Re: [EM] I need an example of Condorcet method being subjected to the spoiler effect if any
On Jan 21, 2010, at 9:29 AM, Kristofer Munsterhjelm wrote: robert bristow-johnson wrote: i think that the answer is no, if a Condorcet winner exists and that all bets are off if a CW does not exist, except, perhaps for these strategy-resistant methods such as Markus Schulze's method. i sorta understand it, but since he hangs here, i think Markus should address the question if the Schulze method is spoiler free. Most Condorcet related problems occur only when there is no Condorcet winner (i.e. there is a top level cycle in the group preferences). Sincere or artificially generated cycles are the root cause of both problems with sincere votes (e.g. spoiler related problems) and strategic voting related problems in Condorcet. Different Condorcet methods (e.g. Ranked Pairs, Schiulze) are quite similar in the sense that the basic vulnerabilities of Condorcet methods exist in all of them (e.g. the basic burial scenarios). Their differences between the most common Condorcet methods are quite small in the sense that in real life elections they almost always elect the same candidate. Their differences are mostly related to how well they can resist strategic voting. Another point of view is to compare which method elects the best/correct winner with sincere votes. What is good in all the common Condorcet methods is that their vulnerabilities to strategies (and their differences in general) may very well be so small in typical real elections (large, public, with independent voter decision making, with changing opinions and less than perfect poll information) that strategic voting will not be a problem. Also their differences with sincere votes (e.g. spoiler related problems) are quite small in real life elections. Here's one simple spoiler related example as a response to Kathy Dopp's request. 35: ABC 33: BCA 32: CAB I this example there are three candidates in a top level cycle. If any of the candidates would not run that would mean that there is a Condorcet winner, and that winner would be different in each case. Let's say that the method we use will pick A as the winner. If B would not run then the votes would be 35: AC, 33: CA, 32: CA and C would win. B is thus a spoiler from C's point of view. I note however that these spoiler cases in Condorcet are not as common as in many of the other methods. In practice it may be that there is no need to worry about these cases. Maybe Kathy Dopp's comparisons will reveal something about how problematic the spoiler effect is or is not in Condorcet. Not also that in the example above B was not a minor party candidate (often term spoiler refers to minor candidates) but a pretty strong candidate. MAM/Ranked Pairs is also pretty strategy-resistant and is easier to understand. Schulze has the advantage of producing better results in some cases (closer to Minmax), but if ability to describe to the public is important, then Ranked Pairs wins there. Could someone please provide me with an example of the spoiler effect occuring with the Condorcet method of counting rank choice ballots or tell me why the spoiler effect doesn't happen with Condorcet in a few words? (...) i would say that (with the CW existing), it's spoiler-proof. Yes. If there's a CW and a candidate is added, and that candidate doesn't create a cycle, then the winner doesn't change. All the tricky stuff happens when there is a cycle, or the candidate makes one. The advanced methods can claim further resistance: Schulze and Ranked Pairs both make their winner decision independent of candidates not in the Smith set. River is independent of Pareto- dominated alternatives - a candidate is Pareto-dominated if everybody who ranks both him and some other (specific) candidate, rank the other candidate above him (e.g. X is Pareto-dominated by Y if all voters who rank both X and Y say YX, and there's at least one such voter). I imagine these resistances would mostly come into play in smaller elections. Still, they're nice to have, and their existence immediately tells parties not to try exploiting certain weaknesses (because it won't work). Yes, in small elections (with few voters only) it may be possible to know the opinions of each voter and agree about the applied strategy with the strategizing voters. In typical large real life elections many of the vulnerabilities are not practical and sincere voting may be the best strategy to most if not all voters. Many of the criteria would be nice to have. One must however remember that often they have two sides. Winning something in some area may mean losing something in another area (e.g. the LNH property of IRV has been discussed widely on this list recently) especially when trying to fix the last remaining problems of the Condorcet methods. And if one assumes that strategic voting will not be meaningful in the planned
Re: [EM] I need an example of Condorcet method being subjected to the spoiler effect if any
Thanks Kristofer for the explanations. Do you know a good place that discusses the Ranked Pairs method of resolving cycles, or all the methods of resolving cycles? I would still like an example of a spoiler in Condorcet no matter how unlikely if possible. Thank you. Kathy On Thu, Jan 21, 2010 at 2:29 AM, Kristofer Munsterhjelm km-el...@broadpark.no wrote: robert bristow-johnson wrote: i think that the answer is no, if a Condorcet winner exists and that all bets are off if a CW does not exist, except, perhaps for these strategy-resistant methods such as Markus Schulze's method. i sorta understand it, but since he hangs here, i think Markus should address the question if the Schulze method is spoiler free. MAM/Ranked Pairs is also pretty strategy-resistant and is easier to understand. Schulze has the advantage of producing better results in some cases (closer to Minmax), but if ability to describe to the public is important, then Ranked Pairs wins there. Could someone please provide me with an example of the spoiler effect occuring with the Condorcet method of counting rank choice ballots or tell me why the spoiler effect doesn't happen with Condorcet in a few words? (...) i would say that (with the CW existing), it's spoiler-proof. Yes. If there's a CW and a candidate is added, and that candidate doesn't create a cycle, then the winner doesn't change. All the tricky stuff happens when there is a cycle, or the candidate makes one. The advanced methods can claim further resistance: Schulze and Ranked Pairs both make their winner decision independent of candidates not in the Smith set. River is independent of Pareto-dominated alternatives - a candidate is Pareto-dominated if everybody who ranks both him and some other (specific) candidate, rank the other candidate above him (e.g. X is Pareto-dominated by Y if all voters who rank both X and Y say YX, and there's at least one such voter). I imagine these resistances would mostly come into play in smaller elections. Still, they're nice to have, and their existence immediately tells parties not to try exploiting certain weaknesses (because it won't work). -- Kathy Dopp Town of Colonie, NY 12304 phone 518-952-4030 cell 518-505-0220 http://utahcountvotes.org http://electionmathematics.org http://kathydopp.com/serendipity/ Realities Mar Instant Runoff Voting http://electionmathematics.org/ucvAnalysis/US/RCV-IRV/InstantRunoffVotingFlaws.pdf Voters Have Reason to Worry http://utahcountvotes.org/UT/UtahCountVotes-ThadHall-Response.pdf Checking election outcome accuracy --- Post-election audit sampling http://electionmathematics.org/em-audits/US/PEAuditSamplingMethods.pdf Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] I need an example of Condorcet method being subjected to the spoiler effect if any
On Jan 21, 2010, at 1:03 AM, Juho wrote: What is good in all the common Condorcet methods is that their vulnerabilities to strategies (and their differences in general) may very well be so small in typical real elections (large, public, with independent voter decision making, with changing opinions and less than perfect poll information) that strategic voting will not be a problem. Also their differences with sincere votes (e.g. spoiler related problems) are quite small in real life elections. Here's one simple spoiler related example as a response to Kathy Dopp's request. 35: ABC 33: BCA 32: CAB I this example there are three candidates in a top level cycle. If any of the candidates would not run that would mean that there is a Condorcet winner, and that winner would be different in each case. Let's say that the method we use will pick A as the winner. If B would not run then the votes would be 35: AC, 33: CA, 32: CA and C would win. B is thus a spoiler from C's point of view. I note however that these spoiler cases in Condorcet are not as common as in many of the other methods. In practice it may be that there is no need to worry about these cases. Maybe Kathy Dopp's comparisons will reveal something about how problematic the spoiler effect is or is not in Condorcet. Not also that in the example above B was not a minor party candidate (often term spoiler refers to minor candidates) but a pretty strong candidate. In that case it might be a good starting point to define spoiler, so we know what we've found when we find it. What's an example of an IRV spoiler who's not a pretty strong candidate? Election-Methods mailing list - see http://electorama.com/em for list info
[EM] Strong Minmal Defense, Top Ratings
