Re: [EM] Mathematical equivalence (was Re: Democracy Chronicles, introductions)
On 4/28/2012 9:48 AM, Kristofer Munsterhjelm wrote: On 04/24/2012 08:37 PM, Richard Fobes wrote: In the non-mathematical world the word "equivalent" means "having similar or identical effects" which allows for not _always_ being _identical_ in _all_ respects. That is the context for usage in the Democracy Chronicles article. A context which is overriden by prefixing the word "equivalent" with "mathematically". I have recommended to Adrian a wording change that does not include the word "mathematical." Consider two functions f and g defined on the integers. f(x, y) = x + y, g(x, y) = x + y when |x-y| > 2 = x * y otherwise. The relationship between these two functions is not similar to the relationship we are discussing. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] Democracy Chronicles, introductions
Adrian, although I do not think that readers will misunderstand the
words "mathematically equivalent," Kristofer has been insistent that he
does not like that wording, so let's change the wording accordingly:
"... VoteFair popularity ranking, which virtually always identifies the
same election winner as the Condorcet-Kemeny method, one of the methods
supported by ..."
Eliminating the word "mathematical" should make Kristofer happy, based
on what he says in a recent message. And the word "virtually" will be
recognized by anyone as a qualification of the word "always." And
referring to the winner without referring to the "results" -- which in
the Condorcet-Kemeny method includes a full ranking of all choices --
will avoid an issue that is not worth explaining in this article.
Clarification for forum purposes: The full ranking from most popular,
second-most popular, and so on down to least popular can differ between
the Condorcet-Kemeny method and the VoteFair ranking software, and that
accounts for the largest number of cases in which there is a difference.
Out of those cases a much smaller number of cases could involve a
difference in who is declared the winner (the highest-ranked choice),
but that can only happen if there are more than six candidates in the
Smith set (although I have not yet had time to reply to Jameson
regarding the proof of this point), which is covered by the word
"election" before the word "winner" because real elections do not have
that many candidates in the Smith set. Non-election situations, such as
the ranking of 100 songs, would have a reasonable (yet still unlikely)
chance of having more than six choices in the Smith set.
Further clarification for forum purposes: The statement in the revised
wording for this Democracy Chronicles article refers to the VoteFair
ranking software, which is not the same as saying that VoteFair
popularity ranking does not intend to duplicate Condorcet-Kemeny results
in all cases. (First I want to characterize the cases in which they
differ as being so convoluted in terms of voter preferences that the
difference is not significant for the purposes of use in an election
[remembering that there must be more than six candidates in the Smith
set in order for the top-ranked-choice difference to occur].)
Adrian, very importantly, I recommend revising the article's words
"diverse group of election experts" and "election reform advocate" to
use the phrases "election-method experts" and "election-method reform
advocate" because our topic ("election-method" reform) is a subset of
"election reform," and we do not claim to be "election experts" -- which
would offend people who study voter registration, voter turnout, and
many other characteristics of real elections.
I approve the article if the phrase "election-method" is used.
The only other edit might be to hyphenate the phrase "round-off" in the
words "... compensate for any round off errors ...". Actually I think
the correct spelling in an academic article might be "roundoff" as a
single word (but I'm not sure), but that would be confusing to
non-academic readers.
Thank you for your great work Adrian! And especially thank you for your
patience in dealing with those of us who choose our words so carefully
as a result of discussing our mathematically rigorous topic.
As for an image, you have my permission to use the graphic in the upper
left of the VoteFair.org home page:
http://www.votefair.org/index.html
(The same image appears on the BanSingleMarkBallots.org website, but
that is not an official choice, just an image that fills the spot until
something better is presented.)
Or, if you prefer to use the cover of my book, you have my permission to
use it at [http://www.solutionscreative.com/ehu_cover.html], but please
keep in mind that my goal for this article is to promote the
VoteFair.org website, and secondarily mentioning (and linking to) my
creative-problem-solving book as credibility for my problem-solving
skills, so this article should not contain a link that (also) promotes
my election-method-reform book unless you use that cover for your image
(in which case for this article I prefer that you link to the Google
Books version at [http://books.google.com/books?id=UOf86S4Lc-YC] where
people can read some of it for free).
Speaking of images, if you are interested in the voting-related cartoons
in "Ending The Hidden Unfairness In U.S. Elections," you are welcome to
use them in a separate cartoon-only series.
Again, thank you Adrian for connecting us with people who can benefit
from our deep understanding of election methods.
