Robert-- http://en.minguo.info/polling_station This is a big topic, and various issues have been raised. I'd like to address them all. First, it's good that you're willing to discuss these voting-system-choice issues, and not shut-out disagreeing opinion. I've been meaning to discuss, with people of Burlington, replacement alternatives for Plurality & IRV. There are 2 main issues here: 1. The merit of Approval, vs rank methods. 2. Which rank-method ...(recognizing that activists & organizations & activists prefer rank-methods, and that some overecompromisers & rival parties are likely to vote better with rankings than with Approval.) I feel that #2 is the more practical & important issue. But #1 is important too, at least partly because of Approval's unmatched easy & entirely cost-free implementation. I'll discuss these issues as they come up in this post that I'm replying to, rather than purposely separate them into separate sections of this post. But most of this post that I'm replying to is about #1, and so I guess that will be this reply's main topic. But, recognizing that people want, and many voters need, a ranking-method, I definitely want to discuss the choice of a rank-method, as an immediate practical municipal choice. Probably in an immediately subsequent post. So I emphasize: However little you like what I say about Approval, don't shut me out just yet, because I recognize that the choice-of-rank-method topic is really more important than the Approval-vs-rank topic, and I intend to address choice-of-rank-method next. On Tue, Nov 22, 2016 at 2:27 AM, robert bristow-johnson < rbj@audioimagination.com> wrote: > >> > >> I don't see any reason why IRV results in the dominance of the two major > >> parties. > >> > > > > ...except that IRV's national track record is one of 2-party dominance. > > > > the national 2-party dominance exists independently of IRV > Where IRV has been used for about a century, including for at least some state, and probably national, offices, 2-party dominance remains in place. Returning to discussion about reform in this country: I'm not saying that IRV can't be any good under any circumstances or conditions. ...only that it isn't always the best choice. You wrote: . in Burlington, we may have a 2-party dominance, but the GOP is not one of the two. and occasionally a GOP candidate is elected mayor and in 2009 there was a legitimate 3-way (or 4-way if you include an Independent) race between candidates of whom none were dominant. [endquote] The broader contexts that Daniel brought up can't be ignored, and inevitably come up.. For example, should the method be chosen tailored for each particular municipality, or, even in municipal reforms, should it be chosen for usefulness in state & national elections (even though it's much less likely for democracy to be achieved at state & national level)? . (...but Maine's adoption of IRV suggests that maybe state democracy isn't entirely impossible, even now. Can that be?). I suggest that, even when choosing a municipal voting-system reform, the method should be chosen for its usefulness at state and national level, because those levels are where the most broadly-important decisions are made. That's relevant to this topic, because, nationally, the meaningful choice is between progressive candidates & parties, vs Republocrat candidates. Could there be any easier choice? Hence my emphasis on the situation where the voter has a strong top-set and a strong bottom-set. ...as every progressive voter does, given the current candidate lineups. And I claim that the 99% would benefit greatly from a progressive (rather than Republocrat) government. ...and therefore has a strong top-set and a strong bottom-set. (Replying a little farther down) : > > >> The problem that IRV suffers is that sometimes IRV does not elect the > >> Condorcet Winner and then voters feel that the election failed when a > >> candidate is elected despite that more voters marked their ballots that > >> they preferred a specific different candidate. Then this cascades into > >> other problems like spoiler and later harm.. > >> > > > > That well characterizes what happened in Burlington, but IRV doesn't > > violate Later-No-Harm. ...or Later-No-Help either. > > well, note that there were no caps. i wasn't referring to > https://en.wikipedia.org/wiki/Later-no-harm_criterion but i *was* > referring to 1500 voters who were promised that they could vote for their > favorite candidate without worrying about helping elect their least > favorite candidate and found out after the election that this was not true. > the promise of IRV had failed. these voters' political interests were > harmed by their expression of their most favorite candidate. > I see what you mean. Yes, certainly. Everywhere where Richie & his FairVote promoted IRV, most likely those same false promises were made, and IRV's problems were concealed. Some of us have been trying to explain to Rob Richie, over the past 30 years, IRV's problem. Over that time, we've been telling Richie about the very problem that happened in Burlngton. He can't say he hadn't heard about it or didn't know about it. > > IRV is a good method. But don't advocate or adopt it unless you > understand, > > accept and want its problem, and its unique offering and tradeoffs. IRV > > isn't for everyone. > > no, no, no, no... > > we don't go to the general public and say that. we, instead, try to > deliver to them a means of electing a candidate out of a field of many, in > such a way that, given one-person-one-vote weighting, these voters are > **relieved** of worry about "understanding" how using or failing to use > tactics will harm their political interests. > > that's what you don't get, Mike. > No, I agree completely. My goal at EM has been to emphasize the importance of making voting easy (less strategic difficulty), and making sincerity safe. There's good reason to believe that, just as you said, IRV doesn't fill the bill. I don't advocate IRV, and I wouldn't include it in a multi-method proposal. [Replying a little farther down] : > > > > > > > In Burlington, you repealed it because you didn't like its elimination > of a > > CWv. Then why did you adopt it??? > > > > > You didn't know that it could do that? > > Did FairVote assure you that it would never happen? > > yes, FairVote assured the citizens of Burlington that they could vote for > their favorite candidate without worry of electing their least favorite > candidate. but 1500 voters for the GOP (that also hated the Prog) found > out that they could have prevented the election of the Prog had they > insincerely elevated the Dem to first choice. but that would not have > happened with Condorcet. > As I mentioned above, we've been explaining about that to Rob Richie for the past 30 years. Could have have not known about it? So: ignorance, or intentional concealment? > <<gobbledigook deleted>> > Part of the "gobbledigook" is about IRV's mitigations, which admittedly aren't relevant to this discussion, because I don't advocate IRV, and because there's no way that Burlington would want IRV. But part of it is about Condoret's favorite-burial need, for some overcompromising voters. Also, it might be in that section that I described