20 ms·

Ranking the voting systems: STAR Voting > Approval Voting > Ranked Choice Voting > Plurality ("pick only one") voting. Ranked Choice Voting is marginally bette

by bhauer 4y ago

Ranking the voting systems: STAR Voting > Approval Voting > Ranked Choice Voting > Plurality ("pick only one") voting. Ranked Choice Voting is marginally better than plurality voting, but it has problems. The chief defect with Ranked Choice Voting is its non-monotonicity, whereby increasing your support for your genuine favorite can actually hurt their odds of winning. This may be what happened in Alaska [1]. STAR Voting is a slight modification to Score Voting, where you simply score each candidate and are not forced to rank them. You are given the discretion to give multiple candidates the same score if you so choose. STAR is highly expressive and simple to count: just sum the scores. Approval Voting is appealing because of its simplicity. Both ballots and how they are counted would require only superficial changes versus plurality, such as changing the prompt from "Vote for only one" to "Vote for as many as you like." Approval has a good balance of utility and simplicity. If we are going to invest time and effort into achieving voting reform, it would be a shame to spend that effort on RCV rather than superior alternatives. [1] https://electionscience.org/commentary-analysis/rcv-fools-palin-voters-into-electing-a-progressive-democrat/ https://electionscience.org/commentary-analysis/rcv-fools-pa...