Richard Fobes
On 4/28/2012 9:38 AM, Adrian Tawfik wrote:
As long as everyone is somewhat comfortable with keeping the "mathematically
equivalent"
wording, I think we can move forward with the article. I put together the more
complete
text of the article with the interview included and some
Re: [EM] Kristofer: Approval vs Condorcet, 4/28/12
On 4/28/12 11:46 PM, Paul Kislanko wrote: It fails when approval is used as vote counting method. In approval COUNTING, if you voted Favorite above Compromise, you vote Favorite EQUAL Compromise, and even though you don't like Compromise, you helped elect the idiot. i've been saying this for months. in Approval voting, how does a voter decide whether to approve of their 2nd choice. they surely want their 2nd choice to beat their most hated candidate, but they don't want to help their 2nd choice to beat their favorite. Approval sucks. similar problem for Score voting. you just cannot say that these two systems speak adequately to the burden of tactical voting they place upon voters. On Apr 28, 2012, at 5:04 PM, Michael Ossipoff wrote: For one thing, Condorcet discourages honesty, this is just stupid. because, even if you top-rank Compromise, top-ranking Favorite too can cause Compromise to lose to Worse. as long as Compromise is ranked above Worse, it doesn't matter what you do to Favorite, you are not affecting your contribution to Compromise's position with respect to Worse's position (your vote increases Compomise's lead over Worse or decreases Worse's lead over Compromise). when ranking Compromise _alone_ in 1st place would have defeated Worse. To do your best to defeat Worse, you have to vote Favorite below Compromise. baloney. unless you're assuming some kind of pathological cycle is to happen. and i don't accept that cycles are anywhere close to common. You have to say with your vote that Condorcet is better than Favorite. ??? you mean "Compromise is better than Favorite."? if that is what you meant to say, then i say you are mistaken. Consider that before you criticize Approval for not letting you vote Favorite over a needed compromise. What is going on here? I properly have to rank Favorite above Compromise. Exactly how can this fail? i am still unimpressed with Mike O's analysis if this is what it is. maybe i should un-plonk him, but i dunno why. -- 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] Paul: Counts
But who says it has to be marked by a machine. A handcount actually doesn't requirea machine-printed ballot! I would not deceive you about that. The voter could mark hir ballot, _hirself_, without a machine! Pencils used to be used, for instance. Pens would be better, for an un-erasable mark.Punched cards have been used too. Read by a machine, but (hopefully) saved for examination.Of course even if punched cards are used, that doesn't mean that they must be read by machine atall. Punched holes are especially un-modifiable. i used to live in Illinois and we had punch cards. they are cheap and the voting "machines" (they are actually just a jig) are inexpensive for the municipalities. but the problem with punch cards is that they do not have the candidate names on the physical instrument themselves. the punch card could be misregistered or mal-aligned. you could punch through the hole on the jig of the candidate you support and it punches out the chad in the card for the candidate you hate. and you wouldn't know it unless you could, by sight, decode the row and column numbers for your candidate. a similar issue exists with those older mechanical voting machines. the voter pulls the lever down for the candidate they select and when they commit their vote (by pulling on the big red lever), they are not assured that the marks left on the internal paper record will be properly interpreted (because of a misalignment in the machine), *even* if hand-counted. if the machine takes your vote and, because of misalignment, assigns that vote to another candidate, there is little that can be done later in a manual recount if this misalignment cannot be connected to the particular machine. that's more of a problem with punch cards since, at a particular precinct, there might be several, ostensibly identical, vote-punching kiosks that the voter can use and only one is misaligned. you won't be able to tell which intermingled ballots went through that particular jig. optical scan technology applied to paper (or light cardboard) ballots, where the candidate names are marked on the very same ballot (the physical instrument), is the only way to avoid this danger at least in the manual recount. sure, there could be an alignment problem in the scanning machine and that would result in an incorrect tabulation, but if someone bitched about it and the decision was made to recount manually, you could do it and be assured that the original intent of the voter was manifest in each ballot. optical scan is the only smart way to do this and i think that a trustworthy, transparent, and democratic system would best use optical scan consistently to the exclusion of all other technologies. BTW, i had (and still have) Mike plonked. so i can only see responses to whatever he says. i am still, for the most part, unimpressed my Mike's content-to-hype ratio. i just had to respond to this dumb punch card thing. -- 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] Kristofer: Approval vs Condorcet, 4/28/12