how Condorcet could meet FBC (and thus not have that favorite-burial need). ...and offered two rank methods that meet FBC and don't have a chicken-dilemma problem (Condorcet has chicken dilemma, as does Approval). You're of course not obligated to reply to any of that, but Condorcet's favorite-burial need is there. And the offer of a way to bring FBC to Condorcet is there, and so is the offer of two rank methods that meet FBC and don't have a chicken-dilemma problem. (Those two methods are Simmons ( MDDA(pt/2) ) and MMPO(pt/2). ...and also Kevin Venzke's IC, MMPO , bringing it to 3 methods) It's a mistake to assume that Condorcet is really free of strategy. Maybe I should say something here about ordinary Condorcet's favorite-burial need: Say that you top-rank Compromise, and 2nd-rank Favorite. Compromise is going to win as CWv (voted CW). But then you (and a few other voters who vote like you) decide to rank Favorite at top, along with Compromise. Oops! By voting Compromise over favorite, you had been keeping Favorite from pair-beating Compromise. But now that you've moved Favorite to top, you're no longer voting Comprmise over favorite. Now Compromise pairwise-beats favorite. It happens that Worst pairwise-beats Compromise, and that Worst is the winner of the resulting cycle. By moving Favorite to the top with Compromise, you've changed the winner from Compromise to Worst. Something similar can happen if, instead of winning as CW, Compromise was merely winning in a cycle, by having the lowest margin-against, if the method is Condorcet(margins). By no longer ranking Favorite over Compromise, you could spoil that win too, by decreasing the Favorite > Compromise total, thereby increasing the margin by which Compromise pairbeats Favorite, giving Compromise now a higher margin-against than Worst has. These scenarios can happen only if there's a sincere cycle. That's because: If Compromise were the CWs (sincere CW), then your sincere ranking obviously can't hurt her win as CWv. If someone else is CWs, then, in Condorcet(wv) the only way you could chanage the winner from hir to Compromise would be by burial. (In Condorcet(margins) you could do it by truncation.) So there has to be a cycle. Sincere cycles are evidently rare (at least in the current (probably inauthentic) political system. Maybe in an authentic political system (one free of media disinformation and count-fraud) there wouldn't be a 1-dimensional political spectrum, and cycles wouldn't be unusual. Anyway, there are insincere cycles too. Strategic cycles. Burial strategy is by no means out of the question. Burial will make a cycle...like the one that Condorcet's FBC failure needs. We've discussed, on EM, how you can protect the CWs from burial (in wv or margins) by moving that CWs up to 1st-place. But, if not enough people do that, then you could need to rank the CWs over your favorite, to prevent successful burial of the CWs under your favorite. Yes, if the method is Condorcet(wv) (but not Condorcet(margins), the burial can likewise be thwarted (& penalized) if the people who prefer the CWs to the buriers' candidate refuse to rank the buriers' candidate. Of course then they need to know or guess who's the CWs, and who's going to bury. Yes, that's predictive information. Their defensive of the CWs is tactical strategy that depends on good predictive information. If you're going to use Condorcet, then you'd better be good at tactical strategy, and have good predictive information. (Well, admittedly, the fact that you'll truncate the (guessed) buriers' candidate if you prefer the (guessed) CWs to hir can (but might not) deter burial, because the would-be buriers don't have better predictive information than you do.) But, if the expected or evident CWs (CWse) is in the would-be buriers' bottom-set, then they don't really care if their burial is penalized by the election of a different (negligibly worse) bottom-set member. What happens then? A burial-fest resembling a sharks' feeding-frenzy. I call it the "perpetual burial fiasco". 'Round about that time, you'd be wishing that the election were an Approval election. Let's not kid ourselves that Condorcet is free of strategy. So: You can change the winner from Compromise to Worst by moving Favorite to top alongside Compromise. ...but also especially more likely by ranking Favorite over Compromise. As I said, burial strategy isn't out of the question, and the need to vote insincerely, as described above, isn't out of the question either. That's why some of us offer ranking-methods that meet FBC. That's why IC Condorcet is offered, to bring FBC to Condorcet. For instance, IC Condorcet could consist of IC RP. RP(wv) is much better than RC(margins), because RC(wv) is completely truncation-proof, whereas RC(margins) leaves the CWs completely vulnerable to truncation, whether strategic or innocent. Also, RP(margins) loses FBC in more scenarios than does RC(wv). Steve Eppley's MAM is just RC(wv) with Steve's clever way of avoiding some problems and meeting more criteria in small electorates. So, if you insist on Condorcet (and are therefore willing to give up chicken-dilemma protection), then consider IC MAM. The version of Condorcet Criterion that IC MAM meets is arguably better and more meaningful than the ordinary Condorcet Criterion. ...but while IC Condorcet meets FBC, it still has chicken-dilemma, as does Approval. So we offer MDDA(pt/2), MMPO(pt/2), and IC,MMPO. ...which meet FBC and don't have a chicken-dilemma problem. [Replying a little farther down] : > > Approve only your favorite if you insist on voting tactically (based on > > predictive information) and if it looks as if your favorite has a win. > > > > Is that a problem to you? > > **what** predictive information?? the polls??? have you learned nothing > in the last two weeks? > There was nothing for me to learn in the past 2 weeks, because what happened in the past 2 weeks exemplifies what I've been saying all along. You haven't been reading my postings, or you'd know that I'm the one who has been saying that the media's intentionally disinformational predictive information is no good. So, in Approval, forget all about tactical voting and predictive information. I said, "Just vote (only) for your strong top-set". By the definition of the strong top-set, electing from it, instead of your strong bottom-set, is incomparably more important than the matter of _which_ strong top-set wins. And, if you insist on trying to choose among your strong top-set, then you increase the probability of electing from your strong bottom-set. And if you don't have strong top & bottom sets? You do, in our state & national elections (the elections that most matter), at least in our current, probably phony, political system. Likely, in a better world, the kind that has honest media and vote-counting, your bottom-set candidates & parties would no longer be in the elections. The bottom-set is likely an artifact of a phony political system. But, if you didn't have strong top & bottom sets, then what? Then, maybe, under those conditions we'd have honest media, and good predictive information, for one thing. ...on which to base good tactical expectation-maximizing strategy. But what if there were no strong top & bottom sets, and also there were no good predictive information? 