scythe 4y ago
Regarding the Score/Approval/STAR trio, I always felt that there was a clear fourth possibility missing from the list, so I invented it: https://pastebin.com/PCgSaBS8 https://pastebin.com/PCgSaBS8 Still, I'm not much of a political campaigner, so it hasn't gone very far.
- dane-pgp 4y ago
  That's a clever system, thanks for sharing it. What I appreciate most about it, though, is the explicit goal of "being easier to count by hand", which I think many advocates of voting-reform overlook. You even say "by sorting ballots into each of thirteen piles corresponding to the thirteen possible rankings of three candidates", which shows you've actually thought through what the process would look like for the people doing the counting. When people are proposing voting systems, they really need to consider that their process is being carried out in an adversarial environment, where different groups have incentives to cheat, or accuse the voting system of producing the "wrong" result (e.g. due to the unintuitive outcome, or the complex tactical thinking needed to optimally fill in the ballot), or claim that voting machines were running backdoored firmware (which is a destabilizing claim whether it is in fact true or false). That last point is perhaps the most important. I think that as time goes on, the pressure to automate the counting process and give instant election results (or worse, do the elections online) is only going to grow, and any voting reform which (if done manually) slows down the counting process, or for any other reason in practice requires people to trust machines that are selected and maintained by officials from one political party, is going to do even more harm to democracy than FPTP does.
buzzy_hacker 4y ago
STAR and Approval have different pros and cons, they are not objectively superior. I think RCV is better, in my opinion. STAR and Approval both have their own tactical voting problems. If you really like A, are meh about B, and dislike C, it’s clear you should rate A high / give A your approval and rate C low / do not give C your approval. But what about B? It has nothing to do with what you actually think about them, it has to do with what you think their odds of beating C and A are. You want B to do well enough to beat C, but not well enough to beat A. This adds too much distortion to people wanting to express their sincere preferences, which they can easily do in RCV: A > B > C And, more importantly, RCV is a stepping stone to multi-member district STV proportional voting (plug for https://www.fixourhouse.org/ https://www.fixourhouse.org/), which is the golden goose; way more important to implement proportional representation than squabble about STAR vs RCV, and RCV gets us closer to proportional representation.
- bombcar 4y ago
  This right here is the problem, these things get way too damn complicated for normal people really fast.
  - godelski 4y ago
    You don't think normal people can understand: "Rate this candidate from 0 to 5 stars?" If so, we really need to rethink a lot of how we ask for people's opinions, because this is probably the most common way.
    - bombcar 4y ago
      They can do that - but the game theory around "don't rate the candidate you don't want ..." stuff is way too complicated. And if it is not as simple as "Bob got the most votes, Bob won" it's going to get harder and harder to explain exactly what is going on, which is going to get people annoyed and angry. As long as the result is the same as would have come from first past the post, nobody will care.
      - 4y ago
lucasmullens 4y ago
RCV has made real ground, and halting its progress for STAR voting seems misguided. There's no consensus here, but you're stating the ranking like it's some objective fact. Imo, RCV's growing popularity only increases the popularity of other new voting systems (note that we're having a conversation about them, because someone used RCV).
kirse 4y ago
Yea, it's been interesting to watch how RCV has filtered into the public consciousness over the past 5-10 years and slowly increased in terms of how many people parrot it as the answer to all our voting ills. There was a site posted to HN years back that mathematically analyzed a ton of these voting schemes, I cannot seem to find it on Google but it was some sort of hybrid scheme that apparently has the best overall outcomes. Edit: Given some of the comments below, some good material on these voting systems... https://news.ycombinator.com/item?id=13161396 https://news.ycombinator.com/item?id=13161396 https://ncase.me/ballot/ https://ncase.me/ballot/ https://rangevoting.org/rangeVirv.html https://rangevoting.org/rangeVirv.html http://zesty.ca/voting/sim/ http://zesty.ca/voting/sim/
- theptip 4y ago
  IRV is a Schelling Point, and I think it's important to consider it as such. The biggest critique of alternatives to FPTP is "they are too complicated". When you look at an individual system like IRV it seems hard to sustain that objection, but then a bunch of voting system nerds enter the room, and now we're comparing Condorcet to Borda, and critiquing a bunch of mathematical properties of each system, and drawing graphs of preference-space; at which point the average voter's eyes gloss over and they say "yup, too complicated". I think this is a perfect case for the maxim "Let not the perfect be the enemy of the good". It seems IRV is sticking, and nobody has made Condorcet or Borda stick; let's go with IRV for now. Once we've abolished FPTP, and everyone is familiar with ranked systems, we can come back and argue for a better option. As far as I'm concerned, IRV captures 99% of the utility of changing voting systems. (All that said, I do like the work that https://electionscience.org https://electionscience.org is doing with Approval Voting, and if we can run experiments in local elections across the country with two new systems, I think that's also positive.)
  - ajmurmann 4y ago
    IRV is hard to evaluate though. How long would it take to teach a member of the general public to evaluate IRV results? You also cannot have meaningful estimates until all results from every district are in. How long it took in New York to get any results might leave people disillusioned and opposed to anything but FPTP. Using Borda and Condorcet is also clearly a straw man, as AFAIK nobody is actually pushing for those in practice. The real runners seem to be IRV, Approval and STAR. Of those IRV is clearly the most compared to evaluate. IMI they are all pretty simple and in an ideal world it wouldn't matter that much. However, we need to be highly aware that as soon as a major election gets won with anything but FPTP, we'll see the losing party "flood the zone with shit" to undermine the election results and the method. We've already seen the beginning of this after the election in Alaska. We need to keep it as simple as possible, otherwise the shit will drown everything else out.
fasthands9 4y ago
One thing that I think gets lost in political debate around voting is that RCV also changes the behavior/selection of the candidates. Its not just that the candidates are identical and you vote with a different mechanism - but candidates have an incentive to moderate. This may mean the outcome is the same (in terms of which candidate wins) but their positions represent the median voter more. I would imagine approval voting and other voting systems have this too (I'm not going to pretend to know how this would turn out for all of them) and I think that gets lost a bit when analyzing the systems.
- ajmurmann 4y ago
  This is also where it would be great to skip primaries, since primaries favor more extreme candidates.
- gamegoblin 4y ago
  This is actually not the case - RCV suffers from the "center squeeze" effect [1], in which moderates are squeezed out in the early rounds. That said, things like Approval have the opposite problem, in which moderates are given an advantage (some approval people will say this is a feature, not a bug). For an edge case, consider an election with 3 factions: socialists, libertarians, and fascists. There is also a 4th candidate, a boring status quo moderate, running. Imagine the populace is split 1/3 between the first 3 factions as their 1st choice, but everyone selects boring status quo moderates as their 2nd choice, because everyone in the first 3 factions hates the other ideologies. The boring status quo moderates get eliminated in round 1, and eventually 1 of the 3 extremes will win, resulting in 2/3 of the country being very angry. Not ideal. Obviously this is just an unrealistic edge case, but it demonstrates the effect. [1] https://electowiki.org/wiki/Center_squeeze https://electowiki.org/wiki/Center_squeeze
  - godelski 4y ago
    > That said, things like Approval have the opposite problem, in which moderates are given an advantage (some approval people will say this is a feature, not a bug). This is a weird way to say that representing the population is a bug. Your example election would be a compelling argument if population preference was more of a multimodal distribution rather than closer to a unimodal distribution. Despite the US's current division, it is nowhere near as divided as your scenario.
yellowapple 4y ago
It would be an even greater shame to be stuck with first-past-the-post while we squabble over its replacement. I vastly prefer score or approval voting over ranked-choice, but I also vastly prefer even ranked-choice over FPTP and will be voting accordingly here in Nevada. We can always adopt score or approval voting later.
- malnourish 4y ago
  Same. "Perfect is the enemy of good" as they say. I used to be vehemently opposed to RCV; I'm still against it. It's leagues better than FPTP and my hope is that it opens the door for improvements like STAR or anything that can satisfy the Condorcet criterion.
  - dllthomas 4y ago
    I'm not convinced that instant runoff is actually better than FPTP. Without tactical voting (and good information to support it) the first thing IRV does (along with most variations that in practice get the label "RCV", though of course this doesn't apply to all ranked methods) is to throw out the compromise candidate that everyone likes. In the current situation, I am concerned that most people will prefer their particular flavor of theocracy to finding ways to get along (which they probably prefer to anyone else's flavor of theocracy). We already struggle with similar dynamics a bit, but it's the devil we know and I really don't want it magnified.
    - manholio 4y ago
      Even if your technical argument would be correct (it's not), you would be trying to fix the problems of democracy by purposefully skewing the vote away from the desires of the electorate. That's a bad idea because it prevents democracy from working and exposing the leaders to the unfiltered democratic will of the people; so you get unaccountable leaders that can maintain power by gaming the system instead of catering for the interest of the public. Essentially, any power-mad dictator and authoritarian regime has some variant of this argument. Yes, we love democracy and voting, but we definitely cannot let those kinds of people vote or get elected. You are just espousing the softer version still accepted in some countries of "yes, the voting system is not ideal but it gives our side an advantage for now, realpolitik always trumps democratic principles".
godelski 4y ago