It fails when approval is used as vote counting method. In approval COUNTING, if you voted Favorite above Compromise, you vote Favorite EQUAL Compromise, and even though you don't like Compromise, you helped elect the idiot. -Original Message- From: election-methods-boun...@lists.electorama.com [mailto:election-methods-boun...@lists.electorama.com] On Behalf Of Dave Ketchum Sent: Saturday, April 28, 2012 9:15 PM To: Michael Ossipoff Cc: election-meth...@electorama.com Subject: Re: [EM] Kristofer: Approval vs Condorcet, 4/28/12 On Apr 28, 2012, at 5:04 PM, Michael Ossipoff wrote: > For one thing, Condorcet discourages honesty, because, even if you > top-rank Compromise, top-ranking Favorite too can cause Compromise to > lose to Worse. when ranking Compromise _alone_ in 1st place would > have defeated Worse. To do your best to defeat Worse, you have to vote > Favorite below Compromise. You have to say with your vote that > Condorcet is better than Favorite. Consider that before you criticize > Approval for not letting you vote Favorite over a needed compromise. > What is going on here? I properly have to rank Favorite above Compromise. Exactly how can this fail? 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] Paul: Counts
Mike, I asked what ballots are you going to hand-count. I vote by touching a touch-screen, and the machine gives me a receipt. You say I COULD give you a paper ballot to hand-count, but if I just voted by pressing a portion of a touch-sensitive-display-screen, what are you going to hand-count? My touch is not verifiable by a hand-count of what the machine recorded. It can only be verified by asking me if what your machine's record matches what the machine printed out for me. And you can't "hand count" that without asking me if what you're counting matched my ballot. And you can't do THAT without violating the principles of all voting systems. From: election-methods-boun...@lists.electorama.com [mailto:election-methods-boun...@lists.electorama.com] On Behalf Of Michael Ossipoff Sent: Saturday, April 28, 2012 9:10 PM To: election-meth...@electorama.com Subject: [EM] Paul: Counts Paul: You said: That wasn't Dave who said that, it was me. [endquote] That's why I said, "My apologies to Paul Kislanko, whom I called by the wrong name." You asked: My point was that what is it you're going to count by hand? The tape printed out by the machine I voted on? [endquote] It wouldn't be printed on a tape, would it. The card that it's printed out on? Sure, if you've had a chance to look at it first, to verify that it says what you intended it to, before you deposit it in the ballot-box slot. But who says it has to be marked by a machine. A handcount actually doesn't require a machine-printed ballot! I would not deceive you about that. The voter could mark hir ballot, _hirself_, without a machine! Pencils used to be used, for instance. Pens would be better, for an un-erasable mark. Punched cards have been used too. Read by a machine, but (hopefully) saved for examination. Of course even if punched cards are used, that doesn't mean that they must be read by machine at all. Punched holes are especially un-modifiable. Someone pointed out that an Approval ballot could be illegally modified by adding more marks. One solution would be for the voter to mark "Approved" or "Unapproved". In any case, video cameras must record every moment that the ballots are handled. That would prevent modifications. The locked ballot-box, in a locked room, should likewise be video-camera protected. I'm skeptical about secure machine-counts, but it would be ok if it could be guaranteed. Until then, counting should only be by hand. That rules out Condorcet and Kemeny. Approval is easily handcountable. Mike Ossipoff If I didn't hand-fill-out a ballot you've no way to know whether the ballots you're counting by hand match what I told the machine was my vote. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] Kristofer: Approval vs Condorcet, 4/28/12
On Apr 28, 2012, at 5:04 PM, Michael Ossipoff wrote: For one thing, Condorcet discourages honesty, because, even if you top-rank Compromise, top-ranking Favorite too can cause Compromise to lose to Worse. when ranking Compromise _alone_ in 1st place would have defeated Worse. To do your best to defeat Worse, you have to vote Favorite below Compromise. You have to say with your vote that Condorcet is better than Favorite. Consider that before you criticize Approval for not letting you vote Favorite over a needed compromise. What is going on here? I properly have to rank Favorite above Compromise. Exactly how can this fail? Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] Dave Ketchum: Handcounts