1. Look, if it's difficult to decide what set of candidates you consider top-set, the better set that deserves approval, then it doesn't make a whole lot of difference how you choose between those sets that you're indifferent between. So don't worry about it. If all the candidates are so good that there's no strong bottom-set, then just be glad of that, and take your best guess for your approval-set, because it doesn't matter much under those conditions. 2. That said, there's a way to make an especially good guess: There's 0-info expectation-maximizing strategy. ...but only if you want to use it. I emphasize that you needn't bother with it. It's only for people who like strategy and want to use it (though they don't need to), to make their best guess. But one guess is reallly as good as another when you're basically indifferent between the various possible approval-sets. But there's no reason to believe that media disinformation, & media exclusion of candidates & policy-solutions other than those of the 1% are the only tactic in use. With our unverifiable vote-count, and the plentiful evidence of presidential election count-fraud in 2000 & 2004, there's no reason to believe that count-fraud isn't a main routine tactic of our rulers. I mean, given all the mass killing that the 1% (or less) ruling-families & their employees don't mind doing to gain more power & more wealth, does anyone think that they'd be willing to let themselves be voted out of office, or to take any chance of that? Why should they leave anything to chance, or to us voters? Count-fraud, at least whenever needed, is an obvious part of their optimal strategy. You wrote: > voters shouldn't have to worry about "predictive information" in casting > their vote. they should not have to face tactical voting in the polls in > any common case. > Exactly. That's one reason why like Approval, and why I agree with David that it's the best method (except that chicken-dilemma protected methods do bring some improvement). I discussed that in more detail a few paragraphs above, in my previous long comment just before this one. This posting is already long, so I'll just post it now, as Part 1. Part 2 will follow, tomorrow. I re-emphasize, don't shut me out just yet, because I recognize that the discussion of the choice between ranking-methods is more important than the discussion of Approval vs ranking. ...and my post on the choice between ranking-methods will follow the replies to the posting that I'm now replying to. Michael Ossipoff (This concludes Part 1 of this reply. Part 2 of this reply will be along tomorrow morning.) (I don't have a convenient way to delete text, but all of this postings replies are above this point). > > > > > > > > Ii suggest that if you're a progressive, then there are a large set of > > candidates whom you so much prefer to the alternatives, that your best > > strategy is to approve them all. > > that is not the fact. how can you tell that to a naive voter, "just > approve them all"??? > > > > >> Voters don't want their second-choice to beat their first-choice. How do > >> they stop that from happening? (By bullet-voting their first choice, and > >> that devolves Approval into Plurality.) > >> > > > > You're worrying too much about voting _among_ the best candidates. > > the voter shouldn't have to worry AT ALL. let the voter just inform the > poll what their contingency votes are. > > > > > > >If > > there's a set of candidates who are qualitatively the best, incomparably > > better than the rest, then your best strategy is to maximize the > > probability that one of them will win. That doesn't include voting > _among_ > > them. Just vote _for_ them. > > and maybe the voter will find out they helped their less-preferred > candidate beat their more-preferred candidate. sure, that'll make the > voter feel like their vote is effective! > > > > In any case, I don't recommend tactical voting (voting based on > predictive > > information)...partly because the corporate mass-media's predictive > > information is disinformational, > > no shit. but that still does nothing to relieve the voter of the tactical > pressure because they **want** their favorite candidate to win, they don't > want their contingency candidate to beat their favorite. but they also > want their contingency candidate to beat their most-hated candidate if the > election comes down to that. Approval does **nothing** to help them > express that contingency. > > >in the service of the people who "own" the > > mass media. > > > > ...but also partly because it's most to your advantage to just approve > the > > whole set of candidates that's qualitatively, incomparably, better than > the > > alternatives. ...to maximize the probability that one of that strong > > top-set will win. > > > > (But until we have verifiable vote-counting, of course the voting system > is > > irrelevant, as is the voting.) > > totally different topic. (everyone should adopt optical-scan paper > ballots.) > > precinct summability is necessary. IRV and Bucklin don't have it, but > Condorcet (as well as Plurality) do have it. > > > > > > > > > > > > > >> > Equal-rank approval votes mean > >> > voters don't get the option to express their preferred ranking, but > >> because > >> > of that, they aren't at all motivated to bias their ranking > dishonestly. > >> > >> That's baloney. They may very-well not Approve of their second-choice > >> candidate (whom they do actually approve of) to protect their > first-choice > >> candidate. > >> > > If they strongly approve of their 2nd choice, but not their 3rd choice, > > then they'll approve their 2nd choice. > > unless they worry about their 2nd choice beating their 1st choice. > > > In our actual candidate lineup, the > > merit differences among the progressives are negligible in comparison to > > the big gulf of a merit difference between them and the Republocrats. > > > > Don't worry about it so much. Just approve all of the best. > > you're repeated assurances are hollow. just like Rob Richie. > > > > > > > > > > > >> > >> > They only have to decide which candidates to approve, or where the > >> cut-off > >> > is between approval and disapproval. > >> > >> No. That is a falsehood. > >> > >> They have to consider whether or not their second-choice candidate has a > >> chance of beating their first-choice. And they have to consider whether > or > >> not their third-choice has a chance of beating either their > second-choice > >> or their first-choice. And then they have to decide if that risk is > worth > >> it to Approve their second or third choices. > >> > > No, they don't have to do any of that. They don't have to try to vote > > tactically. > > bullshit. you have no right to tell a voter what his or her interests > are. they **may** have to consider voting tactically because they want > their favorite candidate to win. but, if their favorite is a poor showing, > they also don't want their contingency candidate to lose to their hated > candidate. > > telling them "You don't have to try to vote tactically" is hollow. it's > like saying "Don't worry. Any of the candidates are okay." > > > > > Partly because there's