{"items": [{"author": "David", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116966220592", "anchor": "fb-10100116966220592", "service": "fb", "text": "I live in Australia, which has used IRV for a century. In the vast majority of cases, it's fine -- we have a mostly two-party system, usually the two major party candidates are the last two in the instant runoff, and the whole thing's basically equivalent to a regular runoff.

In contests where there are three (or more) candidates with a serious chance of winning, it is certainly true that there is scope for tactical voting. A recent election (Wentworth by-election 2018) saw the Greens leader recommend that Greens supporters ignore the party's official advice to preference the Labor Party, and instead preference a centrist independent; to maximise the chance of the conservative candidate losing, it was important that the centrist stay ahead of Labor at the last exclusion, whereupon Labor preferences would elect the centrist.

In some cases, a major party candidate with no chance of winning may unofficially \"run dead\" (i.e. not campaign strongly) in the hope of finishing behind an independent who has a chance of beating the rival major party after preferences are distributed.

These sorts of practical considerations make the IRV winner to very likely be the Condorcet winner.

There is also a wrinkle about whether preferences are compulsory. In states where preferences are optional, I believe that the Condorcet winner almost always wins (e.g. all 88 seats in the last New South Wales election went to the Condorcet winner). If voters have to rank all candidates, then there are stronger preference flows and unusual results are more likely.

(One theoretical objection I have against always choosing a Condorcet winner is a situation like:
49% A-B-C
49% C-B-A
1.1% B-C-A
0.9% B-A-C

This might come up when major parties A and C hate each other, and some no-name independent B picks up some random votes and is the Condorcet winner. I gather that many Condorcet fans would endorse this outcome, but I don't think you're a serious candidate unless you can get a decent percentage of the first-preference vote.)", "timestamp": "1571485077"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116966220592&reply_comment_id=10100116973416172", "anchor": "fb-10100116966220592_10100116973416172", "service": "fb", "text": "→ David Thanks! It's good to have your view as someone who's watched important elections using IRV.

\"These sorts of practical considerations make the IRV winner to very likely be the Condorcet winner.\"

That's true, and probably becomes more true the better polling gets and the more strategic everyone is. But if this is a goal then Condorcet-IRV is better, since it guarantees this property.

\"One theoretical objection I have against always choosing a Condorcet winner is a situation like...\"

I think this is a good example of why preferences should be optional. If B is a complete nobody, then the only reason they're being given as a second choice is that it's mandatory.

On the other hand, in cases where preferences *are* optional, then seeing B on so many ballots is meaningful, and B probably is a decent compromise candidate. Picking A or C when half the population completely hates them is not a great outcome.", "timestamp": "1571491314"}, {"author": "Martin", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116966220592&reply_comment_id=10100116992208512", "anchor": "fb-10100116966220592_10100116992208512", "service": "fb", "text": "→ Having preferences be optional does seem to be critical.", "timestamp": "1571500394"}, {"author": "Eric", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972083842", "anchor": "fb-10100116972083842", "service": "fb", "text": "The Center for Election Science", "timestamp": "1571489914"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262", "anchor": "fb-10100116972373262", "service": "fb", "text": "I'm curious about your thoughts for a voting system in elections where you have multiple seats open for the same position. This happens frequently in local elections in MA and can cause a problem with rank choice voting (I.e. You are trying to fill 2 Selectboard seats from 5 candidates, or fill 5 Town Meeting seats from 8 candidates).", "timestamp": "1571490389"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116973775452", "anchor": "fb-10100116972373262_10100116973775452", "service": "fb", "text": "→ I like approval voting in general, and I think it works well for this kind of case. For example, if eight people are running for five seats the ballot looks like:

__ A
__ B
__ C
__ D
__ E
__ F
__ G
__ H

You mark the names of anyone you would be happy to see elected, and five with the most support win.", "timestamp": "1571491828"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116973860282", "anchor": "fb-10100116972373262_10100116973860282", "service": "fb", "text": "→ Other places use https://en.wikipedia.org/wiki/Single_transferable_vote , including the Cambridge elections for City Council", "timestamp": "1571491954"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116974124752", "anchor": "fb-10100116972373262_10100116974124752", "service": "fb", "text": "→ Got it. So you wouldn't do any rank choice, just a simple \"Pick 5 of 8\". The problem with the Cambridge system, is their results aren't replicable", "timestamp": "1571492227"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116975427142", "anchor": "fb-10100116972373262_10100116975427142", "service": "fb", "text": "→ Ben not \"pick five of eight\", but \"pick as many as you want\"", "timestamp": "1571493509"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116975487022", "anchor": "fb-10100116972373262_10100116975487022", "service": "fb", "text": "→ Ben I think if Cambridge shifted to fractional redistribution it would be replicable?", "timestamp": "1571493548"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116975551892", "anchor": "fb-10100116972373262_10100116975551892", "service": "fb", "text": "→ I haven't heard about fractional redistribution? Could you explain?", "timestamp": "1571493628"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116976914162", "anchor": "fb-10100116972373262_10100116976914162", "service": "fb", "text": "→ Ben which step makes the Cambridge system non-replicable?", "timestamp": "1571493741"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116978176632", "anchor": "fb-10100116972373262_10100116978176632", "service": "fb", "text": "→ When a candidate exceeds the quota of first place votes, their extra first place ballots are redistributed to the second place person, and so on. This is done randomly, which means that a different batch of votes can be redistributed, yielding different results. For a simple explanation,let's say there are 3 candidates running for 2 seats, and the quota is 2,501 votes. 5,002 people cast ballots, 2,501 voting A/B/C, while 2,501 voted A/C/B. 2,501 votes go to A, who has won, and the other 2,501 have to be redistributed. If the random generator happens to choose all 2,501 A/B/C ballots to be redistributed, then B wins the second seat. If it randomly chooses all 2,501 A/C/B ballots, then C wins. So different people can be elected based on which batch of ballots are randomly selected to be redistributed", "timestamp": "1571494245"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116978196592", "anchor": "fb-10100116972373262_10100116978196592", "service": "fb", "text": "→ This also makes recounts impossible", "timestamp": "1571494258"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116979124732", "anchor": "fb-10100116972373262_10100116979124732", "service": "fb", "text": "→ Ben Yup! So the way you make this replicable is that instead of distributing votes randomly you distribute them fractionally. Since A has 2x the number of votes they need, you redistribute *all* their votes but only count each one at 1/2.", "timestamp": "1571494721"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116979743492", "anchor": "fb-10100116972373262_10100116979743492", "service": "fb", "text": "→ That's interesting. Does it work for the second, third, etc. rounds?", "timestamp": "1571495288"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116980272432", "anchor": "fb-10100116972373262_10100116980272432", "service": "fb", "text": "→ Ben yes! Whenever you have \"randomly decide which n of m ballots to redistribute\" you instead do \"redistribute all m ballots, but only value them at n/m each\"", "timestamp": "1571495456"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100116986145662", "anchor": "fb-10100116972373262_10100116986145662", "service": "fb", "text": "→ That's an interesting approach for properly weighting the votes that I hadn't heard before! The local vote issue comes up frequently at the State House when people talk about rank choice voting, so I like hearing people's thoughts", "timestamp": "1571497329"}, {"author": "Marcus", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116972373262&reply_comment_id=10100117771980842", "anchor": "fb-10100116972373262_10100117771980842", "service": "fb", "text": "→ I actually like the idea of multi-member districts combined with some kind of ranked choice voting for federal congressional elections - as in https://www.congress.gov/bill/115th-congress/house-bill/3057...", "timestamp": "1571864558"}, {"author": "Kim", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973241522", "anchor": "fb-10100116973241522", "service": "fb", "text": "Why do you populate this example with only 4 of the 6 possible options?", "timestamp": "1571491153"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973241522&reply_comment_id=10100116973610782", "anchor": "fb-10100116973241522_10100116973610782", "service": "fb", "text": "→ You're talking about the first example, and the lack of people voting Trump > Clinton > Kasich or Clinton > Trump > Kasich, right? I mostly put those at 0% for simplicity, but it's also a decent assumption in a cases where the candidates differ along a single spectrum. If you imagine the candidates as representing \"left\", \"right\", and \"center\" then \"left > right > center\" and \"right > left > center\" are the two of the six possible orderings you don't expect to see.", "timestamp": "1571491612"}, {"author": "opted out", "source_link": "#", "anchor": "unknown", "service": "unknown", "text": "→ this user has requested that their comments not be shown here", "timestamp": "1571494575"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973241522&reply_comment_id=10100116979434112", "anchor": "fb-10100116973241522_10100116979434112", "service": "fb", "text": "→ Elliot yup, which is how you get cycles like in the second example I give", "timestamp": "1571495113"}, {"author": "Molly", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973306392", "anchor": "fb-10100116973306392", "service": "fb", "text": "Brian I imagine you might have thoughts about this", "timestamp": "1571491212"}, {"author": "Brian", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973306392&reply_comment_id=10100117049214272", "anchor": "fb-10100116973306392_10100117049214272", "service": "fb", "text": "→ Yup. 100%. IRV sucks https://bolson.org/irv/ and it feels ragingly stupid when for no additional cost or complexity we could have a better system (Condorcet's method).", "timestamp": "1571521299"}, {"author": "Molly", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973306392&reply_comment_id=10100117053061562", "anchor": "fb-10100116973306392_10100117053061562", "service": "fb", "text": "→ What is the advantage of Condorcet over approval voting?", "timestamp": "1571522273"}, {"author": "Brian", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973306392&reply_comment_id=10100117053206272", "anchor": "fb-10100116973306392_10100117053206272", "service": "fb", "text": "→ Approval is less expressive. With Approval I can say \"I like A and B\" but with a rankings ballot I can say \"I like A and B but I like A better\"", "timestamp": "1571522353"}, {"author": "Molly", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973306392&reply_comment_id=10100117053615452", "anchor": "fb-10100116973306392_10100117053615452", "service": "fb", "text": "→ Got it. Thanks", "timestamp": "1571522575"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973306392&reply_comment_id=10100117055072532", "anchor": "fb-10100116973306392_10100117055072532", "service": "fb", "text": "→ Brian more expressive isn't necessarily better: cardinal voting less you say \"I like A a little better than B, which I like a lot more than C\" but tactical voting means it's not nearly as good as it looks", "timestamp": "1571522990"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973306392&reply_comment_id=10100117055097482", "anchor": "fb-10100116973306392_10100117055097482", "service": "fb", "text": "→ Brian I'm curious what you think of Condorcet-IRV?", "timestamp": "1571523006"}, {"author": "Brian", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973306392&reply_comment_id=10100117055641392", "anchor": "fb-10100116973306392_10100117055641392", "service": "fb", "text": "→ Condorcet-IRV seems like silly runaround. Why start with IRV - except when we decide to skip to Condorcet instead? Why not just go Condorcet?
For the 'a little better' problem I want to go yet _more_ expressive. Rate each candidate on a scale of 0 to 99. The tactical/strategic voting against raw rating summation can be mitigated with complex election algorithms. (see Instant Runoff Normalized Ratings http://bolson.org/voting/methods.html#IRNR )
But practically, Condorcet can be counted by hand on paper and produces pretty good results and that's my sweet spot for the various tradeoffs.", "timestamp": "1571523305"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973306392&reply_comment_id=10100117062981682", "anchor": "fb-10100116973306392_10100117062981682", "service": "fb", "text": "→ Brian any voting method that chooses the Condorcet winner when one exists needs to do something when a Condorcet winner doesn't exist. IRV seems like a better fallback to me? But I could be convinced?