In a recent EM post in another thread, I defined and recommended the Strong Minimal Defense, Top Ratings method (that I first proposed in 2008) as the best of the methods that meet the Favourite Betrayal criterion, and also the best 3-slot ballot method: *Voters fill out 3-slot ratings ballots, default rating is bottom-most (indicating least preferred and not approved). Interpreting top and middle rating as approval, disqualify all candidates with an approval score lower than their maximum approval-opposition (MAO) score. (X's MAO score is the approval score of the most approved candidate on ballots that don't approve X). Elect the undisqualified candidate with the highest top-ratings score.* I gather from one off-list response that this sentence of mine could have been more clear: 'Unlike MCA/Bucklin this fails Later-no-Help (as well as LNHarm) so the voters have a less strong incentive to truncate..' I neglected to mention that I think it is desirable that after top-voting X, ranking Y below X (but above bottom) should be about equally likely to help X as to harm X. This implies that if one of the the two LNhs are failed, it is desirable that the other is also. MCA/Bucklin meets Later-no-Help while failing Later-no-Harm. The voters have a big incentive to truncate, and to equal-rank at the top, so with strategic voters it tends to look like plain Approval. In SMD,TR after top-rating X, middle-rating Y may harm X or may help X. As discussed in 2008, it fails Mono-add-Top (and so Participation). 8: C 3: F 2: XF 2: YF 2: ZF F wins after all other candidates are disqualified, but if 2 FC ballots are added C wins. Of course it is far from uniquely bad in that respect. A big plus for it is that it is virtually alone in meeting my proposed Unmanipulable Majority strategy criterion: Regarding my proposed Unmanipulable Majority criterion: *If (assuming there are more than two candidates) the ballot rules don't constrain voters to expressing fewer than three preference-levels, and A wins being voted above B on more than half the ballots, then it must not be possible to make B the winner by altering any of the ballots on which B is voted above A without raising their ranking or rating of B.* http://lists.electorama.com/pipermail/election-methods-electorama.com/2008-December/023530.html In common with MCA it meets mono-raise (aka ordinary monotonicity) and a 3-slot ballot version of Majority for Solid Coaltions, which says that if majority of the voters rate a subset S of the candidates above all the outside-S candidates, the winner must come from S. From the post that introduced SMD,TR: It is more Condorcetish and has a less severe later-harm problem than MCA, Bucklin, or Cardinal Ratings (aka Range, Average Rating, etc.) 40: AB 35: B 25: C Approval scores: A40, B75, C25 Approval Opp.: A35, B25, C75 Top-ratings scores: A40, B35, C25 They elect B, but SMD,TR elects the Condorcet winner A. Chris Benham __ See what's on at the movies in your area. Find out now: http://au.movies.yahoo.com/session-times/ Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] I need an example of Condorcet method being subjected to the spoiler effect if any
well, since no one else responded... On Jan 21, 2010, at 9:21 AM, Kathy Dopp wrote: Thanks Kristofer for the explanations. Do you know a good place that discusses the Ranked Pairs method of resolving cycles, or all the methods of resolving cycles? Wikipedia. maybe start with http://en.wikipedia.org/wiki/ Voting_systems and see what you find from there. I would still like an example of a spoiler in Condorcet no matter how unlikely if possible. the spoiler would have to either push a nicely resolved election with a CW into a cycle and have the cycle resolved so that the earlier CW does not win or, if the election was in a cycle in the first place (but resolved to pick some winner), the spoiler would have to cause the election algorithm to choose a different winner (and not the spoiler). -- r b-j r...@audioimagination.com Imagination is more important than knowledge. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] I need an example of Condorcet method being subjected tothe spoiler effect if any
Jonathan makes an important point. The term spoiler means a minor candidate with a small percentage of the vote, who changes which of the other candidates wins by running. But Kathy and some others wish to expand the definition to include a front-runner. (Note that these IRV opponents refer to the top plurality vote-getter who narrowly lost the runoff in Burlington, Kurt Wright, as a spoiler who prevented the candidate in third place from winning. This is a dynamic worthy of analysis, but the word spoiler is never used by the media or political scientists when describing the plurality leader. Terry Bouricius - Original Message - From: Jonathan Lundell jlund...@pobox.com To: Juho juho4...@yahoo.co.uk Cc: EM Methods election-methods@lists.electorama.com Sent: Thursday, January 21, 2010 10:33 AM Subject: Re: [EM] I need an example of Condorcet method being subjected tothe spoiler effect if any On Jan 21, 2010, at 1:03 AM, Juho wrote: What is good in all the common Condorcet methods is that their vulnerabilities to strategies (and their differences in general) may very well be so small in typical real elections (large, public, with independent voter decision making, with changing opinions and less than perfect poll information) that strategic voting will not be a problem. Also their differences with sincere votes (e.g. spoiler related problems) are quite small in real life elections. Here's one simple spoiler related example as a response to Kathy Dopp's request. 35: ABC 33: BCA 32: CAB I this example there are three candidates in a top level cycle. If any of the candidates would not run that would mean that there is a Condorcet winner, and that winner would be different in each case. Let's say that the method we use will pick A as the winner. If B would not run then the votes would be 35: AC, 33: CA, 32: CA and C would win. B is thus a spoiler from C's point of view. I note however that these spoiler cases in Condorcet are not as common as in many of the other methods. In practice it may be that there is no need to worry about these cases. Maybe Kathy Dopp's comparisons will reveal something about how problematic the spoiler effect is or is not in Condorcet. Not also that in the example above B was not a minor party candidate (often term spoiler refers to minor candidates) but a pretty strong candidate. In that case it might be a good starting point to define spoiler, so we know what we've found when we find it. What's an example of an IRV spoiler who's not a pretty strong candidate? Election-Methods mailing list - see http://electorama.com/em for list info Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
WARNING: this is a metacommunication, about the communication process here and elsewhere in voting system advocacy, not about voting methods, per se. At 01:48 AM 1/21/2010, robert bristow-johnson wrote: On Jan 20, 2010, at 11:23 AM, Abd ul-Rahman Lomax wrote: Variation on previous post. Silly time! At 02:31 PM 1/16/2010, robert bristow-johnson wrote: On Jan 16, 2010, at 12:05 PM, Abd ul-Rahman Lomax wrote: Robert, your slip is showing. what slip? i don't have nuttin' under me kilt. We already knew that. you do? you keep saying that you can see it. Yes, I said that. Slip is showing is a metaphor, stating that something relatively unmentionable is visible. I can see something. Others can see something. Do you see or know what we see? Perhaps you do, but you are defending yourself as if you cannot see it. Others who do see it might respond differently. This is meta talk, it's about the communication, not election methods. I will therefore limit it to what's relevant to the *extended* purposes of this list, which include voting system advocacy, not merely theory. If you are going to be a public advocate, you will be much more effective if you know how your actions and words will be seen, and if you can learn as much as possible about debate tactics and strategy. Silly hat, Off. Robert, if you want to be effective in public debate, what makes you think i'm not effective? do you actually think you were effective? 1. I suspect you are less effective than you can be. You get caught, easily, in irrelevancies, distracting from the central points to be conveyed. As a public activist, to be effective, you must use polemic and all the skills of advocacy, which is different from discussion. Here, we discuss, and no collective decisions are actually made here. However, I inferred from behavior here what might happen in a public debate. If, in fact, some of this behavior carries over to public debate, you could get creamed. Unnecessarily. That is, over your own style and personality, not over the issue you are advocating. 2. Was I effective? In what? I'm engaged only in a diffuse kind of advocacy here. However, I've also repeated ideas that I've expressed here many times, and this is part of my own learning and polishing process. This is of benefit to those who find it useful to follow my discussions, to explore these topics repeatedly so that they become familiar, and so that deeper understanding spreads. It's my method and approach, and it certainly is not for everyone. Were I to do in a public forum, not a specialized forum like this, what I do here, I'd almost completely fail. (3.) I have, however, come to the point that I'm sufficiently familiar with the issues that I'd engage, if invited, in public debate. I'm an effective speaker, making clear and direct contact with the audience. We'll see if that happens. I have made blog posts in public fora on these issues, they are far briefer, in general. The effort per word and per message is much higher for them. i won't slap on the argumentum verbosium and explode the debate about a single testable issue (like how many piles one needs if there are 3 candidates) into pages and pages, that when i responded, my post was rejected by the list server as too large. Oh, we are crushed at the loss actually, usually it isn't exactly rejected, it is held for moderator approval, which can take some time. Depends. I'd suggest avoiding setting up an immediate victory by the other side by feeding him or her lines like that. you're the one feeding lines. Sure. Like a debate opponent might. Your slip is showing is a metacommunication to the audience, calling attention very briefly to the opponent's behavior, or sometimes to an issue of substance (possibly). As an ad hominem argument, it's irrelevant, but in real debate, it could be very important. People respond to the person, usually, more than to the substance. They judge the substance by the person. Only in careful deliberative process is this effect reduced much. who brought up the slip showing in the first place? Me. A stand-in for your debate opponent. However, it wasn't intended as a debate tactic, but as personal advice, which you could take or leave. You took it, in fact, but as if it were bait in a debate, and you also took, therefore, the hook and the line. And in so doing, you got jerked out of the water. My judgement. Yours might be different, but if you really want to know, ask someone neutral. how does one respond when facing: Your slip is showing, now onto a verbose response that does not speak to the core factual issues at all. How? It's actually terminally easy. No response at all is probably the most efficient. A quick joke, though, may be even more efficient. Learn to think on your feet, if you have to puzzle over this, no response is better. Robert, your slip is showing was very efficient for me, it took,