Computers do well at performing the tasks they are properly told to perform - better than humans given the same directions. Thus it would make sense to direct the computers and expect them to do what is needed accurately. Still, we sometimes wonder exactly what the computers have been told to do. I keep remembering a case where it took me some studying to sort out how a computer program seemed to perform magic: Got a diskette and was told it could not be copied - because the owner of the program wanted to get paid for each new diskette the program was copied to. Curious, though not intending theft, I copied all the records to another diskette and tried running it. New diskette knew it was not an original, and refused to run. Looked at records on the diskette - there were no instructions in the copied records to do such a test - anyway, presumably the records had been copied accurately, so their content could not indicate this was a copy. Look more carefully - there was an odd collection of odd words. Turns out that, when it was time to check, this collection became instructions to look at the noise next to a particular record on the diskette and see if the "password" was there, as it would be on the original. After making the test, put those instructions back into their hidden form. Dave Ketchum On Apr 28, 2012, at 9:28 PM, Paul Kislanko wrote: That wasn’t Dave who said that, it was me. My point was that what is it you’re going to count by hand? The tape printed out by the machine I voted on? If I didn’t hand-fill-out a ballot you’ve no way to know whether the ballots you’re counting by hand match what I told the machine was my vote. From: election-methods-boun...@lists.electorama.com [mailto:election-methods-boun...@lists.electorama.com ] On Behalf Of Michael Ossipoff Sent: Friday, April 27, 2012 9:55 PM To: election-meth...@electorama.com Subject: [EM] Dave Ketchum: Handcounts Dave: You said: And that gets to why I think "hand counting" is no longer useful as "verification" - what is there to "hand count" when there are no paper ballots except those printed by the machines that we're auditing? [endquote] Handcounting was used even before there were voting machines! I kid you not. Mike Ossipoff Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] Dave Ketchum: Handcounts
That wasn't Dave who said that, it was me. My point was that what is it you're going to count by hand? The tape printed out by the machine I voted on? If I didn't hand-fill-out a ballot you've no way to know whether the ballots you're counting by hand match what I told the machine was my vote. From: election-methods-boun...@lists.electorama.com [mailto:election-methods-boun...@lists.electorama.com] On Behalf Of Michael Ossipoff Sent: Friday, April 27, 2012 9:55 PM To: election-meth...@electorama.com Subject: [EM] Dave Ketchum: Handcounts Dave: You said: And that gets to why I think "hand counting" is no longer useful as "verification" - what is there to "hand count" when there are no paper ballots except those printed by the machines that we're auditing? [endquote] Handcounting was used even before there were voting machines! I kid you not. Mike Ossipoff Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] Dave Ketchum: Repetition of previoius Approval discussion
On Apr 28, 2012, at 12:56 AM, Michael Ossipoff wrote: First, my apologies to Paul Kislanko, whom I called by the wrong name when I replied to his posting, a few minutes ago. _This_ reply is to Dave Ketchum: Dave: I'd said: > How to avoid this problem? Why not repeal the rule that makes > Plurality so funny? Let people rate _every_ candidate with a 1 or > a 0. Rate every candidate as "Approved" or "Unapproved". The > candidate with the most "Approved" ratings wins. The result? Well, > we'd be electing the most approved candidate, wouldn't we. Who can > criticize that? > You replied: Anyone who realizes that there is more to wish for. My next sentence was part of completing that thought: Here you can vote for both Favorite and Compromise to help defeat Worse, but cannot vote for both without implying equal liking for each - and thus risking unwanted election of Compromise. [endquote] Ah, "If wishes were horses..." :-) Far be it for to tell you what you should or shouldn't wish for. But you should keep the distinction between wishes, fantasy, and feasible possibilities. Anyway, as I explained to you when we had this same discussion a few days ago, even you can't complain about changing from Plurality to Approval. (At least I assume that you don't believe that you have an argument agains that change). It being an improvement I did not, and do not, argue against changing from Plurality to Approval - the improvement is minor, but the effort is comparatively minor. Going back to your thoughts when starting this series, Plurality does not allow voters to adequately express their desires. They do not want Worse, so vote for Compromise as the best hope of accomplishing this major goal under Plurality - they feel that voting for Favorite may let Worse win. Approval helps by letting them vote for both Favorite and Compromise. However this is only a partial correction since it says they have equal liking for each. Thus I argue for using a stronger method, such as Condorcet, that will let voters more completely indicate which candidates they most prefer when voting for more than one: . It matters little whether Approval is used until we agree on something better - it is better than Plurality but very little different. . While I promote Condorcet, I do not here argue for or against varieties, even such as ICT that Mike talks of. You can say, "But I want something more complicated that I claim will be better." But that isn't an argument against changing from Plurality to Approval. That change (from Plurality to Approval) amounts to nothing more than repealing the ridiculous rule that is Purality's problem. Now, as I've discussed, a proposal to change from Plurality to Condorcet would be a whole other ballgame.If you want to try that, then feel free to. But don't say I told you to. You continued: Here you can vote for both Favorite and Compromise to help defeat Worse, but cannot vote for both without implying equal liking for each [endquote] In a u/a election (there