(Overall I like approval best: simple and easy to explain to voters is an important factor.)", "timestamp": "1571527004"}, {"author": "Brian", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973306392&reply_comment_id=10100117064244152", "anchor": "fb-10100116973306392_10100117064244152", "service": "fb", "text": "→ I think Condorcet cycles are less likely than IRV inversions (I think I had statistics on that once, but can't find them now). I think there are several good cycle resolution rules for a Condorcet election. There is no fixup for an IRV failure, IRV doesn't know it has failed.", "timestamp": "1571527569"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973306392&reply_comment_id=10100117076040512", "anchor": "fb-10100116973306392_10100117076040512", "service": "fb", "text": "→ Brian sure, but I'm not advocating IRV. I'm talking about Condorcet-IRV, where people rank choices, you take a Condorcet winner if there is one, and otherwise you apply IRV.", "timestamp": "1571533678"}, {"author": "Brian", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116973306392&reply_comment_id=10100117105426622", "anchor": "fb-10100116973306392_10100117105426622", "service": "fb", "text": "→ Here's a Condorcet cycle resolution rule I like: for a cycle A beats B beats C beats A, maybe A beats B 60:40, B beats C 58:42, and C beats A 55:45. (A, B, and C as a group beat all other options in the election.) If we discount C beating A because it has the lowest winning score (55), then A becomes virtually undefeated again and can be declared the winner.", "timestamp": "1571546809"}, {"author": "Nix", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116974958082", "anchor": "fb-10100116974958082", "service": "fb", "text": "I like this site for giving intuition about some of the problems that IRV can have: http://zesty.ca/voting/sim/

It does use pretty simplistic computational assumptions which are not going to match reality.", "timestamp": "1571493285"}, {"author": "Nix", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116974958082&reply_comment_id=10100116978151682", "anchor": "fb-10100116974958082_10100116978151682", "service": "fb", "text": "→ Also, worth pointing out that given the assumptions the model makes, there is always a condorcet winner. This is not true in reality, and any voting system has to deal with this possibility (leading to more complicated hybrid systems like Condorcet-IRV)", "timestamp": "1571494194"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116974958082&reply_comment_id=10100116979304372", "anchor": "fb-10100116974958082_10100116979304372", "service": "fb", "text": "→ Nix yeah, if you're going to assume there's always a Condorcet winner then you're leaving out a large portion of the \"voting is weird\" space", "timestamp": "1571494939"}, {"author": "Michael", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116975352292", "anchor": "fb-10100116975352292", "service": "fb", "text": "I'm not sure I get your point... Are you contending that IRV is worse than what we have now, or just that it has flaws?

I don't think anyone would argue with the latter, but it appears to be an improvement on the status quo.", "timestamp": "1571493483"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116975352292&reply_comment_id=10100116976844302", "anchor": "fb-10100116975352292_10100116976844302", "service": "fb", "text": "→ Michael I think it's an improvement over what we have now, but I also think that Condorcet-IRV is strictly better than IRV.", "timestamp": "1571493697"}, {"author": "opted out", "source_link": "#", "anchor": "unknown", "service": "unknown", "text": "→ this user has requested that their comments not be shown here", "timestamp": "1571494844"}, {"author": "Michael", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116975352292&reply_comment_id=10100116979713552", "anchor": "fb-10100116975352292_10100116979713552", "service": "fb", "text": "→ I'm not so sure I agree, Jeff. There's an implicit value judgment in saying the winner should be the \"most tolerated\" instead of \"most liked\". IRV moves the needle in the former direction, but your proposal would seem to be an extreme version. Is that better? I feel like it's subjective.", "timestamp": "1571495239"}, {"author": "Martin", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116975352292&reply_comment_id=10100116991410112", "anchor": "fb-10100116975352292_10100116991410112", "service": "fb", "text": "→ Elliot I have certainly felt more frustrated with ranked voting systems when there are more than 3-4 candidates. I do think that the \u201cbest\u201d voting system is probably different for selecting 8 city councilors out of a field of 20 candidates than for selecting one senator out of a field of 3.", "timestamp": "1571499836"}, {"author": "opted out", "source_link": "#", "anchor": "unknown", "service": "unknown", "text": "this user has requested that their comments not be shown here", "timestamp": "1571494138"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116977802382&reply_comment_id=10100116980147682", "anchor": "fb-10100116977802382_10100116980147682", "service": "fb", "text": "→ Elliot this is something party based proportional representation systems do well", "timestamp": "1571495368"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116977802382&reply_comment_id=10100117582425712", "anchor": "fb-10100116977802382_10100117582425712", "service": "fb", "text": "→ Yeah, I see this as a big strike against IRV. It assumes that voters have lots of nuanced preferences they want to express; but except in rare cases, I'm not sure we do!

This is also why I'm with Jeff in liking approval voting; saying \"Here's a list of candidates I like\" seems closer to the actual structure of voter psychology (or my psychology, at least).

(I find this line of thinking more persuasive than the pathological examples in the post. Arrow's Theorem means that every system will sometimes give objectionable outcomes, and IRV strikes me as no worse or better on that front than most other systems.)", "timestamp": "1571771739"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116977802382&reply_comment_id=10100117583104352", "anchor": "fb-10100116977802382_10100117583104352", "service": "fb", "text": "→ Ben \"Arrow's Theorem means that every system will sometimes give objectionable outcomes, and IRV strikes me as no worse or better on that front than most other systems.\"

To compare the various non-ideal systems you can use voter satisfaction efficiency: http://electionscience.github.io/vse-sim/VSE/

IRV does especially poorly there.", "timestamp": "1571772020"}, {"author": "opted out", "source_link": "#", "anchor": "unknown", "service": "unknown", "text": "→ this user has requested that their comments not be shown here", "timestamp": "1571772539"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116977802382&reply_comment_id=10100117642914492", "anchor": "fb-10100116977802382_10100117642914492", "service": "fb", "text": "→ Interesting - sounds like I have homework to do on VSE and the suite of theorems in social choice theory... but I may skim that homework, since it sounds like I'll wind up landing in the same place I am now (and where Elliott is), on the idea that the important constraints are about implementation, not theory.", "timestamp": "1571783683"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116977802382&reply_comment_id=10100117645958392", "anchor": "fb-10100116977802382_10100117645958392", "service": "fb", "text": "→ Ben VSE is somewhere between theory and practice, since it's based on simulations

The main thing I get out of VSE is that Approval is likely to work well, and we should be running more things with it to gain more experience", "timestamp": "1571784524"}, {"author": "Jai", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116977802382&reply_comment_id=10100117752355172", "anchor": "fb-10100116977802382_10100117752355172", "service": "fb", "text": "→ This is why Single Transferrable Vote is my favorite system. Vote for someone who you trust you represent your values well whether by winning or forwarding your vote to a similar-but--more-viable candidate. Uninformed voters can just vote for one of the major candidates; high information/issue focused/niche voters can vote for whoever they want.