Re: [EM] IRV vs Plurality (back to the pile count controversy)
On Jan 21, 2010, at 3:24 PM, Abd ul-Rahman Lomax wrote: WARNING: this is a metacommunication, about the communication process here and elsewhere in voting system advocacy, not about voting methods, per se. At 01:48 AM 1/21/2010, robert bristow-johnson wrote: On Jan 20, 2010, at 11:23 AM, Abd ul-Rahman Lomax wrote: Variation on previous post. Silly time! At 02:31 PM 1/16/2010, robert bristow-johnson wrote: On Jan 16, 2010, at 12:05 PM, Abd ul-Rahman Lomax wrote: Robert, your slip is showing. what slip? i don't have nuttin' under me kilt. We already knew that. you do? you keep saying that you can see it. Yes, I said that. Slip is showing is a metaphor, stating that something relatively unmentionable is visible. I can see something. Others can see something. Do you see or know what we see? Perhaps you do, but you are defending yourself as if you cannot see it. Others who do see it might respond differently. better say what it is right now, or you're just blowing smoke (to make use of another metaphor). 1. I suspect you are less effective than you can be. You get caught, easily, in irrelevancies, distracting from the central points to be conveyed. it wasn't me that amplified the length of text by a factor of 10. i was trying to keep it focused and my mistake was responding to your asides. As a public activist, to be effective, you must use polemic and all the skills of advocacy, which is different from discussion. Here, we discuss, and no collective decisions are actually made here. However, I inferred from behavior here what might happen in a public debate. If, in fact, some of this behavior carries over to public debate, you could get creamed. Unnecessarily. That is, over your own style and personality, not over the issue you are advocating. blather. (quoting Warren Smith.) 2. Was I effective? In what? I'm engaged only in a diffuse kind of advocacy here. However, I've also repeated ideas that I've expressed here many times, and this is part of my own learning and polishing process. This is of benefit to those who find it useful to follow my discussions, to explore these topics repeatedly so that they become familiar, and so that deeper understanding spreads. It's my method and approach, and it certainly is not for everyone. Were I to do in a public forum, not a specialized forum like this, what I do here, I'd almost completely fail. more blather. (3.) I have, however, come to the point that I'm sufficiently familiar with the issues that I'd engage, if invited, in public debate. I'm an effective speaker, making clear and direct contact with the audience. We'll see if that happens. I have made blog posts in public fora on these issues, they are far briefer, in general. The effort per word and per message is much higher for them. sometimes effective public speakers are successful not because of their efforts to focus the issue, but because of their efforts to distract. e.g. Sarah Palin. i won't slap on the argumentum verbosium and explode the debate about a single testable issue (like how many piles one needs if there are 3 candidates) into pages and pages, that when i responded, my post was rejected by the list server as too large. Oh, we are crushed at the loss actually, usually it isn't exactly rejected, it is held for moderator approval, which can take some time. Depends. i'm not messing with it further. i just ask that you don't amplify the quantity of responses by a factor of 10 and bring your post to 40K so that if anyone actually bothers to read through it and respond to most or all of the points, their effort goes into the trash can. since your name was in the To: header, you got that response, but no one else did. what i have learned from that is to not play your argumentum verbosium game. from now on, i must pick and choose, respond to only one point, delete all the other blather, and keep the issue focussed. thus i am deleting and not bothering to engage in the other text. care to discuss how many piles one needs (for precinct summability) when there are N candidates? or N credible candidates? that's what the issue was before it was buried in blather. -- r b-j r...@audioimagination.com Imagination is more important than knowledge. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
People, please. This is not a debate class, and even if it were, no, I won is really useless even if true. Please take this discussion off list, if you find it important enough not to stop. There's practically no voting system content left. As for what is left: we all know that the number of piles is large, that full ballots can be transmitted, and we can work out the implications to our own perhaps-incorrect satisfaction. Respectfully to you both, but tired of this wordy debate, Jameson Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
i just want to settle the issue about how many piles one needs to be precinct summable when there are N candidates. Kathy was pointing to Abd ul as the qualified actor who refuted the falsifiable assertion that i made that you needed only 9 piles for 3 candidates. She repeated labeled (without any justification other than citing Abd ul's blather) the math that i clearly presented as illogical. Abd ul did nothing to support Kathy's assertion. Kathy, fancying herself as an election security expert, continues to try to taint IRV as being insecure because it's not precinct summable. and that is a demonstrably false claim. i'll leave it to the experts here to judge who was trying to stay on topic and who was decreasing the signal-to-noise ratio with unnecessary text (with aim to distract from the core issue and to denigrate the other side). r b-j On Jan 21, 2010, at 3:56 PM, Jameson Quinn wrote: People, please. This is not a debate class, and even if it were, no, I won is really useless even if true. Please take this discussion off list, if you find it important enough not to stop. There's practically no voting system content left. As for what is left: we all know that the number of piles is large, that full ballots can be transmitted, and we can work out the implications to our own perhaps-incorrect satisfaction. Respectfully to you both, but tired of this wordy debate, Jameson -- r b-j r...@audioimagination.com Imagination is more important than knowledge. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
At 01:48 AM 1/21/2010, robert bristow-johnson wrote: the fact is, transmitting the content (to a central counting location) of *every ballot* is the transfer of a finite amount of information. that is even *more* general than sorting to piles and transmitting the tallies for piles. Yes, of course. And this is an equivalent to carrying all the ballots to a central location, merely, if done, say, over the internet, faster. But ... it raises some security issues. And with central counting there are other issues. This is a red herring, because we are talking about precinct summability, and when the number of candidates is very small, precinct summability isn't relevant, because the raw ballot data may be transmitted. So, back to the real question: is precinct summability an important practical criterion to be applied to voting systems? A related question is the sensitivity of the method to small variations in votes. Noise, if you will. That can be seen with Yee diagrams, in the presence of chaotic regions in issue space, with IRV. But I won't address that here, beyond noting that IRV multiples the probability of ties, and many of the ties will drastically flip the overall result. With most other methods, there is only one relevant tie possible (beyond extraordinarily rare three-way ties) and when this happens, a coin flip doesn't change the expected voter satisfaction much, if at all. With IRV, the effect can be enormous, because the tie can affect a candidate elimination before all the votes for that candidate have been counted. but breaking it down to piles regarding every conceivable permutation of candidate preference is *still* breaking it down to a finite number of piles. Sure. Finite. I'll point out that a google is finite. With computers, this can be done even with moderately large numbers of candidates. It's still a problem with voting security, though. I've argued for Public Ballot Imaging, which would make available actual ballot images, transmitted from polling places, perhaps by fax or more likely through digital camera images -- no touching of ballots necessary except by election officers, all visible openly --, independently by voting watchdog organization through election observers, so that anyone can verify the count in a precinct or as many as they care, or can even just check one serialized ballot (serialized before counting) and mark it as reviewed, in a system that collects and displays such reviews. Many details omitted here! for 3 candidates, that number is 9. Okay, three candidates, A, B, C, the ballot possibilities are, to be complete, much more than 9. I'll assume that write-ins are illegal and void the ballot. Some of the possibilities are legally equivalent to others, and in actual IRV ballot imaging, they are collapsed and reported the same, to the displeasure of voting security people who do want to know the error rate, which includes overvoting and exact overvoting patterns. So-called ballot images are not, generally. They are processed data reducing a ballot to legally equivalent votes. The reduced set is this: A B C AB AC BA BC CA CB Note that this assumes a 2-rank ballot. It also assumes that majority vote isn't important. If it's important, as it would be in an IRV election under Robert's Rules, we have some more possibilities. They are all the three-rank permutations. ABC ACB BAC BCA CAB CBA Each of these is equivalent, for the purposes of finding a plurality winner, to a two-candidate combination. if you or Kathy say it's 15, then you're wrong (and it's your slip that's showing). Well, I won't speak about Kathy, but in terms of practical elections in the U.S., she's right. You did not state enough information to establish your reduced count, and you actually added language that indicated that a larger total would be necessary. You used the qualifier credible to indicate that there might be candidates not credible, and you did not take care to define this. What you have asserted is true under two qualifications: there are only three candidates legally eligible to receive votes. And there are only two ranks on the ballot. If there are three ranks on the ballot, we have a poor situation, an invitation to voters to cast an irrelevant vote, if, in fact, that third rank has any effect on outcome, which, in the general case, it can. If it can affect outcome in some way, the piles must be reported separately. for 4 candidates the number of necessary piles is 40. Under the restricted conditions, perhaps. I haven't checked the math. I distrust formulas compared to exhaustive enumeration, they take more work and there is more room for error. My lists, which I provided before, showed what is shown above, though it may be better explained this time. The slip is an assumption that one's analysis is more complete than that of another, when it may be, instead, ignorant of some of the