I want to say why the cardinal systems (STAR/Score/Approval) are usually better than the ordinal (Ranked/RP/Schulze). It is all about ,,embedding''. More efficient embedding systems mean you can get a higher information gain. Let's look at two different elections but with the same premise. Suppose you REALLY like both candidates A and B. Then suppose that you really dislike candidate C. Ranked: A > B > C (A/B position determined by flipping a coin) Ordinal: A:5 B:5 C:0 What's interesting here is that the first (ranked) system tells us (embeds) that our preference for A over B is the same as our preference of B over C. But in the second case we are able to tell the system that we actually like both A and B equally and that the distance to C is much larger. Ranked does not capture how we don't like C and doesn't capture that we do like B. Because of this the ranked systems frequently fail the favorite betrayer criteria (RCV fails btw). In a realistic scenario this means that in a ranked system Bernie would spoil the election for Biden (or vise verse). Although Gary Johnson wouldn't be a spoiler in either Ordinal nor Cardinal I'd argue that a weak spoiler is less of a worry than a strong spoiler. I will also argue that Ordinal methods are far simpler to calculate and do statistical verification on. Ranked systems require a while loop where we pop an element from our array each iteration and reallocate that element's votes until there is a singular winner. Cardinal systems on the other hand are simply an sum argmax. This sum argmax can be done in parallel and also makes it far easier to verify elections with random subsets of votes (a fairly common practice btw). The simplicity of the algorithm used to tally votes is extremely important and increases the transparency of the election. At the end of the day, the two systems are Ranked: "Put the candidates in order of preference" Cardinal: "Rate each candidate from 0 to 5 stars" People are very used to both types of questions. I see several arguments saying the latter is more complicated but this clearly isn't true or we'd have to change the vast majority of surveys. I also want to state that there are no globally optimal voting systems. Yes, there are plenty of flaws in cardinal systems. But because no global optima exists we have to instead argue that the tradeoffs are worth it. There are some objectively better systems than others though. To play fair like this I will say that I believe that the above advantages are worth the slight reduction of VSE in a 100% honest election that you'd obtain from using either RP or Schultz style elections.
crazygringo 4y ago
I just want to make clear for others: this "ranking" is just bhauer's opinion. Since some people here might interpret it as being presented as objective fact or some consensus opinion. There are a lot of pros/cons to all of the richer voting systems, usually categorized as approval voting, ranked voting, and score voting, in terms of increasing information. But there is absolutely no consensus that approval or score are better than ranked. (In my opinion, ranked comes out ahead, and I talk about why in another comment here.) But most of all, ranked choice is what has the growing momentum, as a lot of reform organizations have concluded it's the best practical solution. But what's most important is that we adopt one of them, and it would be a shame to spend our effort squabbling over which alternative, than giving support to the one that has the most momentum and chance of succeeding now. We're really only even having this conversation now because of the success reform groups have had in promoting ranked voting.
- DemocracyFTW2 4y ago
  There's no way that it would be a good idea to choose a system (ranked choice) that has more complex rules while having known defects over an alternative, simpler system (approval voting) that apparently (based on Wikipedia's short discussion) lacks those defects and has an overall favorable performance when it comes to resistance against tactical voting. FWIW reading the decision rules for ranked choice puts me off. Purely in theory, one would probably prefer a system where all voters have to rank all the candidates; in practice, however, all of voting, validating, counting, applying the rules, and understanding the rules is significantly more difficult than approval voting where a candidate's points is simply the number of times a voter has marked their name on the ballot. You can (generally or for more important decisions) amend approval voting with a quorum, minimum percentage of approval, or minimal distance between first and second places without damaging the system with too much complexity, something I cannot even fathom how to achieve with ranked choice.
- NickM 4y ago
  it would be a shame to spend our effort squabbling over which alternative, than giving support to the one that has the most momentum and chance of succeeding now I think the bigger shame would be to spend so much effort promoting RCV, only to have people become jaded and disillusioned with voting method reform in general when they realize RCVs many glaring problems. The biggest problem with RCV IMO is that it looks like it eliminates the spoiler effect, but it really only eliminates it in simple cases where the third-party candidates are not competitive. This suddenly becomes very obvious and painful in cases like Burlington, 2009 or Alaska, 2022 where voters were told "RCV is great because you can vote honestly", only to realize post-election that this was entirely false, and that they threw away their votes and let the election go to a candidate they despise. Trust is extremely important in political movements, and very easy to lose. I fear RCV is going to poison the well for any FPTP alternatives for a long time to come.
jasonlotito 4y ago
That's a very misleading article that plays loose with the math. e.g. It makes assumptions on things that are flat out not true (for example, assuming people would vote for thing they clearly didn't vote for). If an article is going to sit there and make up votes that didn't occur, I can't really trust it.
FireInsight 4y ago
https://star.vote/ https://star.vote/ is pretty good. I've used it for some votes for a small community with usually more than fifteen different options, and it's worked pretty well!
TimPC 4y ago
Isn’t giving each candidate a score even more gameable than ranking candidates. If I like candidate A and think candidate B is their best opposition I can give candidate A 5.0 and candidate B 0.0 even if my real opinion of candidate B is 3.0. Lowering candidate B’s score artificially definitely helps candidate A.
- avarun 4y ago
  I think you misread the comment you're replying to. Calculating a total score is simply summing the scores assigned to a candidate from each voter, so there's no way to "lower candidate B's score artificially". There's no averaging going on here.
- unholiness 4y ago
  Absolutely. Naive exaggeration is a fantastic way to make your vote count more than others in score voting. It also encourages more negative campaign tactics, since convincing a voter to rank your opponent 0 rather than 4 is literally 4 times as powerful as convincing them to rank you a 5 rather than a 4.
  - ClayShentrup 4y ago
    you have this exactly backwards, as explained by 3 different math phd experts here. http://www.youtube.com/watch?v=JtKAScORevQ http://www.youtube.com/watch?v=JtKAScORevQ https://www.rangevoting.org/NESD https://www.rangevoting.org/NESD https://electionscience.github.io/vse-sim/VSEbasic/ https://electionscience.github.io/vse-sim/VSEbasic/
- naniwaduni 4y ago
  This is a case for approval over score voting. It's not a good case for RCV over approval!
  - ClayShentrup 4y ago
    no, it's not a case for approval voting over score voting, because score voting is better than approval voting if even one voter votes sincerely. if enough people use honest scores, then score voting is better than approval voting even for the "honest suckers". https://www.rangevoting.org/ShExpRes https://www.rangevoting.org/ShExpRes
duxup 4y ago
I read the linked article and I found it kinda scatterbrained. It seemed to dispute claims or tweets about how a vote works, along with what head to head voting outcomes would be. I do find how the math works out interesting. But I don’t find either disputes all that compelling. The head to head polling is interesting… but that’s not how the election works. The fact that two systems would be different seems normal to me. Folks who go to a ranked choice vote and pick just one, that’s their choice too. Same with people who don’t vote and so on. Not to say the article is worthless, it is interesting, but I’m not sure I agree with the conclusions.
unholiness 4y ago
Every voting system has some flaws (see: Arrow's impossibility theorem), and Instant Runoff Voting definitely saw its worst-case scenario rear its ugly head in Alaska: A Condorcet winner existed (Nick Begich), he lost, and voters whose first choice was Plain and second choice was Begich could theoretically, retrospectivally, all changed their first votes from Palin to Begich to have him win rather than Mary Peltola. It's important that I say retrospectively. This was a house race with national coverage, tons of high-quality polling from pollsters with nothing better to do, and it was still absolutely not clear from that polling that Begich would do better than Palin in round 2, or that Palin was likely to beat Begich in round 1. If I had the opinions of those hypothetically strategic voters, and I were completely strategic, and completely tuned into that polling, I still would have voted honestly. The strategy was only apparent after the fact. Thinking about "defense in depth" against this scenario, we'd be remiss not to mention Condorcet ranked choice voting methods, which always elect a Condorcet winner, and only resort to (game-able) elimination procedures if there isn't one. Not only would that have resulted in a better outcome in Alaska, it would have made the possibility of strategic voting much, much more remote, both because a strategic scenario would be much less likely, and because that strategic scenario would be much less visible before the election. With score voting, strategies are always available and almost always obvious: just exaggerate your score differences between the most likely candidates. The fact that strategic dishonesty would have helped in Alaska is a fair criticism of IRV, but the real commonness of strategic scenarios in STAR and score are a more primary feature.
- garaetjjte 4y ago
  >see: Arrow's impossibility theorem Arrow's theorem only applies to ranked systems. That said, cardinal systems are still susceptible to strategic voting (Gibbard's theorem)
  - dane-pgp 4y ago
    Thank you for pointing out how overused "Arrow's impossibility theorem" is as an argument against all voting systems (and therefore often as an argument against all reforms to the current near-worst possible voting system). I would go further and be somewhat dismissive even of "Gibbard's theorem", since it only shows that an insincere vote can strategically produce a better outcome for a voter if they can identify what that that vote should be, and if they are one of the voters for whom such a strategy is available. So it doesn't guarantee (as far as I'm aware) that even a single voter (under every possible cardinal voting system) will necessarily have enough information before or during the election to confidently pick some specific insincere vote that will increase their chances of getting the outcome they want.
garaetjjte 4y ago
I'm not convinced that additional complexity of STAR is worth it compared to extremely simple score voting.
- obelos 4y ago
  As a practical consideration, the runoff in STAR does offer a legal advantage in some districts where a runoff is legally mandated.
- godelski 4y ago
  What's the added complexity of STAR? STAR is phase 1: sort(sum(candidates))[:2] (almost identical to score/approval/FPTP. We'd just take argmax instead of top two) and phase 2: eliminate other candidates, argmax(sum(candidates)) (identical to score/approval/FPTP). On the other hand RCV is a multi-round system with a while loop. In practice it almost never has fewer than 2 rounds (requires overwhelming majority, which almost never happens with more than 3 candidates). From an algorithms perspective RCV is more complex.
  - Wowfunhappy 4y ago
    > From an algorithms perspective RCV is more complex. Perhaps, but people aren't computers. It's admittedly difficult to score this objectively, but I find RCV much easier to understand.
    - godelski 4y ago
      Much easier to understand from the voter perspective of the tallier perspective? From the voter perspective I find them fairly equivalent and both are frequently used in the everyday world (scoring might be even more considering 5 star reviews). But from a tallying perspective I think it is very hard to argue that RCV is substantially easier. Again, the score/approval algorithm is identical to the FPTP algorithm (which is trivial). It is simply argmax(sum(candidates)). Just sum the totals in an excel file and pick the largest number. I don't know a single person that has a hard time understanding this and I suspect if they did they would have an even harder time understanding how to tally RCV.
wizofaus 4y ago