STV + proportional representation is my dream system. For N seats with V votes: on each round, you get a seat if you have at least V/N votes. The winner with the most excess votes (number of votes they have - V/N) forwards their excess votes to candidate(s) of their choice who haven't won yet; if no candidate has an excess, the candidate with the fewest votes is eliminated and must allocate their votes to one or more of the other remaining candidates. Repeat until only N candidates remain.

This doesn't work as well for one-winner-takes-all elections, but for electing bodies it combines simplicity for voters with freedom to vote for whoever you want and meaningfully impact the result.

To adapt this for a single winner system, you can just make it so you're picking the N representatives who will hold an approval vote to select the winner. It's a little convoluted, but I think it still does a pretty good job of reflecting voters' preferences with minimal work required from said voters.", "timestamp": "1571854631"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116977802382&reply_comment_id=10100117755768332", "anchor": "fb-10100116977802382_10100117755768332", "service": "fb", "text": "→ Jai I'm not sure that's what STV is? STV is IRV generalized to multiple representation", "timestamp": "1571856108"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116977802382&reply_comment_id=10100117756127612", "anchor": "fb-10100116977802382_10100117756127612", "service": "fb", "text": "→ https://en.wikipedia.org/wiki/Single_transferable_vote", "timestamp": "1571856328"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116977802382&reply_comment_id=10100117756297272", "anchor": "fb-10100116977802382_10100117756297272", "service": "fb", "text": "→ \"whether by winning or forwarding your vote to a similar-but--more-viable candidate\" sounds like an interesting idea, but isn't STV?", "timestamp": "1571856397"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116977802382&reply_comment_id=10100117756446972", "anchor": "fb-10100116977802382_10100117756446972", "service": "fb", "text": "→ Do you mean Asset Voting?", "timestamp": "1571856545"}, {"author": "Jai", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116977802382&reply_comment_id=10100117756950962", "anchor": "fb-10100116977802382_10100117756950962", "service": "fb", "text": "→ Jeff Kaufman Yes I do! I had my terminology wrong (and apparently have had it wrong for years!) Asset voting is what I'm talking about. Thank you for the correction!", "timestamp": "1571856843"}, {"author": "Ronny", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116979344292", "anchor": "fb-10100116979344292", "service": "fb", "text": "VCG or bust", "timestamp": "1571494976"}, {"author": "opted out", "source_link": "#", "anchor": "unknown", "service": "unknown", "text": "→ this user has requested that their comments not be shown here", "timestamp": "1571531038"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116979344292&reply_comment_id=10100117076419752", "anchor": "fb-10100116979344292_10100117076419752", "service": "fb", "text": "→ Perhaps you just use VCG to auction off the seat?

(Doesn't sound like a good idea to me!)", "timestamp": "1571533722"}, {"author": "Martin", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116981664642", "anchor": "fb-10100116981664642", "service": "fb", "text": "I will preface this by saying that I haven\u2019t done enough research or thinking on Condorcet-IRV or other alternatives, but I am curious to hear your thoughts about the following. In making ranked choices, people often have non-uniform strength of preference. For example, in a 6-way race I might have a strong opinion about my first and second choice but be largely indifferent about 4 other candidates. This non-uniformity is especially likely to happen in a very partisan environment if a 3 person race where the third party candidate is more closely aligned with one party. In other words, rank order is an imperfect measure of the weighted discrete distribution that reflects my preferences because it does not capture distance between rankings. In an extreme case I might want to place zero weight on a candidate. How do you think non-uniform preferences should be considered in constructing a voting system?", "timestamp": "1571495911"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116981664642&reply_comment_id=10100116982942082", "anchor": "fb-10100116981664642_10100116982942082", "service": "fb", "text": "→ Martin people have made systems that try to handle this (https://en.m.wikipedia.org/wiki/Cardinal_voting, not to be confused with the https://en.m.wikipedia.org/wiki/Papal_conclave) but they are very open to tactical voting", "timestamp": "1571496249"}, {"author": "Martin", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116981664642&reply_comment_id=10100116991240452", "anchor": "fb-10100116981664642_10100116991240452", "service": "fb", "text": "→ Jeff Kaufman I guess I was imagining something like Evaluative Voting. What makes cardinal systems more open to tactical voting?", "timestamp": "1571499705"}, {"author": "opted out", "source_link": "#", "anchor": "unknown", "service": "unknown", "text": "→ this user has requested that their comments not be shown here", "timestamp": "1571501649"}, {"author": "Martin", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116981664642&reply_comment_id=10100117012627592", "anchor": "fb-10100116981664642_10100117012627592", "service": "fb", "text": "→ Elliot isn\u2019t that equally true with an ordinal system?", "timestamp": "1571507023"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116981664642&reply_comment_id=10100117027378032", "anchor": "fb-10100116981664642_10100117027378032", "service": "fb", "text": "→ Martin I think the claim is something like \"all the expressive power you'd get with cardinal voting is lost to tactics, and it would turn into something more like first past the post\"", "timestamp": "1571512518"}, {"author": "opted out", "source_link": "#", "anchor": "unknown", "service": "unknown", "text": "→ this user has requested that their comments not be shown here", "timestamp": "1571529676"}, {"author": "Martin", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116981664642&reply_comment_id=10100117072607392", "anchor": "fb-10100116981664642_10100117072607392", "service": "fb", "text": "→ Elliot fair enough", "timestamp": "1571532588"}, {"author": "opted out", "source_link": "#", "anchor": "unknown", "service": "unknown", "text": "this user has requested that their comments not be shown here", "timestamp": "1571501198"}, {"author": "Jacob", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116995581752", "anchor": "fb-10100116995581752", "service": "fb", "text": "I've so far only voted in one IRV election, the SF mayoral election last year. The city elections office published the results of the IRV runs on a daily basis (with votes coming in late because vote-by-mail is open to everyone). I understand how IRV works and followed along, and I was mystified. Even the experts, as embodied by prediction markets, weren't sure who was going to win for a week and more. And there were only three candidates with any significant chance of winning!