Re: [EM] I need an example of Condorcet method being subjected
From: Terry Bouricius ter...@burlingtontelecom.net To: Jonathan Lundell jlund...@pobox.com, Juho juho4...@yahoo.co.uk Jonathan makes an important point. The term spoiler means a minor candidate with a small percentage of the vote, who changes which of the other candidates wins by running. But Kathy and some others wish to expand the definition to include a front-runner. (Note that these IRV opponents refer to the top plurality vote-getter who narrowly lost the runoff in Burlington, Kurt Wright, as a spoiler who prevented the candidate in third place from winning. This is a dynamic worthy of analysis, but the word spoiler is never used by the media or political scientists when describing the plurality leader. Terry Bouricius Thanks for all the information re. Condorcet cycles and the unlikely cases of spoilers in Condorcet method of counting rank choice votes from Juho and Robert. Terry, You cleverly conveniently change all the definitions whenver it is necessary to make yourself and Fairytale Vote right on the facts. Let's see what some of them are: A spoiler is *not* according to you, a nonwinning candidate whose presence in the election changes who would otherwise be the winner, but only a particular type of spoiler that is a minor candidate. A majority of voters is *not* according to you, a majority out of all voters who cast ballots, but only out of voters whose ballots have not been exhausted by the time the final IRV counting round is done. A majority candidate, according to you, is *not* the Condorcet winner who a majority (and indeed the most#) of voters favor above all other candidates, but only a candidate who wins a majority in round one, or in the final IRV counting round out of unexhausted ballots after the Condorcet winner and other more majority-favorite winners are eliminated. Terry, redefine any word you want to and you make yourself right, even if most people do not agree with your definitions. It's a good strategy to mislead the public like Fairytale Vote has done. Kathy -- Kathy Dopp Town of Colonie, NY 12304 phone 518-952-4030 cell 518-505-0220 http://utahcountvotes.org http://electionmathematics.org http://kathydopp.com/serendipity/ Realities Mar Instant Runoff Voting http://electionmathematics.org/ucvAnalysis/US/RCV-IRV/InstantRunoffVotingFlaws.pdf Voters Have Reason to Worry http://utahcountvotes.org/UT/UtahCountVotes-ThadHall-Response.pdf Checking election outcome accuracy --- Post-election audit sampling http://electionmathematics.org/em-audits/US/PEAuditSamplingMethods.pdf Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] Fw: Two simple alternative voting methods that are fairer than IRV/STV and lack most IRV/STV flaws
At 02:28 AM 1/21/2010, robert bristow-johnson wrote: On Jan 20, 2010, at 12:04 PM, Abd ul-Rahman Lomax wrote: At 12:52 AM 1/18/2010, robert bristow-johnson wrote: yes, it's debatable and, since there are 3 different methods all lifting up different declared winners, it's subjective. Well, it's subjective without preference strength information. the debate *might* go into the direct if whatever preference strength information is subjective or not. Once again, narrow interpretation believing that it trumps broader interpretation. preference strength information was here this referring to one of two possibilities: Expressed preference strength. Only certain ballots allow this, but if it is allowed, this information can be objectively used. Is the information subjective. Yes, in the sense that it represents a subjective judgment by the voters, but isn't that what all votes are? The *analysis* can be objective. Since we were discussing various methods of analyzing preference information provided on ballots, we are talking about analysis, not the individual voter process. Real preference strength. This can generally only be used in judging voting systems in simulations, where the preference strength is assumed, typically using models of voter preference, varied randomly according to some sensible distribution. There are possibilities where real preference strength can be measured, typically by setting up some cost to voting. A very relevant example is the cost of actually voting, the cost of turnout. If voters have low preference strength, they are less likely to turn out. Therefore turnout is a factor which indicates preference strength in real elections. We know that when voters have a high preference strength between two candidates, and there is a special cost to turnout, as in a special election or special runoff, voters turn out in unusual numbers. There are other proposed ways of increasing the cost of voting, most particularly the Clarke tax or variations on that model. This isn't the place to explore them, but only to note that preference strength was often neglected in developing and studying voting systems, on the bogus argument that it could not be measured or accurately expressed. That was a narrow understanding, substantially incorrect and even to the extent it was correct, it was misapplied. By people of the stature of Arrow candidate for mayor in Burlington VT in 2009, who also turned out to be the Condorcet winner in an IRV election, is named Andy Montroll. last name Montroll. with two L's no S nor E. if it were me, i would eventually be annoyed if someone consistently mispronounced or mispelled my name, even after the correct name has been offered earlier. Montroll. Chalk Montrose up to my age. Note: in a debate, you gain one point if you generously and courteously correct a minor error by your opponent. You lose ten points if you try to impeach your opponent for a minor error, either directly or indirectly, as by making a big fuss about it to underscore it, particularly with sarcasm or countersinking. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] I need an example of Condorcet method being subjected to the spoiler effect if any
robert bristow-johnson wrote: so, we have a CW... add a candidate, if that candidate does not become the winner, nor cause a cycle, then the Condorcet Winner we had before continues to be the CW with the added candidate. (boy, i guess we're rephrasing the same thing multiple times!) Yup. River is independent of Pareto-dominated alternatives - a candidate is Pareto-dominated if everybody who ranks both him and some other (specific) candidate, rank the other candidate above him (e.g. X is Pareto-dominated by Y if all voters who rank both X and Y say YX, and there's at least one such voter). i wouldn't mind if someone explains this. i don't know what Pareto-dominated is about. can someone expound? A is Pareto-dominated by B if all voters who express any difference in preference between A and B, prefer B to A. Note that a voter may simply leave both unranked - that wouldn't count towards either's Pareto dominance. If all voters equal-rank B and A, that doesn't count towards any Pareto-dominance, either. A is Pareto-dominated (period) if there's some other candidate by which it is Pareto-dominated. The strong Pareto criterion states that Pareto-dominated candidates shouldn't win. This makes sense, because say X won and was Pareto-dominated. Then people could (rightly) complain that everybody who expressed some preference between X and some other candidate Y, preferred Y, and therefore Y should have won. Independence from Pareto-dominated alternatives then simply means that Pareto-dominated candidates can't be spoilers either - they can't even change who wins. I imagine these resistances would mostly come into play in smaller elections. Still, they're nice to have, and their existence immediately tells parties not to try exploiting certain weaknesses (because it won't work). yes, that's the whole point. this is why i am not yet afraid of someone strategically voting to push a Condorcet election into a cycle. it would be an unsafe way to accomplish a political goal. how could anyone predict what would happen? A party might still try, thinking they could pull it off; but complying with a strategic criterion stops that dead, because there's no way it's going to work. It doesn't even have to entertain the thought of trying. With advanced methods, the barrier imposed by these criteria might be so steep that the only remaining strategies are those where a sizable fraction of the electorate has to cooperate, and then it's practically strategy-proof in large public elections (barring disorganized strategy like the a sizable fraction goes on a Burial spree because each individual voter thinks they have nothing to lose of Warren's DH3). Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] I need an example of Condorcet method being subjected to the spoiler effect if any
Jonathan Lundell wrote: In that case it might be a good starting point to define spoiler, so we know what we've found when we find it. What's an example of an IRV spoiler who's not a pretty strong candidate? A very abstract concept of spoiler might be: denote f(X) the minimal number of ballot changes/additions required to make X the winner. Then a spoiler is a candidate with a high f-value relative to the number of ballots (thus hard to get to win), who, by his presence, still changes who wins. Determining f(x) for the various candidates would be very hard, though, and one also runs into the question of what threshold to say above this f-value, spoiler, below it, not a spoiler. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