As far as I can tell RCV is equivalent to what we've been using in Australia since federation to elect lower house candidates (we call it prefential voting and use it at all levels of government). I'm curious in what case "increasing your support for your genuine favorite can actually hurt their odds of winning"? I skimmed the link to the article you referenced and it appears that there were two republican candidates which is the main reason neither won - I don't believe I've ever seen a case here where one party had two candidates on the same ballot paper and I suspect it's probably illegal. I've also never seen a ballot paper with more than about 15 choices (and it's very rarely more than 10), so vote splitting is rarely a problem (it can be for minor parties with similar policies/leanings etc.). On that basis I'm pretty sure there's never been a case in this country where anyone has lost an election due to gaining higher primary levels of support - indeed it's pretty rare that the winning candidate doesn't have at least 35% of the primary vote (though exactly that happened in our recent federal election, as a total across all seats), and exceedingly rare for it to be less than 30%. There are certainly plenty of cases where the winner isn't the one with the highest primary vote, but it's typically only a few percentage points lower, and mostly affects the more progressive candidate because for many seats there are two such candidates with solid levels of support but only one mainstream conservative candidate.
- tim-- 4y ago
  I don't believe that there is anything stopping two candidates from one party. I believe I have even seen it before, but I can't find a reference.
  - wizofaus 4y ago
    Doesn't seem compatible with the idea of "two-party preferred" voting which is what's actually used in practice here, though I'm not 100% clear as to whether that's explicitly different to preferential voting/RCV.
    - aldonius 4y ago
      Two party preferred is more a statistical count done on the assumption that {Labor, Coalition} = {Government, Opposition}. Of course it'll be decreasingly relevant now if the teals can stick.
JTbane 4y ago
STAR voting can be easily gamed- tell your supporters "vote MAX for me and 0 for everyone else". Anyone who gives more than a 0 to their second choice candidate will hurt their first choice.
- garaetjjte 4y ago
  Doing so would be ill-advised because it increases chances of winning for candidates that you actually dislike.
- notriddle 4y ago
  The only voting system that isn't "vulnerable" to bullet voting is FPTP, since it obligates you to bullet vote.
  - dane-pgp 4y ago
    Actually that's not the only invulnerable system; there's also "Asset Voting", which I've mentioned in a comment above[0]. It also obligates you to bullet vote, since the voter's honest first preference is the only information needed, but it avoids the spoiler problems of FPTP, and is thus superior to it. [0] https://news.ycombinator.com/item?id=33117670 https://news.ycombinator.com/item?id=33117670
  - crazygringo 4y ago
    Huh? Ranked voting isn't vulnerable to bullet voting. If you only rank a single candidate, there's zero benefit, you're just throwing away the rest of your vote. It's not giving that candidate any extra influence, the way it would in approval voting.
    - notriddle 4y ago
      > Ranked voting isn't vulnerable to bullet voting. "Vulnerable to bullet voting" is the same thing as "violating the later-no-harm principle," which means that some forms of ranked voting are vulnerable, while others aren't. - Borda Count is vulnerable to bullet voting, for the same reason as Score (giving points to anyone other than your first choice makes it less likely that your first choice wins, because it increases the score of your second and third choice). - Condorcet systems are vulnerable to bullet voting, because the ordering of your second and third choices can potentially change who the Condorcet winner is. - Instant Runoff is "immune" to bullet voting because the ranking of choices beyond the highest non-eliminated choice are ignored. This is also directly responsible for it being vulnerable to the spoiler effect, which makes non-bullet-voting pointless (your top choice must also be the choice most likely to win the 1-1 fight at the end, which means the remainder of your choices are highly unlikely to win, so why bother?).
barbariangrunge 4y ago
Ranked voting is supposed to prevent winner take all strategic voting. Voting by score encourages you to score everybody at zero except your favourite(s). This arguably means there will be less information in the vote tallies, although it gives voters more control over how their opinion will be responded to, and that is valuable
- ClayShentrup 4y ago
  this is obviously false. simply consider that green party supporters usually vote democrat. with score or approval voting, they'd ALSO want to vote for the green. whereas with ranked voting methods, they'd still strategically rank the democrat in 1st place, crushing any hope for the greens. http://www.youtube.com/watch?v=JtKAScORevQ http://www.youtube.com/watch?v=JtKAScORevQ
bfung 4y ago
There’s always going to be a trade-off when dealing with humans: voting result optimization vs ease of voting. That’s always been the issue, not that we don’t know of better result counting theories. Even in the 2020s, the ballots are explained over and over, “Connect the horizontal line”, or “fill in the box completely” — it’s a binary input to capture the best information from the lowest denominator. STAR would change the input from binary to an order N, and the number of mistakes or “that’s not what I intended” will happen. The average HN reader probably can fill in boxes pretty good, but the outside world not so much. RCV is a decent compromise.
SoftTalker 4y ago
A problem with all of these systems is that they are harder to understand, resulting in reduced voter confidence, or they demand more from the voter. Any kind of ranking is going to mean you need to know enough about each candidate to put them in some kind of order. Maybe OK if there are two candidates but much harder after that. Most people will rank their favorite on top and then just randomly rank the remainder. Simple plurality or majority vote is easiest to understand and easiest for the voter. You vote for the one candidate you like the best. If the isn't a majority for one and that is required, you have a runoff between the top two.
- Wowfunhappy 4y ago
  I agree, but I'd argue that instant-runoff/RCV is the second-easiest to understand, by far, while also being much better than simple plurality voting. Which is why I strongly favor instant-runoff.
  - dane-pgp 4y ago
    Actually there's another option which hasn't been mentioned so far, which is just as easy to implement as plurality (for both voting and counting by hand), and has pretty much all the benefits of IRV/RCV. That method is "Asset Voting", which works exactly like a plurality vote, except that once the totals for each candidate are known, the candidates are eliminated one by one (from least to most votes) and get to assign their vote counts to any remaining candidate. If you agree with me and the parent post that voter confidence (including avoiding electronic counting machines) is the most important factor when evaluating possible alternatives to plurality (assuming the alternatives are no worse than plurality) then hopefully you will agree that Asset Voting is preferable to IRV/RCV.
breuleux 4y ago
I don't think the voting system matters as much as we think. The main attack vector that interest groups use against democracy is shaping the public opinion in the direction that benefits them. Certainly, some systems may be easier to game, but ultimately interest groups will do neither more nor less than the minimum required to win. If they only need to make 30% of the population rabid in order to win, that's what they will do, but let's not fool ourselves into thinking they couldn't make 51% of the population rabid if they needed to. Basically, if people's stated preferences are ill-thought-out and/or manipulated, then even the perfect voting system that perfectly reflects the people's preferences will yield bad results. Voting reform in this case would be fixing the symptoms, not the disease. The disease, in my view, is that given the sheer complexity of the issues, the poor attention budget of people, the outsized influence of media, the impossibility of directly interacting with decision-makers, there's basically no way people can make an informed decision. Personally I think we should ditch elections and just pick representatives purely at random, and then pay them to think about their preferences full time, discuss them with each other, directly interview candidates for executive positions, and so on. The basic idea is that I think collective decision making probably reaches an optimum around Dunbar's number and falls off precipitously beyond that.
zajio1am 4y ago
Approval voting fails in a situation that is pretty common in Europe (e.g. French presidential elections) - you have center left, center right, far left and far right candidate. If you vote for center left/right, you have to decide whether to also approve center right/left candidate, which may be contrary to your values, corrupt and incompetent, but at least not communist or fascist. That is why later-no-harm property of IRV is important.
zajio1am 4y ago
> Ranking the voting systems: STAR Voting > Approval Voting > Ranked Choice Voting > Plurality ("pick only one") voting. Funny is that you ranked them and not scored them nor just say whether you approve them or not. If you have to vote for these voting systems using approval voting, which ones would you approve?
- ClayShentrup 4y ago
  scores would be more like: STAR or score or approval = 5 IRV/RCV = 2 plurality = 0 how i'd approve them in a real election would be based on electability, which would be true with any voting method used in an actual contentious election.
dragonwriter 4y ago
> Ranking the voting systems: STAR Voting > Approval Voting > Ranked Choice Voting > Plurality (“pick only one”) voting. I’m going to disagree. For essentially single-winner, public, candidate, secret ballot elections: Any Condorcet method* > Bucklin Voting* (without limited # of ranks) > Instant Runoff Voting* > Majority-Runoff** > Plurality Voting** > Approval***, Score***, or STAR*** Voting (or any other *** system) For public, secret ballot, legislative elections: Mixed Member Proportional (MMP) using PR-STV* + Party List for leveling) > MMP using any single-member district system + Party List for leveling > Single Transferrable Vote (PR-STV)* > Party List PR (optionally, using STV to minimize wasted votes, but vote-for-one is close behind) > single-member districts with single winner systems in the order above (but, compared to the preceding proportional systems, all are close enough as to make no real difference) > any system using some or all multimember constituencies (including staggered elections in the same district for multiple seats in the same body) not described earlier in this list. For public, secret ballot, chief executive + designated successor elections: The single winner methods, in the order above, modified to use ranked ballots if they don’t already, in two rounds where the first elects the chief executive, then that winner is eliminated from all ballots, then a second counting of the ballots is done to chose the designated successor > The single winner methods, in the order above, separately for each office > The single winner methods, i the order above, for slates of pairs of one candidate for each office > Plurality voting** in a two round system (with normal, vote-for-one ballots), where after a chief is chosen on the first round, the successor is chosen by eliminating all ballots for the prior winner and taking the highest remaining vote count. (methods with * are ranked ballots methods, methods with ** are vote-for-one methods commonly in use in the US, methods with *** are either artificially limited-ranks or more-detail-than-ranks methods, which for candidate elections have no clear, unique honest mapping from actual preferences to ballot markings; they are very good when their particular system of ballot markings maps to something concrete – e.g., for approval for selecting group activities in non-secret-ballot elections where approving is a commitment to participate if that option is selected or disapproval forfeits the right to participate, or, for score voting, where chosen score maps to a commitment of resources in the event the scored item is chosen.)