Beyond the cognitive burden concerns Elliot mentioned (which I agree with), I am very down on IRV for this reason, which I think can be glossed as \"transparency\". It's hard for even a dedicated layman to understand the results, which seriously damages public faith in the legitimacy of elections. FPTP has serious flaws, but it's legible to nearly everyone. And Condorcet-IRV is worse on this axis, not better.", "timestamp": "1571502526"}, {"author": "Jacob", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100116995581752&reply_comment_id=10100116995965982", "anchor": "fb-10100116995581752_10100116995965982", "service": "fb", "text": "→ In general, I don't think any election system should be trusted until it's been used in highly-partisan, high-stakes elections and we see how much tactical voting is used, and of what kinds. But my experience in SF has made me tentatively think that voting reform advocacy should strongly favor simple but tactics-able systems like approval voting over complex but more honest ones like IRV.", "timestamp": "1571502696"}, {"author": "Evan Rysdam", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#ZPNkWjeY9gqaRqNk8", "anchor": "lw-ZPNkWjeY9gqaRqNk8", "service": "lw", "text": "About 2/3 of people prefer Kasich to any other candidate on offer [...]

I think this is misleadingly phrased. It's true that Kasich wins by 2/3 to 1/3 no matter which other candicate you pit him against, but it's not true that 2/3 of people prefer him to any other candidate on offer. Only 14% + 17% = 1/3 of people have him as their first choice.

Your thesis stands, though, and I've updated on it.", "timestamp": 1571522834}, {"author": "Clay S", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#zRJ449d4Ry4LpoXbr", "anchor": "lw-zRJ449d4Ry4LpoXbr", "service": "lw", "text": "→ A 2/3 majority prefer Kasich to any other candidate. The point is that Independence of Irrelevant Alternatives dictates that if a group prefers X to Y, and prefers X to Z, that they logically must prefer X if all three are options. It's like if I ask you to choose between chocolate and vanilla and you pick chocolate; if I tell you strawberry is also an option, that shouldn't make you switch to vanilla.", "timestamp": 1571551699}, {"author": "Evan Rysdam", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#d39HhMZHLTtaDFMQw", "anchor": "lw-d39HhMZHLTtaDFMQw", "service": "lw", "text": "→ → if a group prefers X to Y, and prefers X to Z

There's no group that prefers Kasich to Trump and also prefers Kasich to Clinton. It's 2/3 in each case, but those two groups of 2/3 only have an overlap of 1/3.

I'm not familiar with voting theory, so I might be missing the point, but the sentence "there exists a 2/3 majority of the voting population all of whom prefer Kasich to any other candidate" is false. (The problem might be the ambiguity of the English language: it is true that "for any candidate besides Kasich, there exists a 2/3 majority who prefers Kasich to that candidate".)", "timestamp": 1571552940}, {"author": "Clay S", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#cjWdp7KPNv7itnJ4A", "anchor": "lw-cjWdp7KPNv7itnJ4A", "service": "lw", "text": "→ → → \n

There's no group that prefers Kasich to Trump and also prefers Kasich to Clinton.\n\n

That is irrelevant.\n", "timestamp": 1571569777}, {"author": "jkaufman", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#wQC7FksSdoJpitiPc", "anchor": "lw-wQC7FksSdoJpitiPc", "service": "lw", "text": "→ The statement is ambiguous, and depends on whether you're binding \"2/3 of people\" or \"any other candidate\" to a value first. It can mean either the intended:\nfor each other candidate on offer:\n about 2/3 of people prefer Kasich to that candidate\n\n

Or:\nthere exists a group of about 2/3 of people that:\n for each other candidate on offer:\n everyone in the group prefers Kasich to that candidate\n\n

I think the ambiguity is clear in context, however, since the latter is so clearly false.\n", "timestamp": 1571571846}, {"author": "Pattern", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#fk5abZ5ND6QwgxHWm", "anchor": "lw-fk5abZ5ND6QwgxHWm", "service": "lw", "text": "How common are Condorcet winners?", "timestamp": 1571540148}, {"author": "Clay S", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#7ybjZiD7KTKTkQ4C9", "anchor": "lw-7ybjZiD7KTKTkQ4C9", "service": "lw", "text": "→ See table 1.

http://scorevoting.net/RandElect", "timestamp": 1571551471}, {"author": "Clay S", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#szE88WHCXybrPcJnw", "anchor": "lw-szE88WHCXybrPcJnw", "service": "lw", "text": "every voting system has downsides

You linked to an article about Arrow's Theorem, which only applies to ordinal (ranked) voting methods, not cardinal (rated) voting methods like Score Voting and Approval Voting.

In any case, it's worth noting that Approval Voting was adopted in Fargo, ND in November 2018, and will be used in June 2020.

There will be a 2020 ballot initiative in St. Louis to convert their current partisan March primary + April general with a non-partisan March open primary with Approval Voting, followed by an April top-two.

There will be twin 2020 ballot initiatives to get STAR Voting in Eugene, OR as well as in its surrounding Lane County, OR.

Cardinal voting methods will win in the long run.
", "timestamp": 1571552494}, {"author": "Vladimir_Nesov", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#GKpquBGeyF6fB2Px4", "anchor": "lw-GKpquBGeyF6fB2Px4", "service": "lw", "text": "→ \n

Cardinal voting methods will win in the long run.\n\n

(What kind of long run? Why is this to be expected?) Popularity is not based on only merit, being more complicated than the simplest most familiar method sounds like a near-fatal disadvantage. Voting being involved with politics makes it even harder for good arguments to influence what actually happens.\n", "timestamp": 1571552073}, {"author": "Clay S", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#ygo22hbvade7FnD6E", "anchor": "lw-ygo22hbvade7FnD6E", "service": "lw", "text": "→ → Australia has been using a much more complicated ranked system since 1918, and Ireland has used an even more complicated weighted proportional system. The entire state of Maine adopted IRV, and cardinal systems are much simpler. It's not a fatal disadvantage.\n