On Jan 21, 2010, at 4:26 PM, Abd ul-Rahman Lomax wrote: But ... it raises some security issues. And with central counting there are other issues. This is a red herring, because we are talking about precinct summability, and when the number of candidates is very small, precinct summability isn't relevant, because the raw ballot data may be transmitted. no, the problem is that the raw ballot data may be the only practical information to transmitt if the number of candidates is *large*, not very small. when the number of candidates is very small, then it makes sense to transmit the tallies for piles because the number of piles, which are precinct summable, is manageable. So, back to the real question: is precinct summability an important practical criterion to be applied to voting systems? i would ask instead if precinct summability is important for security? i believe that it is. and i believe that it is perfectly practical when the number of *credible* candidates is small. doesn't matter what the voting system is. IRV, or whatever. for 3 candidates, that number is 9. Okay, three candidates, A, B, C, the ballot possibilities are, to be complete, much more than 9. I'll assume that write-ins are illegal and void the ballot. Some of the possibilities are legally equivalent to others, and in actual IRV ballot imaging, they are collapsed and reported the same, to the displeasure of voting security people who do want to know the error rate, which includes overvoting and exact overvoting patterns. So-called ballot images are not, generally. They are processed data reducing a ballot to legally equivalent votes. The reduced set is this: A B C AB AC BA BC CA CB Note that this assumes a 2-rank ballot. no, it can be a 3-rank ballot where the voter declines to rate their last choice. 3rd choice is left unmarked. It also assumes that majority vote isn't important. bullshit. it (the number of consequential ballot permutations) has nothing to do with it (whether or not majority vote is important). If it's important, as it would be in an IRV election under Robert's Rules, we have some more possibilities. They are all the three-rank permutations. ABC ACB BAC BCA CAB CBA Each of these is equivalent, for the purposes of finding a plurality winner, to a two-candidate combination. it's equivalent for the purposes of IRV or Condorcet or *any* method that relies solely on the relative rank of candidates. those 6 markings are equivalent to the corresponding 6 above. if you or Kathy say it's 15, then you're wrong (and it's your slip that's showing). Well, I won't speak about Kathy, but in terms of practical elections in the U.S., she's right. You did not state enough information to establish your reduced count, ... yes i did state enough information. may i remind you? i said that there is *no* consequential difference in these two marked ballots (in the case of N=3). there is no consequential difference between a ballot marked AB to one marked ABC . there is no election scenario, whether it's IRV, Condorcet, Borda or any other method using ranked ballots that will count those two ballots differently. there is no need to separate the AB and ABC into two piles. for N candidates, the number of piles necessary, P(N) is N-1 P(N) = SUM{ N!/n! } n=1 not N-1 P(N) = SUM{ N!/n! } n=0 which is appears to be the formula you and Kathy continue to insist is correct. Which it is under some conditions and yours is correct under some conditions. I assume. I haven't checked them because it's more work than I can put in now. want me to spell it out. it's a simple application of combinatorial analysis, what is the first chapter of my introductory probability textbook (of a course i took more than 3 decades ago). you're doing it already for the specific case of 3 candidates A, B, and C. if you want to look it up, look for language that says something like: how many unique ways can a group of n items be selected from a pool of N items when the order of selection is relevant? and the answer to that is N!/n! . but there is one more fact that you need to toss in. and that fact is that all candidates unmarked or unranked are tied for last place. if there is only one candidate left unmarked, we know how all N candidates are ranked, including the unmarked candidate. everything else between is deleted without comment A vote of ABC, is that the same as AB? Robert assumes, yes. But what about write-ins? ABC is equivalent to ABCW. that's not 3 candidates. that's four. you just changed the premise. that's an official logical fallacy. a form of straw man. if the write-ins are insignificant (usually the case) we can sweep them all into a single insignificant candidate and we have 4 candidates and 40 piles. but we'll see that
Re: [EM] I need an example of Condorcet method being subjected to the spoiler effect if any
On Jan 21, 2010, at 5:33 PM, Jonathan Lundell wrote: On Jan 21, 2010, at 1:03 AM, Juho wrote: What is good in all the common Condorcet methods is that their vulnerabilities to strategies (and their differences in general) may very well be so small in typical real elections (large, public, with independent voter decision making, with changing opinions and less than perfect poll information) that strategic voting will not be a problem. Also their differences with sincere votes (e.g. spoiler related problems) are quite small in real life elections. Here's one simple spoiler related example as a response to Kathy Dopp's request. 35: ABC 33: BCA 32: CAB I this example there are three candidates in a top level cycle. If any of the candidates would not run that would mean that there is a Condorcet winner, and that winner would be different in each case. Let's say that the method we use will pick A as the winner. If B would not run then the votes would be 35: AC, 33: CA, 32: CA and C would win. B is thus a spoiler from C's point of view. I note however that these spoiler cases in Condorcet are not as common as in many of the other methods. In practice it may be that there is no need to worry about these cases. Maybe Kathy Dopp's comparisons will reveal something about how problematic the spoiler effect is or is not in Condorcet. Not also that in the example above B was not a minor party candidate (often term spoiler refers to minor candidates) but a pretty strong candidate. In that case it might be a good starting point to define spoiler, so we know what we've found when we find it. What's an example of an IRV spoiler who's not a pretty strong candidate? In Plurality a typical spoiler scenario is one where the spoiler is a minor candidate (e.g. Nader in the US presidential elections). In IRV the spoilers are typically stronger. Here's one IRV example where the centrist candidate (C) wins. 30: RCL 35: CLR 35: LCR Then we add one more candidate (C2, spoiler) that the R and L supporters strongly dislike. 30: RCLC2 15: CC2LR 20: C2CLR 35: LCRC2 As a result C will be eliminated first, R next, and since C2 is not a strong candidate L will win. C2 thus was a spoiler from C's point of view. C2 is not fully a minor candidate. Although C2 has no chances to win C2 has more first preference votes than C. In IRV this kind of chains of influence could be also longer (5 candidates, 6 candidates etc.), and as a result the spoilers could be more and more minor. But on the other hand the probability of such minor candidates spoiling the election is very low. So, in theory also very minor spoilers are possible but they don't seem probable in practice. This is related to the observation that while Plurality may be in trouble already when there are only two major candidates, main problems of IRV (and Approval and Range) seem to appear only when there are at least three credible candidates. Juho Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] I need an example of Condorcet method being subjected to the spoiler effect if any
On Jan 22, 2010, at 12:05 AM, Kristofer Munsterhjelm wrote: Jonathan Lundell wrote: In that case it might be a good starting point to define spoiler, so we know what we've found when we find it. What's an example of an IRV spoiler who's not a pretty strong candidate? A very abstract concept of spoiler might be: denote f(X) the minimal number of ballot changes/additions required to make X the winner. Then a spoiler is a candidate with a high f-value relative to the number of ballots (thus hard to get to win), who, by his presence, still changes who wins. Determining f(x) for the various candidates would be very hard, though, and one also runs into the question of what threshold to say above this f-value, spoiler, below it, not a spoiler. Maybe one should add also the requirement that the spoiler makes the result worse from spoiler's or spoiler's supporters' point of view. Another possible modification is not to require f(X) to be high. One would just see what would have happened with and without the spoiler. According to that definition also strong candidates (but not actual winners) could be spoilers. (Typically term spoiler refers to minor candidates since these discussions typically refer to a two-party set- up, but the corresponding scientific term might or might not be limited to minor candidates and/or this particular set-up.) Juho Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
robert bristow-johnson Sent: Thursday, January 21, 2010 6:49 AM but breaking it down to piles regarding every conceivable permutation of candidate preference is *still* breaking it down to a finite number of piles. for 3 candidates, that number is 9. if you or Kathy say it's 15, then you're wrong (and it's your slip that's showing). for 4 candidates the number of necessary piles is 40. for N candidates, the number of piles necessary, P(N) is N-1 P(N) = SUM{ N!/n! } n=1 not N-1 P(N) = SUM{ N!/n! } n=0 I do not intend to comment on your formula, but I calculate the numbers of possible unique preference profiles for increasing numbers of candidates (N) as follows: N Unique Preference Profiles 2 4 3 15 4 64 5 325 6 1,956 7 13,699 8 109,600 9 986,409 10 9,864,100 11 108,505,111 12 1,302,061,344 13 16,926,797,485 14 236,975,164,804 15 3,554,627,472,075 16 56,874,039,553,216 17 966,858,672,404,689 18 17,403,456,103,284,400 19 330,665,665,962,404,000 20 6,613,313,319,248,080,000 Where there are large numbers of candidates, the maximum possible number of unique preference profiles will be limited by the number of voters. Thus if there are 10,000 valid votes and 12 candidates, the maximum possible number of preference profiles would be 10,000 and not 1,302,061,344. In practice the actual number of preference profiles would be even lower, as significant numbers of voters would record identical patterns of preferences. Thus in the Meath constituency for the Dáil Éireann election in 2002 with 14 candidates (236,975,164,804 possibilities), there were 64,081 valid votes, but only 25,101 unique preference profiles. The Minneapolis STV (RCV) ballots were all hand sorted to unique preference profiles for each precinct and hand counted. This was unnecessary but feasible as the voters could not record more than three preferences (rankings), no matter the numbers of candidates. I understand the full preference profiles, probably at precinct level, will be published on the City website, but they are not there yet. James Gilmour No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 9.0.730 / Virus Database: 271.1.1/2636 - Release Date: 01/21/10 07:34:00 Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