no reliable information on which to base > > tactical voting, > > which is just one reason we should relieve voters of that burden. > tactical voting is **not** a perk. it's a burden. no one feels good > about it. > > > and partly because their best strategy is just to approve > > all of the set that is incomparably, qualitatively better than the other > > candidates. > > says YOU. > > you cannot tell voters what their best strategy is and they find out later > that following your glib assurance screwed them. that is what FairVote did > with Burlington and there was a city that was quite pissed. we called it > the "IRV Happy Talk". > > you are no authority to advise voters what their best strategy is. the > best we can do is give them an voting system that normally requires no > strategy of voters nor benefits voters from using strategy. after a few > generations, Duverger's Law will be forgotten and they can confidently > simply vote their political expression. > > > > > >> > >> > >> > >> Robert says: > >> > > > > > > I think most of your gushing for Approval voting is not well-founded. > What > > if the voter prefers Candidate A as their favorite, but suspects that > > Candidate B and Candidate C are the leading candidates? How should that > > voter vote with an Approval ballot? Please answer that question with > > specifics. > > > > [endquote] > > > > We'd need better "specifics" about the situation that you've incompletely > > described. It depends on how much you like A better than B & C, and how > > much you like A & B better than C. > > > > NO, we DON'T. that is your failure. you cannot force voters to research > these "specifics". > > All the voter knows is that he/she prefers A above all others. If he/she > can't get A, then they'll take B. If they can't get A or B, then they'll > settle for C. But this voter really doesn't want to help candidate D at > all. The only ballot that expresses the will of this voter is the Ranked > Ballot. Approval does not gather enough information (forcing the voter to > vote tactically) and Score requires too much information (also forcing the > voter to score tactically). With the exceptions of goofy things coming out > of a cycle (which would be rare), only Condorcet resolves that election in > exactly the same manner as would a Simple Majority race between two > candidates with equal weighting for all voters. Only Condorcet does that. > > > > > You don't mention any set's qualitative or incomparable better-ness,and > you > > suggest that there's good predictive information that the winner will be > B > > or C. > > i am saying the voters should not be required to have "good predictive > information" for them to vote their interests. And no political group > should benefit from "good predictive information". That's the Duverger > issue. It's because, with plurality voting, "good predictive information" > tells voters that one of the two major party candidates is expected to win, > so that their vote for third-party or independent is wasted. > > > > So, if you don't have a strong top or bottom set, and the merit > differences > > between your successive choices are uniform, and if there's reliable > > predictive information that the winner will be B or C, then of course > > approving {A,B} is a good way to vote. > > But you DON'T necessarily know that and you shouldn't burden the voter > with that tactical need. They shouldn't have to be thinking "Hmmmm, what > would be a good way to vote to benefit my political interest?" That's your > problem and you do NOTHING to address it. > > > > But in our actual political lineup, the media have people believing that > > there are 2 parties that always must and will win. Say your favorite is > > Jill Stein, and that the other 2 candidates are Hillary & Donald. You're > > told that only Hillary &I Donald are well-liked by everyone enough so > that > > one of those two beloved and popular candidates is sure to win. > > > > For one thing, you should find something fishy about such a claim, > > You are expecting too much from voters. Voting should not be a political > science exam. It's a **service** to voters, not a **burden** for voters. > Or at least it should be. > > > given > > what is known about those 2 candidates being the most despised > presidential > > candidates in this country's history. > > > > For another thing, the merit of Jill is so much greater than that of > > Hillary or Donald that, no matter what your'e told or believe about > > predictive information, your best strategy is to vote for the only > > acceptable, honest and progressive, genuinely pro-people candidate. > > > > That latter interpretation of your scenario is the one that you probably > > intended, because it's the one that actually obtains in elections here. > > with your own words, you are making it clear that voters will have to > explicitly consider strategy when voting in an Approval system. > > Again, you cannot simply answer the question: "What should the voter do > with their second-favorite candidate?" > > With Ranked-Choice Voting, the answer is clear (they mark their > second-favorite with a "2"). > > With Approval Voting, the voter has a tactical burden deciding whether or > not to approve their second-choice. And with Score Voting, the voter has a > tactical burden deciding how much to score their second-choice. > > Mike, you did nothing but verify my objections to Approval voting for > governmental elections. > > > -- > > > > > r b-j rbj@audioimagination.com > > > > > "Imagination is more important than knowledge." > > ---- > Election-Methods mailing list - see http://electorama.com/em for list info > >

In case I seem unfair, in my comparison of Approval & Condorcet, let me briefly summarize & clarify: If you have strong top & bottom sets, then, by their definition, using Approval, and approving (only) all of your strong top-set is optimal. It can't get any more optimal than that. If, hypothetically, you have a choice in voting approval-like, top-ranking your strong top-set, and having that count fully, as in approval, or sincerely ranking in order to choose _among_ your strong top-set, then you increase the probability of electing from your bottom-set. So, sincere ranking with Condorcet helps your strong bottom-set, in comparison to optimal strategy in Approval. And, struggling to avoid a strong bottom-set win, when ranking in Condorcet, you're struggling with inadequate, unreliable predictive "information", trying to vote tactically. ...in order to minimize how much more likely you're making the election of someone from your strong bottom-set. Non-tactical Approval-voting is a completely different kind of voting, a completely different paradigm in voting. And, even if you don't have strong top & bottom sets, and even if there's no predictive information, making tactical voting impossible (you don't need it anyway), that's ok too, because if it isn't clear, if you don't have a strong feeling for one approval-set choice over another, then guess what: ...it doesn't matter. Then, in your A>B>C example, if you don't feel any differently about {A,B} vs C, compared to {A} vs {B,C}, then either is fine. Vote whichever approval set you want to. If you're indifferent between them, you can flip a coin. Either way, you're approving what you want, as nearly as you can judge. And how could it be any nearer or better than that? Don't bother with 0-info expectation-maximizing strategy if you don't like strategy. It isn't necessary. Just approve what you like. If you don't have a clear preference between {A,B} vs {C} and {A} vs {B,C}, than it doesn't matter which you vote. Michael Ossipoff On Tue, Nov 22, 2016 at 8:25 PM, Michael Ossipoff <email9648742@gmail.com> wrote: > Robert-- > > http://en.minguo.info/polling_station > > This is a big topic, and various issues have been raised. I'd like to > address them all. > > First, it's good that you're willing to discuss these voting-system-choice > issues, and not shut-out disagreeing opinion. I've been meaning to discuss, > with people of Burlington, replacement alternatives for Plurality & IRV. > > There are 2 main issues here: > > 1. The merit of Approval, vs rank methods. > > 2. Which rank-method > > ...