http://scorevoting.net/Complexity\n

The Approval Voting system is arguably simpler than the status quo, because you remove a rule. The one that stays your vote is invalid if you vote for multiple candidates.\n", "timestamp": 1571554278}, {"author": "Vladimir_Nesov", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#he4antF7zDrrmyYkz", "anchor": "lw-he4antF7zDrrmyYkz", "service": "lw", "text": "→ → → Maybe \"near-fatal\" is too strong a word, the comment I replied to also had examples. Existence of examples doesn't distinguish winning from survival, seeing some use. I understand the statement I replied to as meaning something like \"In 200 years, if the world remains mostly as we know it, the probability that most elections use cardinal voting methods is above 50%\". This seems implausible to me for the reasons I listed, hence the question about what you actually meant, perhaps my interpretation of the statement is not what you intended. (Is \"long run\" something like 200 years? Is \"winning\" something like \"most elections of some kind use cardinal voting methods\"?)\n", "timestamp": 1571582680}, {"author": "jkaufman", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#HxrEFHBXpWatwAZvk", "anchor": "lw-HxrEFHBXpWatwAZvk", "service": "lw", "text": "→ \n\n

every voting system has downsides\n\n

You linked to an article about Arrow's Theorem, which only applies to ordinal (ranked) voting methods\n\n

While Arrow's Theorem applies only to ordinal voting methods, other voting methods still have downsides, primarily that you need to be tactical: voting honestly won't always give you your best result.\n\n

cardinal (rated) voting methods like Score Voting and Approval Voting\n\n

Why are you classifying Approval Voting as cardinal?\n\n

Cardinal voting methods will win in the long run.\n\n

Why do you think this? You can think of there as being a range of complexity in terms of how much information is elicited from voters: single top preference (FPTP), all acceptable candidates (Approval), candidates in order (Ranked), candidates with ratings (Range). Where we end up on this seems like more a question of what ends up working well in practice than something we can answer from thinking about humans. I like Approval a lot, as I said in the post, but after seeing more hotly contested Approval elections I might understand new aspects of how it works in practice that would change my mind.\n", "timestamp": 1571571641}, {"author": "Vaniver", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#tyNi7MMLX5cxJwWQW", "anchor": "lw-tyNi7MMLX5cxJwWQW", "service": "lw", "text": "→ → Why are you classifying Approval Voting as cardinal?

Ordinal voting systems force you to express a strict preference between all candidates; cardinal voting systems allow you to have candidates tie. For any election with at least three candidates, there has to be a tie if you're using approval voting.", "timestamp": 1571707251}, {"author": "jkaufman", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#r8Yva7YLYJFk4CgPN", "anchor": "lw-r8Yva7YLYJFk4CgPN", "service": "lw", "text": "→ → → Defining \"cardinal\" as \"allows you to rank two candidates as equal\" is a very weird use of the term, but it does appear to be standard: https://en.wikipedia.org/wiki/Cardinal_voting\n", "timestamp": 1571745880}, {"author": "MakoYass", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#iwaBHLaBKTdwfX8KB", "anchor": "lw-iwaBHLaBKTdwfX8KB", "service": "lw", "text": "→ Is utilitarianism an ordinal voting system?", "timestamp": 1571628107}, {"author": "jkaufman", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#fnkwH5kdEK3DSL2Qg", "anchor": "lw-fnkwH5kdEK3DSL2Qg", "service": "lw", "text": "→ → If you could get people to honestly report the change in utility they would experience, then you could choose the things that would give the largest utility increases. But since you can't assume honesty it doesn't work.\n", "timestamp": 1571668321}, {"author": "MakoYass", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#cm9dD2M5niEzGL4rN", "anchor": "lw-cm9dD2M5niEzGL4rN", "service": "lw", "text": "→ → → So you agree that it's a voting system.

I don't think it's intuitive that "give me a full account of all of your desires" wont end up working better than "give me an extremely partial picture of your desires"", "timestamp": 1571709677}, {"author": "philh", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#ZrkyyokPgyAWSKhHJ", "anchor": "lw-ZrkyyokPgyAWSKhHJ", "service": "lw", "text": "Further reading on this: a voting theory primer for rationalists, which in particular mentions that RCV has lower \"voter satisfaction efficiency\" than just about any proposed alternative except first-past-the-post. (Approval does indeed do better, and the author supports it as a first step.)\n", "timestamp": 1571563476}, {"author": "Ravi D'Elia", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#oKu5dcwBG59S8Puo4", "anchor": "lw-oKu5dcwBG59S8Puo4", "service": "lw", "text": "Am I correct in thinking that these problems are more obvious when using ranked-choice to elect a single candidate? When electing a number of people, it would be much better than just picking in order of number of votes.", "timestamp": 1571593136}, {"author": "jkaufman", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#5mAynzkGkELTgSj5Y", "anchor": "lw-5mAynzkGkELTgSj5Y", "service": "lw", "text": "→ This is all focused on single-winner elections, which is the main kind we have in the US.\n

The standard way elections for multiple slots are handled in the US is you vote for up to N people, and the N with the most votes win. Switching to approval voting would be as simple as changing \"vote for up to N\" to \"vote for any number\".\n

(There's also generalization of Instant Runoff called Single Transferrable Vote, which Cambridge MA uses for City Council)\n", "timestamp": 1571596293}, {"author": "jkaufman", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#baMZYEZgQoxzHALRg", "anchor": "lw-baMZYEZgQoxzHALRg", "service": "lw", "text": "→ → Here's an example \"vote for up to 4\" ballot, for Boston City Council: https://www.cityofboston.gov/images_documents/Specimen Ballots for Nov 3 2015_tcm3-52198.pdf\n", "timestamp": 1571597095}, {"author": "Marcus", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100117217397232", "anchor": "fb-10100117217397232", "service": "fb", "text": "You don't need to make up an example. The 2009 Burlington Vermont mayoral election shows how Concordant voting would have worked if it had been implemented.

This election was controversial since the instant runoff winner, who was from the Progressive Party, came in second in the first round. The Republicans managed to convince the voters to repeal the instant runoff and go back to two round voting. (The argument they made was that the number of exhausted ballots was greater than the final margin of victory, essentially disenfranchising voters who didn't express a preference.)