by bhauer 4y ago

scythe 4y ago

Regarding the Score/Approval/STAR trio, I always felt that there was a clear fourth possibility missing from the list, so I invented it: https://pastebin.com/PCgSaBS8 https://pastebin.com/PCgSaBS8 Still, I'm not much of a political campaigner, so it hasn't gone very far.

dane-pgp 4y ago
That's a clever system, thanks for sharing it. What I appreciate most about it, though, is the explicit goal of "being easier to count by hand", which I think many advocates of voting-reform overlook. You even say "by sorting ballots into each of thirteen piles corresponding to the thirteen possible rankings of three candidates", which shows you've actually thought through what the process would look like for the people doing the counting. When people are proposing voting systems, they really need to consider that their process is being carried out in an adversarial environment, where different groups have incentives to cheat, or accuse the voting system of producing the "wrong" result (e.g. due to the unintuitive outcome, or the complex tactical thinking needed to optimally fill in the ballot), or claim that voting machines were running backdoored firmware (which is a destabilizing claim whether it is in fact true or false). That last point is perhaps the most important. I think that as time goes on, the pressure to automate the counting process and give instant election results (or worse, do the elections online) is only going to grow, and any voting reform which (if done manually) slows down the counting process, or for any other reason in practice requires people to trust machines that are selected and maintained by officials from one political party, is going to do even more harm to democracy than FPTP does.

buzzy_hacker 4y ago

STAR and Approval have different pros and cons, they are not objectively superior. I think RCV is better, in my opinion. STAR and Approval both have their own tactical voting problems. If you really like A, are meh about B, and dislike C, it’s clear you should rate A high / give A your approval and rate C low / do not give C your approval. But what about B? It has nothing to do with what you actually think about them, it has to do with what you think their odds of beating C and A are. You want B to do well enough to beat C, but not well enough to beat A. This adds too much distortion to people wanting to express their sincere preferences, which they can easily do in RCV: A > B > C And, more importantly, RCV is a stepping stone to multi-member district STV proportional voting (plug for https://www.fixourhouse.org/ https://www.fixourhouse.org/), which is the golden goose; way more important to implement proportional representation than squabble about STAR vs RCV, and RCV gets us closer to proportional representation.

bombcar 4y ago
This right here is the problem, these things get way too damn complicated for normal people really fast.
- godelski 4y ago
  You don't think normal people can understand: "Rate this candidate from 0 to 5 stars?" If so, we really need to rethink a lot of how we ask for people's opinions, because this is probably the most common way.
  - bombcar 4y ago
    They can do that - but the game theory around "don't rate the candidate you don't want ..." stuff is way too complicated. And if it is not as simple as "Bob got the most votes, Bob won" it's going to get harder and harder to explain exactly what is going on, which is going to get people annoyed and angry. As long as the result is the same as would have come from first past the post, nobody will care.
    - 4y ago

lucasmullens 4y ago

RCV has made real ground, and halting its progress for STAR voting seems misguided. There's no consensus here, but you're stating the ranking like it's some objective fact. Imo, RCV's growing popularity only increases the popularity of other new voting systems (note that we're having a conversation about them, because someone used RCV).

kirse 4y ago

Yea, it's been interesting to watch how RCV has filtered into the public consciousness over the past 5-10 years and slowly increased in terms of how many people parrot it as the answer to all our voting ills. There was a site posted to HN years back that mathematically analyzed a ton of these voting schemes, I cannot seem to find it on Google but it was some sort of hybrid scheme that apparently has the best overall outcomes. Edit: Given some of the comments below, some good material on these voting systems... https://news.ycombinator.com/item?id=13161396 https://news.ycombinator.com/item?id=13161396 https://ncase.me/ballot/ https://ncase.me/ballot/ https://rangevoting.org/rangeVirv.html https://rangevoting.org/rangeVirv.html http://zesty.ca/voting/sim/ http://zesty.ca/voting/sim/

theptip 4y ago
IRV is a Schelling Point, and I think it's important to consider it as such. The biggest critique of alternatives to FPTP is "they are too complicated". When you look at an individual system like IRV it seems hard to sustain that objection, but then a bunch of voting system nerds enter the room, and now we're comparing Condorcet to Borda, and critiquing a bunch of mathematical properties of each system, and drawing graphs of preference-space; at which point the average voter's eyes gloss over and they say "yup, too complicated". I think this is a perfect case for the maxim "Let not the perfect be the enemy of the good". It seems IRV is sticking, and nobody has made Condorcet or Borda stick; let's go with IRV for now. Once we've abolished FPTP, and everyone is familiar with ranked systems, we can come back and argue for a better option. As far as I'm concerned, IRV captures 99% of the utility of changing voting systems. (All that said, I do like the work that https://electionscience.org https://electionscience.org is doing with Approval Voting, and if we can run experiments in local elections across the country with two new systems, I think that's also positive.)
- ajmurmann 4y ago
  IRV is hard to evaluate though. How long would it take to teach a member of the general public to evaluate IRV results? You also cannot have meaningful estimates until all results from every district are in. How long it took in New York to get any results might leave people disillusioned and opposed to anything but FPTP. Using Borda and Condorcet is also clearly a straw man, as AFAIK nobody is actually pushing for those in practice. The real runners seem to be IRV, Approval and STAR. Of those IRV is clearly the most compared to evaluate. IMI they are all pretty simple and in an ideal world it wouldn't matter that much. However, we need to be highly aware that as soon as a major election gets won with anything but FPTP, we'll see the losing party "flood the zone with shit" to undermine the election results and the method. We've already seen the beginning of this after the election in Alaska. We need to keep it as simple as possible, otherwise the shit will drown everything else out.

fasthands9 4y ago

One thing that I think gets lost in political debate around voting is that RCV also changes the behavior/selection of the candidates. Its not just that the candidates are identical and you vote with a different mechanism - but candidates have an incentive to moderate. This may mean the outcome is the same (in terms of which candidate wins) but their positions represent the median voter more. I would imagine approval voting and other voting systems have this too (I'm not going to pretend to know how this would turn out for all of them) and I think that gets lost a bit when analyzing the systems.

ajmurmann 4y ago
This is also where it would be great to skip primaries, since primaries favor more extreme candidates.
gamegoblin 4y ago
This is actually not the case - RCV suffers from the "center squeeze" effect [1], in which moderates are squeezed out in the early rounds. That said, things like Approval have the opposite problem, in which moderates are given an advantage (some approval people will say this is a feature, not a bug). For an edge case, consider an election with 3 factions: socialists, libertarians, and fascists. There is also a 4th candidate, a boring status quo moderate, running. Imagine the populace is split 1/3 between the first 3 factions as their 1st choice, but everyone selects boring status quo moderates as their 2nd choice, because everyone in the first 3 factions hates the other ideologies. The boring status quo moderates get eliminated in round 1, and eventually 1 of the 3 extremes will win, resulting in 2/3 of the country being very angry. Not ideal. Obviously this is just an unrealistic edge case, but it demonstrates the effect. [1] https://electowiki.org/wiki/Center_squeeze https://electowiki.org/wiki/Center_squeeze
- godelski 4y ago
  > That said, things like Approval have the opposite problem, in which moderates are given an advantage (some approval people will say this is a feature, not a bug). This is a weird way to say that representing the population is a bug. Your example election would be a compelling argument if population preference was more of a multimodal distribution rather than closer to a unimodal distribution. Despite the US's current division, it is nowhere near as divided as your scenario.