(recognizing that activists & organizations & activists prefer > rank-methods, and that some overecompromisers & rival parties are likely to > vote better with rankings than with Approval.) > > I feel that #2 is the more practical & important issue. But #1 is > important too, at least partly because of Approval's unmatched easy & > entirely cost-free implementation. > > I'll discuss these issues as they come up in this post that I'm replying > to, rather than purposely separate them into separate sections of this > post. But most of this post that I'm replying to is about #1, and so I > guess that will be this reply's main topic. > > But, recognizing that people want, and many voters need, a ranking-method, > I definitely want to discuss the choice of a rank-method, as an immediate > practical municipal choice. Probably in an immediately subsequent post. > > So I emphasize: However little you like what I say about Approval, don't > shut me out just yet, because I recognize that the choice-of-rank-method > topic is really more important than the Approval-vs-rank topic, and I > intend to address choice-of-rank-method next. > > On Tue, Nov 22, 2016 at 2:27 AM, robert bristow-johnson < > rbj@audioimagination.com> wrote: > >> >> >> >> I don't see any reason why IRV results in the dominance of the two >> major >> >> parties. >> >> >> > >> > ...except that IRV's national track record is one of 2-party dominance. >> > >> >> the national 2-party dominance exists independently of IRV >> > Where IRV has been used for about a century, including for at least some > state, and probably national, offices, 2-party dominance remains in place. > > Returning to discussion about reform in this country: > > I'm not saying that IRV can't be any good under any circumstances or > conditions. ...only that it isn't always the best choice. > > You wrote: > > . in Burlington, we may have a 2-party dominance, but the GOP is not one > of the two. and occasionally a GOP candidate is elected mayor and in 2009 > there was a legitimate 3-way (or 4-way if you include an Independent) race > between candidates of whom none were dominant. > > [endquote] > > The broader contexts that Daniel brought up can't be ignored, and > inevitably come up.. For example, should the method be chosen tailored for > each particular municipality, or, even in municipal reforms, should it be > chosen for usefulness in state & national elections (even though it's much > less likely for democracy to be achieved at state & national level)? > . > (...but Maine's adoption of IRV suggests that maybe state democracy isn't > entirely impossible, even now. Can that be?). > > I suggest that, even when choosing a municipal voting-system reform, the > method should be chosen for its usefulness at state and national level, > because those levels are where the most broadly-important decisions are > made. > > That's relevant to this topic, because, nationally, the meaningful choice > is between progressive candidates & parties, vs Republocrat candidates. > > Could there be any easier choice? Hence my emphasis on the situation where > the voter has a strong top-set and a strong bottom-set. ...as every > progressive voter does, given the current candidate lineups. And I claim > that the 99% would benefit greatly from a progressive (rather than > Republocrat) government. ...and therefore has a strong top-set and a > strong bottom-set. > > (Replying a little farther down) : > > >> >> >> The problem that IRV suffers is that sometimes IRV does not elect the >> >> Condorcet Winner and then voters feel that the election failed when a >> >> candidate is elected despite that more voters marked their ballots that >> >> they preferred a specific different candidate. Then this cascades into >> >> other problems like spoiler and later harm.. >> >> >> > >> > That well characterizes what happened in Burlington, but IRV doesn't >> > violate Later-No-Harm. ...or Later-No-Help either. >> >> well, note that there were no caps. i wasn't referring to >> https://en.wikipedia.org/wiki/Later-no-harm_criterion but i *was* >> referring to 1500 voters who were promised that they could vote for their >> favorite candidate without worrying about helping elect their least >> favorite candidate and found out after the election that this was not true. >> the promise of IRV had failed. these voters' political interests were >> harmed by their expression of their most favorite candidate. >> > > I see what you mean. Yes, certainly. Everywhere where Richie & his > FairVote promoted IRV, most likely those same false promises were made, and > IRV's problems were concealed. > > Some of us have been trying to explain to Rob Richie, over the past 30 > years, IRV's problem. Over that time, we've been telling Richie about the > very problem that happened in Burlngton. He can't say he hadn't heard about > it or didn't know about it. > >> > IRV is a good method. But don't advocate or adopt it unless you >> understand, >> > accept and want its problem, and its unique offering and tradeoffs. IRV >> > isn't for everyone. >> >> no, no, no, no... >> >> we don't go to the general public and say that. we, instead, try to >> deliver to them a means of electing a candidate out of a field of many, in >> such a way that, given one-person-one-vote weighting, these voters are >> **relieved** of worry about "understanding" how using or failing to use >> tactics will harm their political interests. >> >> that's what you don't get, Mike. >> > > No, I agree completely. > > My goal at EM has been to emphasize the importance of making voting easy > (less strategic difficulty), and making sincerity safe. > > There's good reason to believe that, just as you said, IRV doesn't fill > the bill. > > I don't advocate IRV, and I wouldn't include it in a multi-method proposal. > > [Replying a little farther down] : > > > >> >> >> >> > >> > In Burlington, you repealed it because you didn't like its elimination >> of a >> > CWv. Then why did you adopt it??? >> >> >> >> > You didn't know that it could do that? >> > Did FairVote assure you that it would never happen? >> >> yes, FairVote assured the citizens of Burlington that they could vote for >> their favorite candidate without worry of electing their least favorite >> candidate. but 1500 voters for the GOP (that also hated the Prog) found >> out that they could have prevented the election of the Prog had they >> insincerely elevated the Dem to first choice. but that would not have >> happened with Condorcet. >> > As I mentioned above, we've been explaining about that to Rob Richie for > the past 30 years. Could have have not known about it? So: ignorance, or > intentional concealment? > > > >> <<gobbledigook deleted>> >> > > Part of the "gobbledigook" is about IRV's mitigations, which admittedly > aren't relevant to this discussion, because I don't advocate IRV, and > because there's no way that Burlington would want IRV. > > But part of it is about Condoret's favorite-burial need, for some > overcompromising voters. > > Also, it might be in that section that I described how Condorcet could > meet FBC (and thus not have that favorite-burial need). ...and offered two > rank methods that meet FBC and don't have a chicken-dilemma problem > (Condorcet has chicken dilemma, as does Approval). > > You're of course not obligated to reply to any of that, but Condorcet's > favorite-burial need is there. And the offer of a way to bring FBC to > Condorcet is there, and so is the offer of two rank methods that meet FBC > and don't have a chicken-dilemma problem. > > (Those two methods are Simmons ( MDDA(pt/2) ) and MMPO(pt/2). > ...and also Kevin Venzke's IC, MMPO , bringing it to 3 methods) > > It's a mistake to assume that Condorcet is really free of strategy. > > Maybe I should say something here about ordinary Condorcet's > favorite-burial need: > > Say that you top-rank Compromise, and 2nd-rank Favorite. Compromise is > going to win as CWv (voted CW). > > But then you (and a few other voters who vote like you) decide to rank > Favorite at top, along with Compromise. > > Oops! > > By voting Compromise over favorite, you had been keeping Favorite from > pair-beating Compromise. But now that you've moved Favorite to top, you're > no longer voting Comprmise over favorite. Now Compromise pairwise-beats > favorite. It happens that Worst pairwise-beats Compromise, and that Worst > is the winner of the resulting cycle. > > By moving Favorite to the top with Compromise, you've changed the winner > from Compromise to Worst. > > Something similar can happen if, instead of winning as CW, Compromise was > merely winning in a cycle, by having the lowest margin-against, if the > method is Condorcet(margins). By no longer ranking Favorite over > Compromise, you could spoil that win too, by decreasing the Favorite > > Compromise total, thereby increasing the margin by which Compromise > pairbeats Favorite, giving Compromise now a higher margin-against than > Worst has. > > These scenarios can happen only if there's a sincere cycle. > > That's because: > > If Compromise were the CWs (sincere CW), then your sincere ranking > obviously can't hurt her win as CWv. > > If someone else is CWs, then, in Condorcet(wv) the only way you could > chanage the winner from hir to Compromise would be by burial. (In > Condorcet(margins) you could do it by truncation.) > > So there has to be a cycle. > > Sincere cycles are evidently rare (at least in the current (probably > inauthentic) political system. Maybe in an authentic political system (one > free of media disinformation and count-fraud) there wouldn't be a > 1-dimensional political spectrum, and cycles wouldn't be unusual. > > Anyway, there are insincere cycles too. Strategic cycles. Burial strategy > is by no means out of the question. Burial will make a cycle...like the one > that Condorcet's FBC failure needs. > > We've discussed, on EM, how you can protect the CWs from burial (in wv or > margins) by moving that CWs up to 1st-place. But, if not enough people do > that, then you could need to rank the CWs over your favorite, to prevent > successful burial of the CWs under your favorite. > > Yes, if the method is Condorcet(wv) (but not Condorcet(margins), the > burial can likewise be thwarted (& penalized) if the people who prefer the > CWs to the buriers' candidate refuse to rank the buriers' candidate. > > Of course then they need to know or guess who's the CWs, and who's going > to bury. Yes, that's predictive information. Their defensive of the CWs is > tactical strategy that depends on good predictive information. > > If you're going to use Condorcet, then you'd better be good at tactical > strategy, and have good predictive information. > > (Well, admittedly, the fact that you'll truncate the (guessed) buriers' > candidate if you prefer the (guessed) CWs to hir can (but might not) deter > burial, because the would-be buriers don't have better predictive > information than you do.) > > But, if the expected or evident CWs (CWse) is in the would-be buriers' > bottom-set, then they don't really care if their burial is penalized by the > election of a different (negligibly worse) bottom-set member. > > What happens then? A burial-fest resembling a sharks' feeding-frenzy. I > call it the "perpetual burial fiasco". > > 'Round about that time, you'd be wishing that the election were an > Approval election. > > Let's not kid ourselves that Condorcet is free of strategy. > > So: > > You can change the winner from Compromise to Worst by moving Favorite to > top alongside Compromise. ...but also especially more likely by ranking > Favorite over Compromise. > > As I said, burial strategy isn't out of the question, and the need to vote > insincerely, as described above, isn't out of the question either. > > That's why some of us offer ranking-methods that meet FBC. > > That's why IC Condorcet is offered, to bring FBC to Condorcet. > > For instance, IC Condorcet could consist of IC RP. > > RP(wv) is much better than RC(margins), because RC(wv) is completely > truncation-proof, whereas RC(margins) leaves the CWs completely vulnerable > to truncation, whether strategic or innocent. > > Also, RP(margins) loses FBC in more scenarios than does RC(wv). > > Steve Eppley's MAM is just RC(wv) with Steve's clever way of avoiding some > problems and meeting more criteria in small electorates. > > So, if you insist on Condorcet (and are therefore willing to give up > chicken-dilemma protection), then consider IC MAM. > > The version of Condorcet Criterion that IC MAM meets is arguably better > and more meaningful than the ordinary Condorcet Criterion. > > ...but while IC Condorcet meets FBC, it still has chicken-dilemma, as does > Approval. > > So we offer MDDA(pt/2), MMPO(pt/2), and IC,MMPO. > > ...which meet FBC and don't have a chicken-dilemma problem. > > [Replying a little farther down] : > > >> > Approve only your favorite if you insist on voting tactically (based on >> > predictive information) and if it looks as if your favorite has a win. >> > >> > Is that a problem to you? >> >> **what** predictive information?? the polls??? have you learned nothing >> in the last two weeks? >> > > There was nothing for me to learn in the past 2 weeks, because what > happened in the past 2 weeks exemplifies what I've been saying all along. > > You haven't been reading my postings, or you'd know that I'm the one who > has been saying that the media's intentionally disinformational predictive > information is no