Imagine instead the Democrat had won. He got less than a quarter of the first round vote. You really think everyone would have suddenly hailed it as great outcome? Both sides would have been pissed off. I have no problem eliminating a candidate who has so little support. Electing them would be extremely counterintuitive, far more so than a runoff, and would cause voters to lose faith in their government.", "timestamp": "1571607501"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100117217397232&reply_comment_id=10100117240885162", "anchor": "fb-10100117217397232_10100117240885162", "service": "fb", "text": "→ Marcus I strongly disagree. We want to get to a place where people are often putting candidates they love first even when those candidates are unlikely to win. In that world \"it's important that the winning candidate have a large number of first place votes\" is clearly no longer an important consideration. No president I've ever voted for has been my number one choice for the position out of all over-35 natural born citizens.

(I reference the 2009 election later in the post.)", "timestamp": "1571618354"}, {"author": "Marcus", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100117217397232&reply_comment_id=10100117330984602", "anchor": "fb-10100117217397232_10100117330984602", "service": "fb", "text": "→ Literally no one thought IRV was bad in that election because the Democrat lost. They thought it was bad because it led to undervoting and, honestly, because their guy lost a close election. This is an issue with any ranked choice system. Burlington went back to having a two round system. The issue was with the \"instant\" part of IRV, not the \"runoff\".", "timestamp": "1571652320"}, {"author": "Marcus", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100117217397232&reply_comment_id=10100117334846862", "anchor": "fb-10100117217397232_10100117334846862", "service": "fb", "text": "→ I do think you could potentially implement a Concordant system that worked, but you would need to explain it in a way that made intuitive sense to voters. Maybe in terms of vote switching? To use the Burlington example again: \"The Democrat won because the Republicans switched their votes to him when it was clear their guy would lose.\"", "timestamp": "1571655733"}, {"author": "Jesse", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100117301663362", "anchor": "fb-10100117301663362", "service": "fb", "text": "Let's start with Arrow & note that no perfect multi-candidate voting system exists.
https://en.wikipedia.org/.../Arrow%27s_impossibility_theorem

Comparing Schulze (a case of Condorcet-IRV) to pure IRV:
- \"Transparency\" to general public is similar: moderately complex analysis is required to determine (or verify) results.
... as such I'm not clear that either Schulze or IRV is resilient to campaigns against it (especially following close elections).

Per https://en.wikipedia.org/.../Comparison_of_electoral_systems
Criteria met by IRV & Failed by Schulze:
Later-no-harm (adding lower-ranked candidates to voter's ballot will not cause higher-ranked candidates to lose)
Later-no-help (adding lower-ranked candidates to voter's ballot will not cause a higher-ranked candidate to win)

Criteria met by Schulze & Failed by IRV
Condorcet (if exists a winner who is directly preferred to all other candidates, they win)
Smith/ISDA (broader Condorcet, winner always in set of candidates who are majority-preferred directly to all other candidates)

Having been in (college organization) elections where tactical voting was openly practiced, I happen to think avoiding the risk of manipulation is worth something.

Specifically, I'd be curious in the following:
1) Assuming 1000 randomly-distributed ranked votes between three candidates, how often does a Condorcet winner exist? How often is that winner not chosen by IRV? I'd like to quantify the \"harm\" here.
2) Given the same output, how often could the voters preferring the lowest-ranked candidate deny victory to the Condorcet winner if they uniformly changed votes to put the (otherwise Condorcet winner) at the end of their ballot?
3) Under different scenarios, what happens if a majority or all voters attempt tactical voting in their lower choices? How frequently does this flip the outcome to those voters' first choice?", "timestamp": "1571622417"}, {"author": "Linchuan", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100117301663362&reply_comment_id=10100117587575392", "anchor": "fb-10100117301663362_10100117587575392", "service": "fb", "text": "→ Jesse Arrow only applies to ordered preferences without cardinality right?", "timestamp": "1571773798"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100117301663362&reply_comment_id=10100117612924592", "anchor": "fb-10100117301663362_10100117612924592", "service": "fb", "text": "→ Jesse: have a look at http://electionscience.github.io/vse-sim/VSE/ for another way of approaching the \"they're all imperfect but what is lease imperfect\" question", "timestamp": "1571774876"}, {"author": "Jesse", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100117301663362&reply_comment_id=10100117687594952", "anchor": "fb-10100117301663362_10100117687594952", "service": "fb", "text": "→ Linchuan - It's been a while since my polisci courses, so I'll defer to Wikipedia here - there's an extension of Arrow for cardinal rankings as well (Gibbard's theorem). Any choice mechanism is dictatorial, two-option, or strategic.

Jeff - thanks! That's a good first pass at getting more data here. Taking the site's models and population for granted:
- IRV on all six scenarios outscores all but two of Plurality (50%/100% strategic).
- IRV has a lower cap on total score than most Condorcet or cardinal systems
- IRV is less negatively affected than most other models by 1-sided strategic voting.

The model winds up with roughly these outcomes:
a) Easy (plurality = Condorcet) = ~46%
b) Condorcet cycle (indeterminate to Condorcet) = ~2%
c) Spoiler (Condorcet wins among top 3) = ~9%
d) Center-squeeze (Condorcet loses among top 3) = ~9%
e) Chicken-dilemma (Condorcet-3rd wins among top 3) = ~7%
d) Other ~25%

I'm all in favor of developing, testing, and trying out better models - but I think for now I have no issue with IRV as a realistic road-tested alternative to plurality (and starting to test Condorcet or cardinal systems like IRV has been tested)

Other notes:
- Plurality's best scores are dual-strategic (50%/100%). All other plurality scores are below the lowest IRV, Condorcet, or cardinal option.
- Four options (Ranked Pairs, Star2/10 & 3-2-1) score better than best plurality.under all scenarios.
- Only 3-2-1 scores better than best IRV under all scenarios (Star0to10 comes close)
- Condorcet (RP, Schulze) score highest on fully honest votes and are dragged down hard by one-sided strategic voting

... if I had to pick based on this info I'd try Star0to10 (highest max outside Condorcet, higher floor than Condorcet) or 3-2-1 (highest floor, tight grouping across scenarios).

Curious to see how Condorcet-IRV/WoodSIRV would perform in this lineup. These were the explanation pieces I found for it:
https://www.daneckam.com/?p=374
http://scorevoting.net/SmithIRV.html

Might see if I can get their code working and check my (and their) assumptions... but it's somewhere to start.