yellowapple 4y ago

It would be an even greater shame to be stuck with first-past-the-post while we squabble over its replacement. I vastly prefer score or approval voting over ranked-choice, but I also vastly prefer even ranked-choice over FPTP and will be voting accordingly here in Nevada. We can always adopt score or approval voting later.

malnourish 4y ago
Same. "Perfect is the enemy of good" as they say. I used to be vehemently opposed to RCV; I'm still against it. It's leagues better than FPTP and my hope is that it opens the door for improvements like STAR or anything that can satisfy the Condorcet criterion.
- dllthomas 4y ago
  I'm not convinced that instant runoff is actually better than FPTP. Without tactical voting (and good information to support it) the first thing IRV does (along with most variations that in practice get the label "RCV", though of course this doesn't apply to all ranked methods) is to throw out the compromise candidate that everyone likes. In the current situation, I am concerned that most people will prefer their particular flavor of theocracy to finding ways to get along (which they probably prefer to anyone else's flavor of theocracy). We already struggle with similar dynamics a bit, but it's the devil we know and I really don't want it magnified.
  - manholio 4y ago
    Even if your technical argument would be correct (it's not), you would be trying to fix the problems of democracy by purposefully skewing the vote away from the desires of the electorate. That's a bad idea because it prevents democracy from working and exposing the leaders to the unfiltered democratic will of the people; so you get unaccountable leaders that can maintain power by gaming the system instead of catering for the interest of the public. Essentially, any power-mad dictator and authoritarian regime has some variant of this argument. Yes, we love democracy and voting, but we definitely cannot let those kinds of people vote or get elected. You are just espousing the softer version still accepted in some countries of "yes, the voting system is not ideal but it gives our side an advantage for now, realpolitik always trumps democratic principles".

godelski 4y ago

I want to say why the cardinal systems (STAR/Score/Approval) are usually better than the ordinal (Ranked/RP/Schulze). It is all about ,,embedding''. More efficient embedding systems mean you can get a higher information gain. Let's look at two different elections but with the same premise. Suppose you REALLY like both candidates A and B. Then suppose that you really dislike candidate C. Ranked: A > B > C (A/B position determined by flipping a coin) Ordinal: A:5 B:5 C:0 What's interesting here is that the first (ranked) system tells us (embeds) that our preference for A over B is the same as our preference of B over C. But in the second case we are able to tell the system that we actually like both A and B equally and that the distance to C is much larger. Ranked does not capture how we don't like C and doesn't capture that we do like B. Because of this the ranked systems frequently fail the favorite betrayer criteria (RCV fails btw). In a realistic scenario this means that in a ranked system Bernie would spoil the election for Biden (or vise verse). Although Gary Johnson wouldn't be a spoiler in either Ordinal nor Cardinal I'd argue that a weak spoiler is less of a worry than a strong spoiler. I will also argue that Ordinal methods are far simpler to calculate and do statistical verification on. Ranked systems require a while loop where we pop an element from our array each iteration and reallocate that element's votes until there is a singular winner. Cardinal systems on the other hand are simply an sum argmax. This sum argmax can be done in parallel and also makes it far easier to verify elections with random subsets of votes (a fairly common practice btw). The simplicity of the algorithm used to tally votes is extremely important and increases the transparency of the election. At the end of the day, the two systems are Ranked: "Put the candidates in order of preference" Cardinal: "Rate each candidate from 0 to 5 stars" People are very used to both types of questions. I see several arguments saying the latter is more complicated but this clearly isn't true or we'd have to change the vast majority of surveys. I also want to state that there are no globally optimal voting systems. Yes, there are plenty of flaws in cardinal systems. But because no global optima exists we have to instead argue that the tradeoffs are worth it. There are some objectively better systems than others though. To play fair like this I will say that I believe that the above advantages are worth the slight reduction of VSE in a 100% honest election that you'd obtain from using either RP or Schultz style elections.

crazygringo 4y ago

I just want to make clear for others: this "ranking" is just bhauer's opinion. Since some people here might interpret it as being presented as objective fact or some consensus opinion. There are a lot of pros/cons to all of the richer voting systems, usually categorized as approval voting, ranked voting, and score voting, in terms of increasing information. But there is absolutely no consensus that approval or score are better than ranked. (In my opinion, ranked comes out ahead, and I talk about why in another comment here.) But most of all, ranked choice is what has the growing momentum, as a lot of reform organizations have concluded it's the best practical solution. But what's most important is that we adopt one of them, and it would be a shame to spend our effort squabbling over which alternative, than giving support to the one that has the most momentum and chance of succeeding now. We're really only even having this conversation now because of the success reform groups have had in promoting ranked voting.

DemocracyFTW2 4y ago
There's no way that it would be a good idea to choose a system (ranked choice) that has more complex rules while having known defects over an alternative, simpler system (approval voting) that apparently (based on Wikipedia's short discussion) lacks those defects and has an overall favorable performance when it comes to resistance against tactical voting. FWIW reading the decision rules for ranked choice puts me off. Purely in theory, one would probably prefer a system where all voters have to rank all the candidates; in practice, however, all of voting, validating, counting, applying the rules, and understanding the rules is significantly more difficult than approval voting where a candidate's points is simply the number of times a voter has marked their name on the ballot. You can (generally or for more important decisions) amend approval voting with a quorum, minimum percentage of approval, or minimal distance between first and second places without damaging the system with too much complexity, something I cannot even fathom how to achieve with ranked choice.
NickM 4y ago
it would be a shame to spend our effort squabbling over which alternative, than giving support to the one that has the most momentum and chance of succeeding now I think the bigger shame would be to spend so much effort promoting RCV, only to have people become jaded and disillusioned with voting method reform in general when they realize RCVs many glaring problems. The biggest problem with RCV IMO is that it looks like it eliminates the spoiler effect, but it really only eliminates it in simple cases where the third-party candidates are not competitive. This suddenly becomes very obvious and painful in cases like Burlington, 2009 or Alaska, 2022 where voters were told "RCV is great because you can vote honestly", only to realize post-election that this was entirely false, and that they threw away their votes and let the election go to a candidate they despise. Trust is extremely important in political movements, and very easy to lose. I fear RCV is going to poison the well for any FPTP alternatives for a long time to come.

jasonlotito 4y ago

That's a very misleading article that plays loose with the math. e.g. It makes assumptions on things that are flat out not true (for example, assuming people would vote for thing they clearly didn't vote for). If an article is going to sit there and make up votes that didn't occur, I can't really trust it.

FireInsight 4y ago

https://star.vote/ https://star.vote/ is pretty good. I've used it for some votes for a small community with usually more than fifteen different options, and it's worked pretty well!

