much worse when everyone strategic votes.

The new results contrast with an older simulation (graph shown down the page): https://electionscience.org/library/tactical-voting-basics/ Those results show honest voting always being (globally) better. That's one reason why I find the new results surprising.