On Jan 21, 2010, at 6:30 PM, James Gilmour wrote: robert bristow-johnson Sent: Thursday, January 21, 2010 6:49 AM but breaking it down to piles regarding every conceivable permutation of candidate preference is *still* breaking it down to a finite number of piles. for 3 candidates, that number is 9. if you or Kathy say it's 15, then you're wrong (and it's your slip that's showing). for 4 candidates the number of necessary piles is 40. for N candidates, the number of piles necessary, P(N) is N-1 P(N) = SUM{ N!/n! } n=1 not N-1 P(N) = SUM{ N!/n! } n=0 I do not intend to comment on your formula, but I calculate the numbers of possible unique preference profiles for increasing numbers of candidates (N) as follows: N Unique Preference Profiles 2 4 3 15 ... then your calculation is mistaken. the fact that you ostensibly need 4 piles when there are only two candidates should serve as a clue. -- r b-j r...@audioimagination.com Imagination is more important than knowledge. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
N Unique Preference Profiles 2 4 3 15 ... then your calculation is mistaken. the fact that you ostensibly need 4 piles when there are only two candidates should serve as a clue. If there are two candidates, A and B, then the possible unique preference profiles are: A B A B B A Anything that does not conform to this is an incorrect use of the term preference profile. James Gilmour No virus found in this outgoing message. Checked by AVG - www.avg.com Version: 9.0.730 / Virus Database: 271.1.1/2636 - Release Date: 01/21/10 07:34:00 Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
On Jan 21, 2010, at 7:05 PM, James Gilmour wrote: N Unique Preference Profiles 2 4 3 15 ... then your calculation is mistaken. the fact that you ostensibly need 4 piles when there are only two candidates should serve as a clue. If there are two candidates, A and B, then the possible unique preference profiles are: A B AB BA what, on a ballot, is the consequential difference in meaning between A and AB? what effect does a ballot marked AB have over one marked just A (or vise versa) in *any* election method that uses ranked ballots? -- r b-j r...@audioimagination.com Imagination is more important than knowledge. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] I need an example of Condorcet method being subjected
At 04:41 PM 1/21/2010, Jonathan Lundell wrote: On Jan 21, 2010, at 1:32 PM, Kathy Dopp wrote: Terry, You cleverly conveniently change all the definitions whenver it is necessary to make yourself and Fairytale Vote right on the facts. Define spoiler, please, unambiguously. The term has usage, and must be understood from that. A formal definition, nailing it down, would be arbitrary. But Kathy's definition is one reasonable one that matches common usage. But there is another which is broader and, to distinguish this from the common usage, I call the common usage the first order spoiler effect. It refers to minor candidates, with hopelessly low support, who alter the outcome between two major candidates by drawing away votes preferentially from one, from voters who would otherwise vote for that one. The application most common is with plurality, but also top-two runoff and, similarly, IRV, where as little as one vote and some back luck in the resolution of a tie can cause the effect. To define this l.e. spoiler effect more crisply would be arbitrary. But then there is a more generalized spoiler effect, more commonly referred to as center squeeze. It's a spoiler effect, all right, in substance, because an extremist candidates, who would lose in a direct contest between either the centrist or the other extremist, draws enough higher preference votes away from the centrist to reduce that centrist below second rank in first preference. So this is an IIA problem. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
ballot to legally equivalent votes. The reduced set is this: A B C AB AC BA BC CA CB Note that this assumes a 2-rank ballot. no, it can be a 3-rank ballot where the voter declines to rate their last choice. 3rd choice is left unmarked. It also assumes that majority vote isn't important. bullshit. it (the number of consequential ballot permutations) has nothing to do with it (whether or not majority vote is important). If it's important, as it would be in an IRV election under Robert's Rules, we have some more possibilities. They are all the three-rank permutations. ABC ACB BAC BCA CAB CBA Well, I won't speak about Kathy, but in terms of practical elections in the U.S., she's right. You did not state enough information to establish your reduced count, ... yes i did state enough information. may i remind you? i said that there is *no* consequential difference in these two marked ballots (in the case of N=3). there is no consequential difference between a ballot marked AB to one marked ABC . there is no election scenario, whether it's IRV, Condorcet, Borda or any other method using ranked ballots that will count those two ballots differently. there is no need to separate the AB and ABC into two piles. OK. I understand now why you are confused Robert: 1. on the formula for the number of possible unique candidate orderings for any rank choice voting method you incorrectly assume that the number of possible ballot rankings that a voter may fill out is always equal to the number of candidates running for office and so you can collapse N of the rankings, but this simply is not the case in US IRV elections and it would just be unnecessarily confusing to collapse rankings for the special (and unusual) case when there are three candidates and three rankings, when a more general formula that always applies to all situations regardless of the number of candidates and allowed rankings could be used; and 2. on the fact that IRV and Condorcet must be reported similarly and counted similarly, because there are different methods available to count each one. With Condorcet, you can easily count it with an NxN matrix and you cannot count IRV that way at all generally (although I wouldn't put it past you to find an unusual special case where you could). With IRV, you can count it (albeit not easily depending on the number of candidates) with sorting into piles, but you cannot count Condorcet method that way. You can count either Condorcet or IRV by sorting into unique vote orderings, as I gave you the general formula for that works in all cases earlier. However that would be a very difficult and time-consuming way to count Condorcet since Condorcet is precinct-summable in the far simpler n x n matrix. It is the only way to make IRV precinct summable using the formulas I gave you earlier or you can look them up in my IRV report, unless you want to publicly report all voters' individual choices. Minneapolis chose to use the first method. I.e. The counting methods available and ideally used for Condorcet and IRV are different. -- Kathy Dopp Town of Colonie, NY 12304 phone 518-952-4030 cell 518-505-0220 http://utahcountvotes.org http://electionmathematics.org http://kathydopp.com/serendipity/ Realities Mar Instant Runoff Voting http://electionmathematics.org/ucvAnalysis/US/RCV-IRV/InstantRunoffVotingFlaws.pdf Voters Have Reason to Worry http://utahcountvotes.org/UT/UtahCountVotes-ThadHall-Response.pdf Checking election outcome accuracy --- Post-election audit sampling http://electionmathematics.org/em-audits/US/PEAuditSamplingMethods.pdf Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
James, Your formulas below are only correct in the case that voters are allowed to rank all the candidates who run for an election contest. That may be true in Australia, but is not true in the US where typically voters are allowed to rank up to only three candidates. I put the general formula that applies to *all* cases with n candidates and with r rankings allowed in my paper on IRV that I wrote a year or two ago: Realities Mar Instant Runoff Voting http://electionmathematics.org/ucvAnalysis/US/RCV-IRV/InstantRunoffVotingFlaws.pdf Because it's hard to write a summation, fraction formula, etc. here I'll let you look it up. It's on page 6 of the doc linked above. Cheers, Kathy From: James Gilmour jgilm...@globalnet.co.uk I do not intend to comment on your formula, but I calculate the numbers of possible unique preference profiles for increasing numbers of candidates (N) as follows: N Unique Preference Profiles 2 4 3 15 4 64 5 325 6 1,956 7 13,699 8 109,600 9 986,409 10 9,864,100 11 108,505,111 12 1,302,061,344 13 16,926,797,485 14 236,975,164,804 15 3,554,627,472,075 16 56,874,039,553,216 17 966,858,672,404,689 18 17,403,456,103,284,400 19 330,665,665,962,404,000 20 6,613,313,319,248,080,000 Where there are large numbers of candidates, the maximum possible number of unique preference profiles will be limited by the number of voters. Thus if there are 10,000 valid votes and 12 candidates, the maximum possible number of preference profiles would be 10,000 and not 1,302,061,344. In practice the actual number of preference profiles would be even lower, as significant numbers of voters would record identical patterns of preferences. Thus in the Meath constituency for the D?il ?ireann election in 2002 with 14 candidates (236,975,164,804 possibilities), there were 64,081 valid votes, but only 25,101 unique preference profiles. The Minneapolis STV (RCV) ballots were all hand sorted to unique preference profiles for each precinct and hand counted. This was unnecessary but feasible as the voters could not record more than three preferences (rankings), no matter the numbers of candidates. I understand the full preference profiles, probably at precinct level, will be published on the City website, but they are not there yet. James Gilmour -- Kathy Dopp Town of Colonie, NY 12304 phone 518-952-4030 cell 518-505-0220 http://utahcountvotes.org http://electionmathematics.org http://kathydopp.com/serendipity/ Realities Mar Instant Runoff Voting http://electionmathematics.org/ucvAnalysis/US/RCV-IRV/InstantRunoffVotingFlaws.pdf Voters Have Reason to Worry http://utahcountvotes.org/UT/UtahCountVotes-ThadHall-Response.pdf Checking election outcome accuracy --- Post-election audit sampling http://electionmathematics.org/em-audits/US/PEAuditSamplingMethods.pdf Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] I need an example of Condorcet method being subjected