are unacceptable candidates who could win) your best strategy in Condorcet is to rank all of the acceptable candidates in 1st place, and not rank any unacceptable candidates. Doing so doesn't imply that you equally like everyone whom you equal- rank. "1st place" puzzles. Thinking of Favorite and Compromise, I likely want to vote for both in Condorcet, but to rank Favorite higher to indicate my preference. You continued: - and thus risking unwanted election of Compromise. [endquote] Sorry, but you do need to risk that, in Condorcet, in a u/a election. But don't feel too bad, because "unwanted" has a whole other (and stronger) meaning when applied to the unacceptable candidates. I'd said: If you have given 1 point to > Compromise, and 0 points to Worse, then it’s obvious that also > giving a point to Favorite won’t change the fact that you’ve fully > helped Compromise against Worse. > You say: The above sentence emphasizes what happens to Compromise vs Worse, ignoring that it destroys Favorite's desired advantage over Compromise. [endquote] But, with Condorcet, you can't say what I said: Top-ranking Favorite means that you aren't fully helpng Compromise against Worse. There are situations in which Worse will win instead of Compromise because you top-ranked Favorite alongside Compromise. The thinking is getting confused. You are quoting what I said about Approval, and then incorrectly stating what this might do to Condorcet. In Condorcet if I rank Compromise, but not Worse, that is as strong as I can be as to this pair. If I also rank Favorite higher, that is as strong as I can be as to these. Worse, being unranked, is shown as least liked among these. As to top-ranking both Favorite and Compromise, that indicates liking them equally, but does not affect them vs unranked Worse. Dave Ketchum Therefore,
[EM] Kristofer: Approval vs Condorcet, 4/28/12
Kristofer: On 04/27/2012 07:32 PM, Michael Ossipoff wrote:>* ...but wouldn't it be better not to have the disadvantags?* You replied: Sure. It would be better to make Arrow and G&S go away, but it's not going to happen. [endquote] Neither Arrow, Gibbard, not Sattterthwaite has ever said that a voting system has to be a completely novel and complicated contraption, or that every voting system must give incentive for favorite-burial. Therefore, even _without_ making Arrow, Gibbard and Satterthwaite go away, it's _still_ possible to avoid Condorcet's serious faults and disadvantages. :-) You continued: If you want to have Condorcet, you have to give up some desiderata. If you want FBC, you have to give up quite a bit, even for "weak" FBC which is fairly strong by itself. [endquote] It's a accepted that we have to give up some criteria, properties, and attributes to gain others. Election methods could continue debating the relative merits of various rank-counts, for the next 250 years, and, meanwhile, if that's the best we're doing, we'll still be voting by Plurality in our public elections. There's nothing wrong with discussing our ideals, and debating whose ideal is best. But we must not confuse that worthy passtime with actually engaging with the public and the court system, to actually do something about Plurality's blatantly and glaringly ridiculous, and unnecessary, falsification requirement. Give up something? I've said much about the societal benefits that would be gained merely by minimally getting rid of that one Plurality fault. Unless you claim that that wouldn't bring them, then I invite you to tell why you think that they wouldn't be enough. That's why I say that Condorcetists haven't looked at Approval and the results that it would have. Condorcetists are too busy with the distinctions between rank-counts, to actually look at what can be accomplished without rank balloting. >* As you suggested, if Condorcet were proposed municipally, maybe (as*>*was >the case with IRV in some municipalities) big outside money won't *>* come in to emphasize rank-balloting methods complete novelty,*>*unknown-ness, unfamiliarity and unpredictability. *They didn't emphasize any of the weird properties of IRV. Warren et al. did that. What they *have* emphasized is the unnatural nature of any method that isn't Plurality, and that is a hurdle to Approval, too. [endquote] Indeed a hurdle to all voting system reforms. That's why the _minimal_ change of just getting rid of Plurality's "Pretend that you want to give 0 points to all but one candidate" falsification requirement is the more viable reform proposal. Not a completely new method, with a more elaborate new kind of balloting and an elaborate contraption-count. Just an elimination of one ridiculous and unsupportable rule of Plurality. You're underestimating the societal improvement that that one little minimal change would bring, when everyone can fully support whom and what they really like best. You continued: Yes, you may counter by saying you can explain to people that Approval is "setwise voting" and thus is one man one vote [endquote] I prefer to just point out that Approval is the 0,1 point system, in which each voter has the same equal power to give to any candidate either 0 points or 1 point. (or that in Approval, any vote can be countered by an opposite inverted vote), but that still means you have to show people how to look beyond their immediate intuition. [endquote] ...a problem with every proposal for a new voting system. But how much greater a problem for a complicated and completely novel method that's unrelated to and not derived from what we now use! Tradition counts against reform. That's a given. That's how it's always been. People will say, "Plurality goes all the way back to the first days of the Consititution." But do did slavery. It was eventually rejected. Slavery shows that the Constitution can be wrong. ...and that there is precedent for correcting an error or fault in the Consititution. The electoral college mess is of course specified by the Constitution. I don't know if Plurality itself is Constitution-specified, but it was traditional and assumed to be the way to go. I'd said: >* But maybe the reason why that wasn't always done to IRV proposals*>* was because the big opponents knew that IRV wasn't going to really*>* improve on Pluralilty anyway, given current electorates.