TimPC 4y ago

Isn’t giving each candidate a score even more gameable than ranking candidates. If I like candidate A and think candidate B is their best opposition I can give candidate A 5.0 and candidate B 0.0 even if my real opinion of candidate B is 3.0. Lowering candidate B’s score artificially definitely helps candidate A.

avarun 4y ago
I think you misread the comment you're replying to. Calculating a total score is simply summing the scores assigned to a candidate from each voter, so there's no way to "lower candidate B's score artificially". There's no averaging going on here.
unholiness 4y ago
Absolutely. Naive exaggeration is a fantastic way to make your vote count more than others in score voting. It also encourages more negative campaign tactics, since convincing a voter to rank your opponent 0 rather than 4 is literally 4 times as powerful as convincing them to rank you a 5 rather than a 4.
- ClayShentrup 4y ago
  you have this exactly backwards, as explained by 3 different math phd experts here. http://www.youtube.com/watch?v=JtKAScORevQ http://www.youtube.com/watch?v=JtKAScORevQ https://www.rangevoting.org/NESD https://www.rangevoting.org/NESD https://electionscience.github.io/vse-sim/VSEbasic/ https://electionscience.github.io/vse-sim/VSEbasic/
naniwaduni 4y ago
This is a case for approval over score voting. It's not a good case for RCV over approval!
- ClayShentrup 4y ago
  no, it's not a case for approval voting over score voting, because score voting is better than approval voting if even one voter votes sincerely. if enough people use honest scores, then score voting is better than approval voting even for the "honest suckers". https://www.rangevoting.org/ShExpRes https://www.rangevoting.org/ShExpRes

duxup 4y ago

I read the linked article and I found it kinda scatterbrained. It seemed to dispute claims or tweets about how a vote works, along with what head to head voting outcomes would be. I do find how the math works out interesting. But I don’t find either disputes all that compelling. The head to head polling is interesting… but that’s not how the election works. The fact that two systems would be different seems normal to me. Folks who go to a ranked choice vote and pick just one, that’s their choice too. Same with people who don’t vote and so on. Not to say the article is worthless, it is interesting, but I’m not sure I agree with the conclusions.

unholiness 4y ago

Every voting system has some flaws (see: Arrow's impossibility theorem), and Instant Runoff Voting definitely saw its worst-case scenario rear its ugly head in Alaska: A Condorcet winner existed (Nick Begich), he lost, and voters whose first choice was Plain and second choice was Begich could theoretically, retrospectivally, all changed their first votes from Palin to Begich to have him win rather than Mary Peltola. It's important that I say retrospectively. This was a house race with national coverage, tons of high-quality polling from pollsters with nothing better to do, and it was still absolutely not clear from that polling that Begich would do better than Palin in round 2, or that Palin was likely to beat Begich in round 1. If I had the opinions of those hypothetically strategic voters, and I were completely strategic, and completely tuned into that polling, I still would have voted honestly. The strategy was only apparent after the fact. Thinking about "defense in depth" against this scenario, we'd be remiss not to mention Condorcet ranked choice voting methods, which always elect a Condorcet winner, and only resort to (game-able) elimination procedures if there isn't one. Not only would that have resulted in a better outcome in Alaska, it would have made the possibility of strategic voting much, much more remote, both because a strategic scenario would be much less likely, and because that strategic scenario would be much less visible before the election. With score voting, strategies are always available and almost always obvious: just exaggerate your score differences between the most likely candidates. The fact that strategic dishonesty would have helped in Alaska is a fair criticism of IRV, but the real commonness of strategic scenarios in STAR and score are a more primary feature.

garaetjjte 4y ago
>see: Arrow's impossibility theorem Arrow's theorem only applies to ranked systems. That said, cardinal systems are still susceptible to strategic voting (Gibbard's theorem)
- dane-pgp 4y ago
  Thank you for pointing out how overused "Arrow's impossibility theorem" is as an argument against all voting systems (and therefore often as an argument against all reforms to the current near-worst possible voting system). I would go further and be somewhat dismissive even of "Gibbard's theorem", since it only shows that an insincere vote can strategically produce a better outcome for a voter if they can identify what that that vote should be, and if they are one of the voters for whom such a strategy is available. So it doesn't guarantee (as far as I'm aware) that even a single voter (under every possible cardinal voting system) will necessarily have enough information before or during the election to confidently pick some specific insincere vote that will increase their chances of getting the outcome they want.

garaetjjte 4y ago

I'm not convinced that additional complexity of STAR is worth it compared to extremely simple score voting.

obelos 4y ago
As a practical consideration, the runoff in STAR does offer a legal advantage in some districts where a runoff is legally mandated.
godelski 4y ago
What's the added complexity of STAR? STAR is phase 1: sort(sum(candidates))[:2] (almost identical to score/approval/FPTP. We'd just take argmax instead of top two) and phase 2: eliminate other candidates, argmax(sum(candidates)) (identical to score/approval/FPTP). On the other hand RCV is a multi-round system with a while loop. In practice it almost never has fewer than 2 rounds (requires overwhelming majority, which almost never happens with more than 3 candidates). From an algorithms perspective RCV is more complex.
- Wowfunhappy 4y ago
  > From an algorithms perspective RCV is more complex. Perhaps, but people aren't computers. It's admittedly difficult to score this objectively, but I find RCV much easier to understand.
  - godelski 4y ago
    Much easier to understand from the voter perspective of the tallier perspective? From the voter perspective I find them fairly equivalent and both are frequently used in the everyday world (scoring might be even more considering 5 star reviews). But from a tallying perspective I think it is very hard to argue that RCV is substantially easier. Again, the score/approval algorithm is identical to the FPTP algorithm (which is trivial). It is simply argmax(sum(candidates)). Just sum the totals in an excel file and pick the largest number. I don't know a single person that has a hard time understanding this and I suspect if they did they would have an even harder time understanding how to tally RCV.

wizofaus 4y ago

As far as I can tell RCV is equivalent to what we've been using in Australia since federation to elect lower house candidates (we call it prefential voting and use it at all levels of government). I'm curious in what case "increasing your support for your genuine favorite can actually hurt their odds of winning"? I skimmed the link to the article you referenced and it appears that there were two republican candidates which is the main reason neither won - I don't believe I've ever seen a case here where one party had two candidates on the same ballot paper and I suspect it's probably illegal. I've also never seen a ballot paper with more than about 15 choices (and it's very rarely more than 10), so vote splitting is rarely a problem (it can be for minor parties with similar policies/leanings etc.). On that basis I'm pretty sure there's never been a case in this country where anyone has lost an election due to gaining higher primary levels of support - indeed it's pretty rare that the winning candidate doesn't have at least 35% of the primary vote (though exactly that happened in our recent federal election, as a total across all seats), and exceedingly rare for it to be less than 30%. There are certainly plenty of cases where the winner isn't the one with the highest primary vote, but it's typically only a few percentage points lower, and mostly affects the more progressive candidate because for many seats there are two such candidates with solid levels of support but only one mainstream conservative candidate.

tim-- 4y ago
I don't believe that there is anything stopping two candidates from one party. I believe I have even seen it before, but I can't find a reference.
- wizofaus 4y ago
  Doesn't seem compatible with the idea of "two-party preferred" voting which is what's actually used in practice here, though I'm not 100% clear as to whether that's explicitly different to preferential voting/RCV.
  - aldonius 4y ago
    Two party preferred is more a statistical count done on the assumption that {Labor, Coalition} = {Government, Opposition}. Of course it'll be decreasingly relevant now if the teals can stick.