good. > > So, in Approval, forget all about tactical voting and predictive > information. > > I said, "Just vote (only) for your strong top-set". > > By the definition of the strong top-set, electing from it, instead of your > strong bottom-set, is incomparably more important than the matter of > _which_ strong top-set wins. > > And, if you insist on trying to choose among your strong top-set, then you > increase the probability of electing from your strong bottom-set. > > And if you don't have strong top & bottom sets? You do, in our state & > national elections (the elections that most matter), at least in our > current, probably phony, political system. > > Likely, in a better world, the kind that has honest media and > vote-counting, your bottom-set candidates & parties would no longer be in > the elections. The bottom-set is likely an artifact of a phony political > system. > > But, if you didn't have strong top & bottom sets, then what? > > Then, maybe, under those conditions we'd have honest media, and good > predictive information, for one thing. ...on which to base good tactical > expectation-maximizing strategy. > > But what if there were no strong top & bottom sets, and also there were no > good predictive information? > > 1. Look, if it's difficult to decide what set of candidates you consider > top-set, the better set that deserves approval, then it doesn't make a > whole lot of difference how you choose between those sets that you're > indifferent between. So don't worry about it. If all the candidates are so > good that there's no strong bottom-set, then just be glad of that, and take > your best guess for your approval-set, because it doesn't matter much under > those conditions. > > 2. That said, there's a way to make an especially good guess: There's > 0-info expectation-maximizing strategy. ...but only if you want to use > it. I emphasize that you needn't bother with it. It's only for people who > like strategy and want to use it (though they don't need to), to make their > best guess. But one guess is reallly as good as another when you're > basically indifferent between the various possible approval-sets. > > > But there's no reason to believe that media disinformation, & media > exclusion of candidates & policy-solutions other than those of the 1% are > the only tactic in use. With our unverifiable vote-count, and the plentiful > evidence of presidential election count-fraud in 2000 & 2004, there's no > reason to believe that count-fraud isn't a main routine tactic of our > rulers. > > I mean, given all the mass killing that the 1% (or less) ruling-families & > their employees don't mind doing to gain more power & more wealth, does > anyone think that they'd be willing to let themselves be voted out of > office, or to take any chance of that? Why should they leave anything to > chance, or to us voters? Count-fraud, at least whenever needed, is an > obvious part of their optimal strategy. > > You wrote: > >> voters shouldn't have to worry about "predictive information" in casting >> their vote. they should not have to face tactical voting in the polls in >> any common case. >> > Exactly. That's one reason why like Approval, and why I agree with David > that it's the best method (except that chicken-dilemma protected methods do > bring some improvement). > > I discussed that in more detail a few paragraphs above, in my previous > long comment just before this one. > > This posting is already long, so I'll just post it now, as Part 1. > > Part 2 will follow, tomorrow. > > I re-emphasize, don't shut me out just yet, because I recognize that the > discussion of the choice between ranking-methods is more important than the > discussion of Approval vs ranking. ...and my post on the choice between > ranking-methods will follow the replies to the posting that I'm now > replying to. > > Michael Ossipoff > > (This concludes Part 1 of this reply. Part 2 of this reply will be along > tomorrow morning.) > > (I don't have a convenient way to delete text, but all of this postings > replies are above this point). > > > >> >> >> >> >> > >> > Ii suggest that if you're a progressive, then there are a large set of >> > candidates whom you so much prefer to the alternatives, that your best >> > strategy is to approve them all. >> >> that is not the fact. how can you tell that to a naive voter, "just >> approve them all"??? >> >> >> >> >> Voters don't want their second-choice to beat their first-choice. How >> do >> >> they stop that from happening? (By bullet-voting their first choice, >> and >> >> that devolves Approval into Plurality.) >> >> >> > >> > You're worrying too much about voting _among_ the best candidates. >> >> the voter shouldn't have to worry AT ALL. let the voter just inform the >> poll what their contingency votes are. >> >> >> >> >> >> >If >> > there's a set of candidates who are qualitatively the best, incomparably >> > better than the rest, then your best strategy is to maximize the >> > probability that one of them will win. That doesn't include voting >> _among_ >> > them. Just vote _for_ them. >> >> and maybe the voter will find out they helped their less-preferred >> candidate beat their more-preferred candidate. sure, that'll make the >> voter feel like their vote is effective! >> >> >> > In any case, I don't recommend tactical voting (voting based on >> predictive >> > information)...partly because the corporate mass-media's predictive >> > information is disinformational, >> >> no shit. but that still does nothing to relieve the voter of the >> tactical pressure because they **want** their favorite candidate to win, >> they don't want their contingency candidate to beat their favorite. but >> they also want their contingency candidate to beat their most-hated >> candidate if the election comes down to that. Approval does **nothing** to >> help them express that contingency. >> >> >in the service of the people who "own" the >> > mass media. >> > >> > ...but also partly because it's most to your advantage to just approve >> the >> > whole set of candidates that's qualitatively, incomparably, better than >> the >> > alternatives. ...to maximize the probability that one of that strong >> > top-set will win. >> > >> > (But until we have verifiable vote-counting, of course the voting >> system is >> > irrelevant, as is the voting.) >> >> totally different topic. (everyone should adopt optical-scan paper >> ballots.) >> >> precinct summability is necessary. IRV and Bucklin don't have it, but >> Condorcet (as well as Plurality) do have it. >> >> >> > >> > >> > >> > >> > >> >> > Equal-rank approval votes mean >> >> > voters don't get the option to express their preferred ranking, but >> >> because >> >> > of that, they aren't at all motivated to bias their ranking >> dishonestly. >> >> >> >> That's baloney. They may very-well not Approve of their second-choice >> >> candidate (whom they do actually approve of) to protect their >> first-choice >> >> candidate. >> >> >> > If they strongly approve of their 2nd choice, but not their 3rd choice, >> > then they'll approve their 2nd choice. >> >> unless they worry about their 2nd choice beating their 1st choice. >> >> > In our actual candidate lineup, the >> > merit differences among the