*I doubt it. Of course it could be the case that high-profile people who supported IRV, like Howard Dean, were complicit. [endquote] I haven't claimed that. Maybe Dean just knew that it would win points for him among progressives. I'm just saying that maybe no great need was perceived to quash IRV, because it obviously wan't good enough to bother quashing. You continued: Arrow and G&S implies every "boxer" has to hold something behind his back. [endquote] Arrow and G&S most certainly do not imply that every voting system proposal has to have the enactment-disadvantag
Re: [EM] Election layering effect (or why election-method reform is important)
Hi, an other aspect to this is the development of a posteriori to election criteria to measure satisfaction of the electorate from the results. Forget about the electoral method, just focus on the result and the electorate will. Individually taken, it is easy to quantify how much an elector is satisfied with the results from its own participation to an election. Among all the candidates a candidate could have an influence on their election, the ratio of these candidates getting elected represents the satisfaction rate of the elector. How do you get the full will of that elector? Just assume all voters proposed with the same choices than that elector make the same picks. The result represents the 100% satisfaction result. Some examples: With an FPTP election, 3 districts: District 1: Blue* 45% Red 35% Yellow 20% District 2: Blue* 51% Red 30% Yellow 19% District 3: Blue 25% Red 30% Yellow* 45% The 3 elected MP (2 Blue, 1 Yellow) produce an average individual satisfaction rate of 47%. With an STV election, 3 seats in a single super-district, let's assume following ballots and results: 10%: B3 B2 Y2 R1 R2 Y1 Y3 R3 B1 30%: R1 R2 R3 B1 B2 B3 Y1 Y2 Y3 51%: B1 Y1 Y2 Y3 R1 R3 R2 B3 B1 9%: Y1 B1 R1 B2 R2 Y3 Y3 R3 B3 Elected: B1 Y1 R1. Individual satisfaction of the first group of voters: 0% (none elected among the 3 first choices) Individual satisfaction of the second group of voters: 33,3% (one elected among the 3 first choices) Individual satisfaction of the third group of voters: 66,7% (two elected among the 3 first choices) Individual satisfaction of the fourth group of voters: 100% (all elected among the 3 first choices) Global individual satisfaction of all voters: 10% x 0% + 30% x 33,3% + 51% x 66,7% + 9% x 100% = 53% Typically STV produces a global individual satisfaction rates around twice FPTP rates for the simulations I have made yet... Almost all election method can be measured this criteria (it makes no sense for a fully random selection). This does not covers the layering effect of multiple representative levels, but it emphasizes the mismatch between the will of electors and the results. Stéphane Rouillon On 2012-04-27 15:26, Richard Fobes wrote: Recently I realized that in our Declaration, and in our discussions, we have failed to explain and explore the "amplification" effect that occurs as a result of, for a lack of a better term at the moment, "layering." Here is how I explained it in the proposal I referred to earlier: "Winning an election with less than half the votes might seem like a small unfairness, but the effect is huge because of a layering effect. Although each Congressman typically got a ballot mark from about one out of two voters in the general election, he or she got a ballot mark from only about one out of four voters (based on cross-party counting) if the Congressman competed against a strong candidate in the primary election. Another layer occurs because only slightly more than half the members of Congress need to vote in favor of a new law to get it passed, so just those Congressmen got ballot marks from only about one out of eight U.S. voters, which is about 12% of U.S. voters. Yet even more layers are involved because most Congressmen first serve as state-level officials, and the state-level election process similarly filters out the problem-solving leaders that most voters want. Adding in two more layers to account for mainstream-media influence and low voter turnout easily accounts for how each law passed in Congress represents the desires of only 1% of the U.S. population." (The full proposal is at: http://www.the99declaration.org/4408/ban_single_mark_ballots_from_congressional_elections?recruiter_id=4408 ) I'm interested in any ideas for how this concept can be explained more clearly, especially if someone can think of an appropriate analogy or metaphor or diagram. Richard Fobes 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] Kristofer: Yes, maybe Condorcet could overcome its disadvantages.