JTbane 4y ago

STAR voting can be easily gamed- tell your supporters "vote MAX for me and 0 for everyone else". Anyone who gives more than a 0 to their second choice candidate will hurt their first choice.

garaetjjte 4y ago
Doing so would be ill-advised because it increases chances of winning for candidates that you actually dislike.
notriddle 4y ago
The only voting system that isn't "vulnerable" to bullet voting is FPTP, since it obligates you to bullet vote.
- dane-pgp 4y ago
  Actually that's not the only invulnerable system; there's also "Asset Voting", which I've mentioned in a comment above[0]. It also obligates you to bullet vote, since the voter's honest first preference is the only information needed, but it avoids the spoiler problems of FPTP, and is thus superior to it. [0] https://news.ycombinator.com/item?id=33117670 https://news.ycombinator.com/item?id=33117670
- crazygringo 4y ago
  Huh? Ranked voting isn't vulnerable to bullet voting. If you only rank a single candidate, there's zero benefit, you're just throwing away the rest of your vote. It's not giving that candidate any extra influence, the way it would in approval voting.
  - notriddle 4y ago
    > Ranked voting isn't vulnerable to bullet voting. "Vulnerable to bullet voting" is the same thing as "violating the later-no-harm principle," which means that some forms of ranked voting are vulnerable, while others aren't. - Borda Count is vulnerable to bullet voting, for the same reason as Score (giving points to anyone other than your first choice makes it less likely that your first choice wins, because it increases the score of your second and third choice). - Condorcet systems are vulnerable to bullet voting, because the ordering of your second and third choices can potentially change who the Condorcet winner is. - Instant Runoff is "immune" to bullet voting because the ranking of choices beyond the highest non-eliminated choice are ignored. This is also directly responsible for it being vulnerable to the spoiler effect, which makes non-bullet-voting pointless (your top choice must also be the choice most likely to win the 1-1 fight at the end, which means the remainder of your choices are highly unlikely to win, so why bother?).

barbariangrunge 4y ago

Ranked voting is supposed to prevent winner take all strategic voting. Voting by score encourages you to score everybody at zero except your favourite(s). This arguably means there will be less information in the vote tallies, although it gives voters more control over how their opinion will be responded to, and that is valuable

ClayShentrup 4y ago
this is obviously false. simply consider that green party supporters usually vote democrat. with score or approval voting, they'd ALSO want to vote for the green. whereas with ranked voting methods, they'd still strategically rank the democrat in 1st place, crushing any hope for the greens. http://www.youtube.com/watch?v=JtKAScORevQ http://www.youtube.com/watch?v=JtKAScORevQ

bfung 4y ago

There’s always going to be a trade-off when dealing with humans: voting result optimization vs ease of voting. That’s always been the issue, not that we don’t know of better result counting theories. Even in the 2020s, the ballots are explained over and over, “Connect the horizontal line”, or “fill in the box completely” — it’s a binary input to capture the best information from the lowest denominator. STAR would change the input from binary to an order N, and the number of mistakes or “that’s not what I intended” will happen. The average HN reader probably can fill in boxes pretty good, but the outside world not so much. RCV is a decent compromise.

SoftTalker 4y ago

A problem with all of these systems is that they are harder to understand, resulting in reduced voter confidence, or they demand more from the voter. Any kind of ranking is going to mean you need to know enough about each candidate to put them in some kind of order. Maybe OK if there are two candidates but much harder after that. Most people will rank their favorite on top and then just randomly rank the remainder. Simple plurality or majority vote is easiest to understand and easiest for the voter. You vote for the one candidate you like the best. If the isn't a majority for one and that is required, you have a runoff between the top two.

Wowfunhappy 4y ago
I agree, but I'd argue that instant-runoff/RCV is the second-easiest to understand, by far, while also being much better than simple plurality voting. Which is why I strongly favor instant-runoff.
- dane-pgp 4y ago
  Actually there's another option which hasn't been mentioned so far, which is just as easy to implement as plurality (for both voting and counting by hand), and has pretty much all the benefits of IRV/RCV. That method is "Asset Voting", which works exactly like a plurality vote, except that once the totals for each candidate are known, the candidates are eliminated one by one (from least to most votes) and get to assign their vote counts to any remaining candidate. If you agree with me and the parent post that voter confidence (including avoiding electronic counting machines) is the most important factor when evaluating possible alternatives to plurality (assuming the alternatives are no worse than plurality) then hopefully you will agree that Asset Voting is preferable to IRV/RCV.

breuleux 4y ago

I don't think the voting system matters as much as we think. The main attack vector that interest groups use against democracy is shaping the public opinion in the direction that benefits them. Certainly, some systems may be easier to game, but ultimately interest groups will do neither more nor less than the minimum required to win. If they only need to make 30% of the population rabid in order to win, that's what they will do, but let's not fool ourselves into thinking they couldn't make 51% of the population rabid if they needed to. Basically, if people's stated preferences are ill-thought-out and/or manipulated, then even the perfect voting system that perfectly reflects the people's preferences will yield bad results. Voting reform in this case would be fixing the symptoms, not the disease. The disease, in my view, is that given the sheer complexity of the issues, the poor attention budget of people, the outsized influence of media, the impossibility of directly interacting with decision-makers, there's basically no way people can make an informed decision. Personally I think we should ditch elections and just pick representatives purely at random, and then pay them to think about their preferences full time, discuss them with each other, directly interview candidates for executive positions, and so on. The basic idea is that I think collective decision making probably reaches an optimum around Dunbar's number and falls off precipitously beyond that.

zajio1am 4y ago

Approval voting fails in a situation that is pretty common in Europe (e.g. French presidential elections) - you have center left, center right, far left and far right candidate. If you vote for center left/right, you have to decide whether to also approve center right/left candidate, which may be contrary to your values, corrupt and incompetent, but at least not communist or fascist. That is why later-no-harm property of IRV is important.

zajio1am 4y ago

> Ranking the voting systems: STAR Voting > Approval Voting > Ranked Choice Voting > Plurality ("pick only one") voting. Funny is that you ranked them and not scored them nor just say whether you approve them or not. If you have to vote for these voting systems using approval voting, which ones would you approve?

ClayShentrup 4y ago
scores would be more like: STAR or score or approval = 5 IRV/RCV = 2 plurality = 0 how i'd approve them in a real election would be based on electability, which would be true with any voting method used in an actual contentious election.

dragonwriter 4y ago

> Ranking the voting systems: STAR Voting > Approval Voting > Ranked Choice Voting > Plurality (“pick only one”) voting. I’m going to disagree. For essentially single-winner, public, candidate, secret ballot elections: Any Condorcet method* > Bucklin Voting* (without limited # of ranks) > Instant Runoff Voting* > Majority-Runoff** > Plurality Voting** > Approval***, Score***, or STAR*** Voting (or any other *** system) For public, secret ballot, legislative elections: Mixed Member Proportional (MMP) using PR-STV* + Party List for leveling) > MMP using any single-member district system + Party List for leveling > Single Transferrable Vote (PR-STV)* > Party List PR (optionally, using STV to minimize wasted votes, but vote-for-one is close behind) > single-member districts with single winner systems in the order above (but, compared to the preceding proportional systems, all are close enough as to make no real difference) > any system using some or all multimember constituencies (including staggered elections in the same district for multiple seats in the same body) not described earlier in this list. For public, secret ballot, chief executive + designated successor elections: The single winner methods, in the order above, modified to use ranked ballots if they don’t already, in two rounds where the first elects the chief executive, then that winner is eliminated from all ballots, then a second counting of the ballots is done to chose the designated successor > The single winner methods, in the order above, separately for each office > The single winner methods, i the order above, for slates of pairs of one candidate for each office > Plurality voting** in a two round system (with normal, vote-for-one ballots), where after a chief is chosen on the first round, the successor is chosen by eliminating all ballots for the prior winner and taking the highest remaining vote count. (methods with * are ranked ballots methods, methods with ** are vote-for-one methods commonly in use in the US, methods with *** are either artificially limited-ranks or more-detail-than-ranks methods, which for candidate elections have no clear, unique honest mapping from actual preferences to ballot markings; they are very good when their particular system of ballot markings maps to something concrete – e.g., for approval for selecting group activities in non-secret-ballot elections where approving is a commitment to participate if that option is selected or disapproval forfeits the right to participate, or, for score voting, where chosen score maps to a commitment of resources in the event the scored item is chosen.)