While I'm trying not to give too much weight to dedicated IRV activists in my thinking on this, at the end here I did read one FairVote piece, and it makes a point I'd like to emphasize:
http://archive3.fairvote.org/.../why-i-prefer-irv-to.../
- All our discussions were for single-seat elections
- Multi-member or proportional districts may have entirely different benefits (guaranteeing representation to minority groups)
- Single Transferrable Vote is a proportional voting mechanism which shares most properties with IRV (though resolution may differ). That could make it easier to transition to multi-member districts.", "timestamp": "1571802368"}, {"author": "Dagon", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#cKCLfqZPthDcwFiGP", "anchor": "lw-cKCLfqZPthDcwFiGP", "service": "lw", "text": "Most problems and edge-cases in a voting (or preference aggregation or utility aggregation) mechanism are dwarfed by the fundamental incorrect assumption of equal weight per person for any decision. In reality, there are orders of magnitude difference in how much a given choice will impact different people, and in how strongly different people care about their choices.

Once you've decided to ignore that, you're well into the world where legitimacy (acceptance of the result with minimal disruption) is your goal, not optimization of preferences. And for this, there is a very strong advantage to the status quo.", "timestamp": 1571688837}, {"author": "jkaufman", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#7tgYBNi2rdYGbxJ79", "anchor": "lw-7tgYBNi2rdYGbxJ79", "service": "lw", "text": "→ Can you say some about how you think this should work?\n", "timestamp": 1571705467}, {"author": "Dagon", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#dhaFjRFWoEKS6W3uZ", "anchor": "lw-dhaFjRFWoEKS6W3uZ", "service": "lw", "text": "→ → I'm not sure "should work" is a phrase I'd use for this topic (topic being "large-group decision-making"). I'm kind of shocked anything works at all, and I suspect it's mostly because voting isn't the main process for deciding anything important.

If legitimacy/acceptance is the main criterion for a voting system, the first things to "fix" are not the geeky "better aggregation of private preferences" mechanisms that we love so much. The problems in trust come from the ludicrous deviations that multi-level aggregation (electoral college and arbitrary geographic divisions) and indirection/bundling of issues (voting for people or parties rather than issues) cause/reveal.", "timestamp": 1571771643}, {"author": "Vaniver", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#Svdjd7GteqzAx3DCa", "anchor": "lw-Svdjd7GteqzAx3DCa", "service": "lw", "text": "I basically don't buy the Condorcet winner argument, mostly because the utility and disutility of winning or losing isn't fixed. This is one of the reasons why I like score voting (or range voting) so much; candidates who are massively disliked lose heavily, whereas candidates who are broadly liked win, and from the candidate's point for view, increasing your score in the eyes of anyone is useful, regardless of their score for other candidates.

Yes, there are concerns about comparing utilities across people, but people tend to be pretty reasonable about this in the score voting framework. (It's strategic to give your favorite a 10, and your least favorite a 0, but empirically people often compress their scores much more, say giving everyone a rating between 6 and 8, which implicitly makes their vote a fifth as strong.) The main problem is when you add candidates who make differences so large that it dwarfs all other variation (at least among voters who think they have tiered preferences). That is, suppose you have an "Anyone But Trump" voter; their vote that maximizes the chance of someone besides Trump winning is to give Trump a 0 and every other candidate a 10. But now whether Clinton or Kasich wins depends mostly on the people who thought Trump was ok (or it was worth putting Kasich in the "as bad as Trump" camp). This is probably fine for the rare election where there's a surprise Trump, and is not great if there's always a Trump-like candidate running.", "timestamp": 1571708242}, {"author": "jkaufman", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#B5bbH43WGsbZTmM37", "anchor": "lw-B5bbH43WGsbZTmM37", "service": "lw", "text": "→ \n

It's strategic to give your favorite a 10, and your least favorite a 0, but empirically people often compress their scores much more, say giving everyone a rating between 6 and 8, which implicitly makes their vote a fifth as strong\n\n

Can you say more about this? I would expect that it wouldn't take many competitive elections before most people would be submitting votes that are as strong as possible.\n", "timestamp": 1571745702}, {"author": "Vaniver", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#FPbKuTmojK8TEvriQ", "anchor": "lw-FPbKuTmojK8TEvriQ", "service": "lw", "text": "→ → I would expect that it wouldn't take many competitive elections before most people would be submitting votes that are as strong as possible.

That seems right; the surprising thing is that people are willing to 'undervote' at all, and that they do so deliberately instead of through ignorance. But it makes sense, especially for downballet candidates where you vaguely suspect parks commissioner candidate A is better than B but don't want to put your full force behind A (because if other people are well-informed on the parks commissioner race, you want them to settle it).

Beyond ignorance, altruism sometimes motivates this, which is easiest to see for simple things like picking what restaurant to go to. If you're mostly indifferent between them but have a weak preference, it makes sense to do a compressed vote if you suspect other people might have strong preferences that you don't want to overwhelm.", "timestamp": 1571845356}, {"author": "jmh", "source_link": "https://www.lesswrong.com/posts/9LKecLdEz3kjEmtKY#qMHjxmLExSe38xuXL", "anchor": "lw-qMHjxmLExSe38xuXL", "service": "lw", "text": "This will all be interesting to see play out and I do agree that we can improve the election process.

I do wonder just how well it will actually improve the end outcome: better government and better representatives. From my perspective one of the main problems is that representatives are simply not representative of any real majority of the population or exposed to any real incentives to pursue what might be call common/general good over the special interests and partisan policies. ", "timestamp": 1571752302}, {"author": "Christopher", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100118085377792", "anchor": "fb-10100118085377792", "service": "fb", "text": "Ranked Choice Voting is a component of Anti-corruption Act, which is likely why it's being promoted. See: https://represent.us/anticorruption-act/ and https://www.youtube.com/watch?v=TfQij4aQq1k", "timestamp": "1572025673"}, {"author": "Bob", "source_link": "https://www.facebook.com/jefftk/posts/10100116963925192?comment_id=10100120166686832", "anchor": "fb-10100120166686832", "service": "fb", "text": "https://xkcd.com/2225/", "timestamp": "1573071057"}]}