progressives are negligible in comparison to >> > the big gulf of a merit difference between them and the Republocrats. >> > >> > Don't worry about it so much. Just approve all of the best. >> >> you're repeated assurances are hollow. just like Rob Richie. >> >> >> >> >> > >> > >> > >> >> >> >> > They only have to decide which candidates to approve, or where the >> >> cut-off >> >> > is between approval and disapproval. >> >> >> >> No. That is a falsehood. >> >> >> >> They have to consider whether or not their second-choice candidate has >> a >> >> chance of beating their first-choice. And they have to consider >> whether or >> >> not their third-choice has a chance of beating either their >> second-choice >> >> or their first-choice. And then they have to decide if that risk is >> worth >> >> it to Approve their second or third choices. >> >> >> > No, they don't have to do any of that. They don't have to try to vote >> > tactically. >> >> bullshit. you have no right to tell a voter what his or her interests >> are. they **may** have to consider voting tactically because they want >> their favorite candidate to win. but, if their favorite is a poor showing, >> they also don't want their contingency candidate to lose to their hated >> candidate. >> >> telling them "You don't have to try to vote tactically" is hollow. it's >> like saying "Don't worry. Any of the candidates are okay." >> >> >> >> > Partly because there's no reliable information on which to base >> > tactical voting, >> >> which is just one reason we should relieve voters of that burden. >> tactical voting is **not** a perk. it's a burden. no one feels good >> about it. >> >> > and partly because their best strategy is just to approve >> > all of the set that is incomparably, qualitatively better than the other >> > candidates. >> >> says YOU. >> >> you cannot tell voters what their best strategy is and they find out >> later that following your glib assurance screwed them. that is what >> FairVote did with Burlington and there was a city that was quite pissed. >> we called it the "IRV Happy Talk". >> >> you are no authority to advise voters what their best strategy is. the >> best we can do is give them an voting system that normally requires no >> strategy of voters nor benefits voters from using strategy. after a few >> generations, Duverger's Law will be forgotten and they can confidently >> simply vote their political expression. >> >> >> > >> >> >> >> >> >> >> >> Robert says: >> >> >> > >> > >> > I think most of your gushing for Approval voting is not well-founded. >> What >> > if the voter prefers Candidate A as their favorite, but suspects that >> > Candidate B and Candidate C are the leading candidates? How should that >> > voter vote with an Approval ballot? Please answer that question with >> > specifics. >> > >> > [endquote] >> > >> > We'd need better "specifics" about the situation that you've >> incompletely >> > described. It depends on how much you like A better than B & C, and how >> > much you like A & B better than C. >> > >> >> NO, we DON'T. that is your failure. you cannot force voters to research >> these "specifics". >> >> All the voter knows is that he/she prefers A above all others. If he/she >> can't get A, then they'll take B. If they can't get A or B, then they'll >> settle for C. But this voter really doesn't want to help candidate D at >> all. The only ballot that expresses the will of this voter is the Ranked >> Ballot. Approval does not gather enough information (forcing the voter to >> vote tactically) and Score requires too much information (also forcing the >> voter to score tactically). With the exceptions of goofy things coming out >> of a cycle (which would be rare), only Condorcet resolves that election in >> exactly the same manner as would a Simple Majority race between two >> candidates with equal weighting for all voters. Only Condorcet does that. >> >> >> >> > You don't mention any set's qualitative or incomparable better-ness,and >> you >> > suggest that there's good predictive information that the winner will >> be B >> > or C. >> >> i am saying the voters should not be required to have "good predictive >> information" for them to vote their interests. And no political group >> should benefit from "good predictive information". That's the Duverger >> issue. It's because, with plurality voting, "good predictive information" >> tells voters that one of the two major party candidates is expected to win, >> so that their vote for third-party or independent is wasted. >> >> >> > So, if you don't have a strong top or bottom set, and the merit >> differences >> > between your successive choices are uniform, and if there's reliable >> > predictive information that the winner will be B or C, then of course >> > approving {A,B} is a good way to vote. >> >> But you DON'T necessarily know that and you shouldn't burden the voter >> with that tactical need. They shouldn't have to be thinking "Hmmmm, what >> would be a good way to vote to benefit my political interest?" That's your >> problem and you do NOTHING to address it. >> >> >> > But in our actual political lineup, the media have people believing that >> > there are 2 parties that always must and will win. Say your favorite is >> > Jill Stein, and that the other 2 candidates are Hillary & Donald. You're >> > told that only Hillary &I Donald are well-liked by everyone enough so >> that >> > one of those two beloved and popular candidates is sure to win. >> > >> > For one thing, you should find something fishy about such a claim, >> >> You are expecting too much from voters. Voting should not be a political >> science exam. It's a **service** to voters, not a **burden** for voters. >> Or at least it should be. >> >> > given >> > what is known about those 2 candidates being the most despised >> presidential >> > candidates in this country's history. >> > >> > For another thing, the merit of Jill is so much greater than that of >> > Hillary or Donald that, no matter what your'e told or believe about >> > predictive information, your best strategy is to vote for the only >> > acceptable, honest and progressive, genuinely pro-people candidate. >> > >> > That latter interpretation of your scenario is the one that you probably >> > intended, because it's the one that actually obtains in elections here. >> >> with your own words, you are making it clear that voters will have to >> explicitly consider strategy when voting in an Approval system. >> >> Again, you cannot simply answer the question: "What should the voter do >> with their second-favorite candidate?" >> >> With Ranked-Choice Voting, the answer is clear (they mark their >> second-favorite with a "2"). >> >> With Approval Voting, the voter has a tactical burden deciding whether or >> not to approve their second-choice. And with Score Voting, the voter has a >> tactical burden deciding how much to score their second-choice. >> >> Mike, you did nothing but verify my objections to Approval voting for >> governmental elections. >> >> >> -- >> >> >> >> >> r b-j rbj@audioimagination.com >> >> >> >> >> "Imagination is more important than knowledge." >> >> ---- >> Election-Methods mailing list - see http://electorama.com/em for list >> info >> >> >