On 4/28/12 1:13 PM, Kristofer Munsterhjelm wrote: But maybe the reason why that wasn't always done to IRV proposals was because the big opponents knew that IRV wasn't going to really improve on Pluralilty anyway, given current electorates. I doubt it. Of course it could be the case that high-profile people who supported IRV, like Howard Dean, were complicit. But if so, they're doing a *really* good job at faking their enthusiasm. i supported Howard Dean, both for governor and also for president. in fact, because lacking Single Payer, the State of Vermont has the closest thing going compared to any U.S. state (it's called VHAP, Vermont Health Access Program) and it was something signed into law by Dean and i was a beneficiary of it (it saved me from a heart attack). in fact, i got to tell this story and introduce this presidential candidate to a town hall meeting two days after the infamous "Dean scream" and just before the New Hampshire primary (where Kerry beat Dean) in Wentworth NH. got to talk with George Stephanopolis. so i like Howard Dean. but Howard's endorsement of IRV during our IRV repeal slugfest did little good. *during* his endorsement, with the news cameras rolling, he said (sorta asked) that this (2009) was the first time we used it. in fact, it was the second time, the first time was in 2006. but it made him look like he was outa touch with the town he lives in (at the time he was the Chair of the Democratic National Committee and probably spent more time away from home than at home) and outa touch with the facts of the debate. it was sorta embarrassing for those of us against the repeal (and was touted big-time by the pro-repeal people). Bernie also endorsed IRV. even Barack Obama had been noted to have endorsed it. but all of this is in the context of FairVote selling IRV and Ranked-Choice Voting as one-and-the-same, which of course they are not. this was why i wrote my first paper on the subject (in March 2009 as soon as i found out that there was a Condorcet winner and IRV failed to elect him) about how IRV *failed* to accomplish the very purposes we had adopted it for. but i was not for the repeal. i was hoping the repeal effort would lose, just barely, and then i wanted to try to get the anti-IRV people to consider Condorcet. but as it happened, the "single mark ballot" people have been vindicated for at least a generation. sad. -- 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] Kristofer: Yes, maybe Condorcet could overcome its disadvantages.
On 04/27/2012 07:32 PM, Michael Ossipoff wrote: ...but wouldn't it be better not to have the disadvantags? Sure. It would be better to make Arrow and G&S go away, but it's not going to happen. If you want to have Condorcet, you have to give up some desiderata. If you want FBC, you have to give up quite a bit, even for "weak" FBC which is fairly strong by itself. As you suggested, if Condorcet were proposed municipally, maybe (as was the case with IRV in some municipalities) big outside money won't come in to emphasize rank-balloting methods complete novelty, unknown-ness, unfamiliarity and unpredictability. They didn't emphasize any of the weird properties of IRV. Warren et al. did that. What they *have* emphasized is the unnatural nature of any method that isn't Plurality, and that is a hurdle to Approval, too. Yes, you may counter by saying you can explain to people that Approval is "setwise voting" and thus is one man one vote (or that in Approval, any vote can be countered by an opposite inverted vote), but that still means you have to show people how to look beyond their immediate intuition. But maybe the reason why that wasn't always done to IRV proposals was because the big opponents knew that IRV wasn't going to really improve on Pluralilty anyway, given current electorates. I doubt it. Of course it could be the case that high-profile people who supported IRV, like Howard Dean, were complicit. But if so, they're doing a *really* good job at faking their enthusiasm. And you further suggested, when people become familiar with Condorcet's properties, by experience in municipal use, maybe they'll realize that FBC violations are rare--usually no one will regret ranking their favorite in 1st place. Right, and a good enough boxer might be able to win while holding one hand behind his back. Heroic, but maybe not the best idea. Arrow and G&S implies every "boxer" has to hold something behind his back. Why would we want to have the disadvantages of Condorcet, if they can be avoided? Because we want the *advantages* of Condorcet. Advantages like rank expressivity when at least some significant fraction of the voters are honest. Advantages like the ease that significant fraction can vote without having to use dice, know