On Thu, Jan 21, 2010 at 7:34 PM, Abd ul-Rahman Lomax a...@lomaxdesign.com wrote: At 04:41 PM 1/21/2010, Jonathan Lundell wrote: On Jan 21, 2010, at 1:32 PM, Kathy Dopp wrote: Terry, You cleverly conveniently change all the definitions whenver it is necessary to make yourself and Fairytale Vote right on the facts. Define spoiler, please, unambiguously. The term has usage, and must be understood from that. A formal definition, nailing it down, would be arbitrary. But Kathy's definition is one reasonable one that matches common usage. But there is another which is broader and, to distinguish this from the Abd ul, actually your definition below is *narrower* not broader, because it narrows the number of cases that fit the definition. I simply took my broader definition that includes all cases of nonwinning candidates who alter election outcomes by their presence, from Arrow's fairness criteria which is describes less simply here: Arrow's Fairness Criteria http://www.ctl.ua.edu/math103/Voting/whatdowe.htm#The%20Independence%20of%20Irrelevant%20Alternatives%20Criterion So you see that my definition is actually Arrow's definition of spoiler, not my own original definition. Kathy common usage, I call the common usage the first order spoiler effect. It refers to minor candidates, with hopelessly low support, who alter the outcome between two major candidates by drawing away votes preferentially from one, from voters who would otherwise vote for that one. The application most common is with plurality, but also top-two runoff and, similarly, IRV, where as little as one vote and some back luck in the resolution of a tie can cause the effect. To define this l.e. spoiler effect more crisply would be arbitrary. But then there is a more generalized spoiler effect, more commonly referred to as center squeeze. It's a spoiler effect, all right, in substance, because an extremist candidates, who would lose in a direct contest between either the centrist or the other extremist, draws enough higher preference votes away from the centrist to reduce that centrist below second rank in first preference. So this is an IIA problem. -- Kathy Dopp Town of Colonie, NY 12304 phone 518-952-4030 cell 518-505-0220 http://utahcountvotes.org http://electionmathematics.org http://kathydopp.com/serendipity/ Realities Mar Instant Runoff Voting http://electionmathematics.org/ucvAnalysis/US/RCV-IRV/InstantRunoffVotingFlaws.pdf Voters Have Reason to Worry http://utahcountvotes.org/UT/UtahCountVotes-ThadHall-Response.pdf Checking election outcome accuracy --- Post-election audit sampling http://electionmathematics.org/em-audits/US/PEAuditSamplingMethods.pdf Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] I need an example of Condorcet method being subjected
Kathy, I ask that you stop smearing me on this (and other) discussion lists. I did not alter any standard definition of spoilers. Webster's online for example defines it as: 1. A candidate with no chance of winning but who may draw enough votes to prevent one of the leading candidates from winning. This means a spoiler is a non-leading candidate with almost no chance of winning (I think the term minor is a fair way of stating that concisely) and not one of the leading candidates. Note also that the concept of having a chance to win suggests the term can be applied prospectively, prior to knowing what the ballots reveal. Kurt Wright, being perceived as a likely winners and who was in first place in the initial tally had an EXCELLENT chance of winning, and almost did in the runoff, and thus does not meet the standard definition of a spoiler. I will refrain from the majority discussion, as that is off topic. Terry Bouricius - Original Message - From: Kathy Dopp kathy.d...@gmail.com To: election-methods@lists.electorama.com Sent: Thursday, January 21, 2010 4:32 PM Subject: Re: [EM] I need an example of Condorcet method being subjected From: Terry Bouricius ter...@burlingtontelecom.net To: Jonathan Lundell jlund...@pobox.com, Juho juho4...@yahoo.co.uk Jonathan makes an important point. The term spoiler means a minor candidate with a small percentage of the vote, who changes which of the other candidates wins by running. But Kathy and some others wish to expand the definition to include a front-runner. (Note that these IRV opponents refer to the top plurality vote-getter who narrowly lost the runoff in Burlington, Kurt Wright, as a spoiler who prevented the candidate in third place from winning. This is a dynamic worthy of analysis, but the word spoiler is never used by the media or political scientists when describing the plurality leader. Terry Bouricius Thanks for all the information re. Condorcet cycles and the unlikely cases of spoilers in Condorcet method of counting rank choice votes from Juho and Robert. Terry, You cleverly conveniently change all the definitions whenver it is necessary to make yourself and Fairytale Vote right on the facts. Let's see what some of them are: A spoiler is *not* according to you, a nonwinning candidate whose presence in the election changes who would otherwise be the winner, but only a particular type of spoiler that is a minor candidate. A majority of voters is *not* according to you, a majority out of all voters who cast ballots, but only out of voters whose ballots have not been exhausted by the time the final IRV counting round is done. A majority candidate, according to you, is *not* the Condorcet winner who a majority (and indeed the most#) of voters favor above all other candidates, but only a candidate who wins a majority in round one, or in the final IRV counting round out of unexhausted ballots after the Condorcet winner and other more majority-favorite winners are eliminated. Terry, redefine any word you want to and you make yourself right, even if most people do not agree with your definitions. It's a good strategy to mislead the public like Fairytale Vote has done. Kathy -- Kathy Dopp Town of Colonie, NY 12304 phone 518-952-4030 cell 518-505-0220 http://utahcountvotes.org http://electionmathematics.org http://kathydopp.com/serendipity/ Realities Mar Instant Runoff Voting http://electionmathematics.org/ucvAnalysis/US/RCV-IRV/InstantRunoffVotingFlaws.pdf Voters Have Reason to Worry http://utahcountvotes.org/UT/UtahCountVotes-ThadHall-Response.pdf Checking election outcome accuracy --- Post-election audit sampling http://electionmathematics.org/em-audits/US/PEAuditSamplingMethods.pdf Election-Methods mailing list - see http://electorama.com/em for list info Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
On Jan 21, 2010, at 7:42 PM, Kathy Dopp wrote: James, Your formulas below are only correct in the case that voters are allowed to rank all the candidates who run for an election contest. James didn't put forth any formulae. but he did put forth a table which appears to be consistent with N-1 P(N) = SUM{ N!/n! } n=0 he, appears to miss the same point as Abd and you do. That may be true in Australia, but is not true in the US where typically voters are allowed to rank up to only three candidates. where do you get your information, Kathy? that is *not* at all the case in the IRV election in Burlington VT. or is Burlington untypical? -- r b-j r...@audioimagination.com Imagination is more important than knowledge. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] I need an example of Condorcet method being subjected
On Jan 21, 2010, at 8:58 PM, Terry Bouricius wrote: Kathy, I ask that you stop smearing me on this (and other) discussion lists. rots o' ruk. I did not alter any standard definition of spoilers. Webster's online for example defines it as: 1. A candidate with no chance of winning but who may draw enough votes to prevent one of the leading candidates from winning. This means a spoiler is a non-leading candidate with almost no chance of winning (I think the term minor is a fair way of stating that concisely) and not one of the leading candidates. Note also that the concept of having a chance to win suggests the term can be applied prospectively, prior to knowing what the ballots reveal. Kurt Wright, being perceived as a likely winners and who was in first place in the initial tally had an EXCELLENT chance of winning, and almost did in the runoff, and thus does not meet the standard definition of a spoiler. Terry, you may have read that i take some responsibility for also associating Wright as the spoiler by replacing almost no chance of winning to having lost in the definition. and i know that they are not the same thing. strictly speaking, Kurt Wright was not a spoiler because it is uncontroversial whether or not he had a chance of winning. that said, i believe that a spoiler-lite (a candidate who loses and whose presence in an election changes who the winner is) problem is still a problem. i think, in these parts, we call it Independence of irrelevant alternatives. IIA is spoiler-lite even if it is not always the spoiler scenario. -- r b-j r...@audioimagination.com Imagination is more important than knowledge. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] IRV vs Plurality (back to the pile count controversy)
At 05:17 PM 1/21/2010, robert bristow-johnson wrote: On Jan 21, 2010, at 4:26 PM, Abd ul-Rahman Lomax wrote: But ... it raises some security issues. And with central counting there are other issues. This is a red herring, because we are talking about precinct summability, and when the number of candidates is very small, precinct summability isn't relevant, because the raw ballot data may be transmitted. no, the problem is that the raw ballot data may be the only practical information to transmitt if the number of candidates is *large*, not very small. when the number of candidates is very small, then it makes sense to transmit the tallies for piles because the number of piles, which are precinct summable, is manageable. This is correct. I actually stated it oppositely, slip of the pen, so to speak. It's still a red herring, because the topic is precinct summability and the general use of precinct sums. The only precinct sum that can be used with IRV is the relevant-ballot-pattern summary, which becomes extremely large very rapidly. Forget about it with manual ballots and more than a quite small number of candidates. Remember, as well, that preferential voting, like top-two runoff, encourages lots of candidates to run, since they can do so with relative safety and get a payoff: some first rank votes that show support. They can turn that into cash in the next election when they are seeking the office again, or in other ways. So, back to the real question: is precinct summability an important practical criterion to be applied to voting systems? i would ask instead if precinct summability is important for security? i believe that it is. Good. So do I. Or was that a slip? and i believe that it is perfectly practical when the number of *credible* candidates is small. doesn't matter what the voting system is. IRV, or whatever. Yes. But how small? Don't use the bogus numbers that aren't at all realistic given real-world election rules, and since we are talking about the U.S., there must be accomodation for every ballot candidate that gets any votes at all in the precinct, plus a write-in provision at a minimum, and God help the election officials if there are a *lot* of write-in candidates, with the sum being more than enough to alter the elimination sequence for the remaining candidate. Write-ins, in all the actual election reported counts, are counted as a category and then dropped when the total for all of them was insufficient to do other then batch-eliminate them, possibly with other candidates as I've seen. I.e., one assumes the simplest case, that all the write-in votes are for the same candidate. Now, if it turns out that the write-ins are relevant, suppose that we set up some rule to lump all candidates with only one vote and report all the others explicitly. But the problem rapidly gets hairy. One has to report another candidate as relevant in addition to all the write-ins. For voting system security issues, one must be able to count the votes manually, as part of an audit. I'm sure that Kathy could explain audit process, but, again, it gets very hairy rapidly with IRV, because vote samples aren't enough, given the sensitivity of the method to many small differences in vote patterns. What is actually being done? Only ballot images, with machines that collect and report them, in toto, from the precinct. In other words, the only solution in actual usage that doesn't involve toting all the ballots to a central location involves reporting ballot images. But this is precisely a system that is quite vulnerable to hacking and some very real voting security issues. If there are no paper ballots or at least bulletproof paper records that the voter personally verified, it's impossible to verify that there were no shenanigans. Precinct summable methods are not nearly as sensitive to manipulation as are IRV totals, it appears. It can only take a very small shift in voting patterns to shift an IRV result, under some conditions, and this isn't merely a very close election in terms of overall support for a candidate, it gets down to exact preference order and how it interacts with elimination sequence, which is determined sometimes at many places in the election process. And, note: if it's ballot images, these images don't include, generally, the actual write-in votes. If it turns out that write-ins need to be counted, only manual counting can do it, the name was hand written on a record, if I'm correct. for 3 candidates, that number is 9. Okay, three candidates, A, B, C, the ballot possibilities are, to be complete, much more than 9. I'll assume that write-ins are illegal and void the ballot. Some of the possibilities are legally equivalent to others, and in actual IRV ballot imaging, they are collapsed and reported the same, to the displeasure of voting security people who do want to know the error rate, which includes overvoting and
Re: [EM] I need an example of Condorcet method being subjected
Start telling the truth about IRV at Fairytale Vote, and then when people speak the truth about Fairytale Vote, it won't sound like a smear. If you don't like the sound of your own behavior when retold, try some different behavior, rather than falsely accusing others of smearing you when they tell the truth. The list of deliberate misinformation told by the Fairytale Vote group, if put together in one place, would reach to the moon most likely, as this one youtube video alludes to. Experts in election methods have had occasions to debunk the disinformation by Fairytale Vote sufficiently to know that. http://www.youtube.com/watch?v=bPCS-zWuel8 So continue to alter the definitions of commonly used words in order to make yourself right because that's the only way you can do that Terry. Your definition of spoiler alters the definition that was given by Arrow's theorem which is the one I choose to use, not the narrower one your organization uses in order to make the false claim that IRV solves the spoiler problem even though it obviously does not, and the Republican candidate acted as a spoiler to knock out the most popular Democratic candidate so that their least favorite, the left wing Progressive could win. Terry, just do not imagine that people do not see the trick you use of redefining words that have had a common meaning for decades. Cheers, Kathy On Thu, Jan 21, 2010 at 8:58 PM, Terry Bouricius ter...@burlingtontelecom.net wrote: Kathy, I ask that you stop smearing me on this (and other) discussion lists. I did not alter any standard definition of spoilers. Webster's online for example defines it as: 1. A candidate with no chance of winning but who may draw enough votes to prevent one of the leading candidates from winning. This means a spoiler is a non-leading candidate with almost no chance of winning (I think the term minor is a fair way of stating that concisely) and not one of the leading candidates. Note also that the concept of having a chance to win suggests the term can be applied prospectively, prior to knowing what the ballots reveal. Kurt Wright, being perceived as a likely winners and who was in first place in the initial tally had an EXCELLENT chance of winning, and almost did in the runoff, and thus does not meet the standard definition of a spoiler. I will refrain from the majority discussion, as that is off topic. Terry Bouricius - Original Message - From: Kathy Dopp kathy.d...@gmail.com To: election-methods@lists.electorama.com Sent: Thursday, January 21, 2010 4:32 PM Subject: Re: [EM] I need an example of Condorcet method being subjected From: Terry Bouricius ter...@burlingtontelecom.net To: Jonathan Lundell jlund...@pobox.com, Juho juho4...@yahoo.co.uk Jonathan makes an important point. The term spoiler means a minor candidate with a small percentage of the vote, who changes which of the other candidates wins by running. But Kathy and some others wish to expand the definition to include a front-runner. (Note that these IRV opponents refer to the top plurality vote-getter who narrowly lost the runoff in Burlington, Kurt Wright, as a spoiler who prevented the candidate in third place from winning. This is a dynamic worthy of analysis, but the word spoiler is never used by the media or political scientists when describing the plurality leader. Terry Bouricius Thanks for all the information re. Condorcet cycles and the unlikely cases of spoilers in Condorcet method of counting rank choice votes from Juho and Robert. Terry, You cleverly conveniently change all the definitions whenver it is necessary to make yourself and Fairytale Vote right on the facts. Let's see what some of them are: A spoiler is *not* according to you, a nonwinning candidate whose presence in the election changes who would otherwise be the winner, but only a particular type of spoiler that is a minor candidate. A majority of voters is *not* according to you, a majority out of all voters who cast ballots, but only out of voters whose ballots have not been exhausted by the time the final IRV counting round is done. A majority candidate, according to you, is *not* the Condorcet winner who a majority (and indeed the most#) of voters favor above all other candidates, but only a candidate who wins a majority in round one, or in the final IRV counting round out of unexhausted ballots after the Condorcet winner and other more majority-favorite winners are eliminated. Terry, redefine any word you want to and you make yourself right, even if most people do not agree with your definitions. It's a good strategy to mislead the public like Fairytale Vote has done. Kathy -- Kathy Dopp Town of Colonie, NY 12304 phone 518-952-4030 cell 518-505-0220 http://utahcountvotes.org http://electionmathematics.org http://kathydopp.com/serendipity/ Realities Mar Instant Runoff Voting
Re: [EM] I need an example of Condorcet method being subjected
On Jan 21, 2010, at 9:03 PM, Kathy Dopp wrote: Start telling the truth about IRV at Fairytale Vote, and then when people speak the truth about Fairytale Vote, it won't sound like a smear. Ha ha! I've been meaning to compliment you, Ms Dopp, on that sidesplitting line. It was really funny the first time, and it's gotten more hilarious each time you've used it since. How many times? I've lost count, to tell the truth, but please keep it up. Fairytale! Fantastic! I can't get enough. If you don't like the sound of your own behavior when retold, try some different behavior, rather than falsely accusing others of smearing you when they tell the truth. The list of deliberate misinformation told by the Fairytale Vote group, if put together in one place, would reach to the moon most likely, as this one youtube video alludes to. Experts in election methods have had occasions to debunk the disinformation by Fairytale Vote sufficiently to know that. Three times in one post. Just in case we readers don't get it the first or second time--the mark of a master comic. Thanks! http://www.youtube.com/watch?v=bPCS-zWuel8 So continue to alter the definitions of commonly used words in order to make yourself right because that's the only way you can do that Terry. Your definition of spoiler alters the definition that was given by Arrow's theorem which is the one I choose to use, Arrow never used, never mind defined, the word spoiler. not the narrower one your organization uses in order to make the false claim that IRV solves the spoiler problem even though it obviously does not, and the Republican candidate acted as a spoiler to knock out the most popular Democratic candidate so that their least favorite, the left wing Progressive could win. Terry, just do not imagine that people do not see the trick you use of redefining words that have had a common meaning for decades. Election-Methods mailing list - see http://electorama.com/em for list info