their preferences to the level of olympic judges, or keep updated with iterations of polling; and advantages like defensibility of the victor when there's a CW, both directly and indirectly. The CW would beat any other candidate one-on-one in a runoff; on the other hand, Approval's "the candidate most people approve of" begs the question because Approval "approvals" are so often strategic. If enough people are honest, Condorcet's defensibility is real, too. It means no angry majority can be unified behind a repeal simply by the cause "you didn't like the winner". Not only would approval have an incomparably better chance for relatively quick national adoption, but Approval (nothing other than an elimination of Plurality's nonsensical falsification-requirement) rightfully qualifies as a voting-rights issue, a rights remedy that could and should be required by the courts. Alright, if you can push directly for a national change, do so. I'm not going to stop you :-) However, I think (and this may be wrong, of course, but it's the impression I get) that national change is pretty locked tight, at least for something as unfamiliar as voting reform. Voting affects everybody, and it's brittle - use a nonfunctioning method and democracy is compromised. It's not like the tax code or an environmental tweak where small changes to the regulation leads to small changes in the outcome. And, if anyone is going to propose something more complicated than Approval, it should be something that better gets rid of the co-operation/defection problem (C/D). ICT greately mitigates that problem, is defection-resistant. That can't be said for Condorcet. Compare Condorcet and ICT in the usual Approval bad-example, the 27,24,49 example. CT is Condorcet (Smith, even), and you said CT handles the defection problem. So Condorcet in itself doesn't preclude defection-resistance. So if you can find a Condorcet method that passes defection resistance as well as the criteria that the advanced Condorcet methods do (clone independence, ISDA), that would be something. But we might have to pick our criteria compliances even then. Schulze argues that his method, by finding strongest beatpaths, makes it possible to rebut any arguments of the form "X won yet Y beats X" by a strong claim of the sort "but X beats Z beats W beats Y and each of those victories are stronger than Y's victory over X". It might very well be true that beatpath methods can't solve the defection problem; if so, one would have to choose between strategic resistance within the method ( defection problem resistance) and resistance outside of the method (from arguments that the winner did not des
[EM] Mathematical equivalence (was Re: Democracy Chronicles, introductions)
On 04/24/2012 08:37 PM, Richard Fobes wrote: In the non-mathematical world the word "equivalent" means "having similar or identical effects" which allows for not _always_ being _identical_ in _all_ respects. That is the context for usage in the Democracy Chronicles article. A context which is overriden by prefixing the word "equivalent" with "mathematically". Even in a rigorous academic mathematical context, "equivalent" means "having virtually identical or corresponding parts." In this context VoteFair popularity ranking is "virtually identical" to the Condorcet-Kemeny method because the word "virtually" allows for the _extremely_ _rare_ cases in which there are more than six candidates in the Smith set (which can possibly cause a difference in which candidate is declared the winner), and allows for an election involving, say, 30 candidates that _can_ (but may not) result in different full rankings between the two methods. Consider two functions f and g defined on the integers. f(x, y) = x + y, g(x, y) = x + y when |x-y| > 2 = x * y otherwise. Is f(x,y) mathematically equivalent to g(x, y)? I doubt it, even though the range of (x,y) pairs for which f(x, y) = g(x, y) is enormously greater than the range of (x,y) pairs for which that is not true*. But, of course, I should find something more weighty than my opinion and Andy's. The paper at http://arxiv.org/abs/1009.3068 has the title "two mathematically equivalent versions of Maxwell's Equations". The synopsis then contains: "Here, we first show that there are two versions of Maxwell's equations...", i.e. that there are two ways of formulating the same equations. Furthermore, the Wikipedia page on Gaussian gravity ( https://en.wikipedia.org/wiki/Gauss%27s_law_for_gravity ) uses the term "mathematically equivalent" multiple times -- to mean "mathematically the same" in each case, and http://cvxmod.net/builtinatoms.html states its Euclidean norm function is equivalent to sqrt(sum(square(expr))) - i.e. square root of sum of squares, which again is an exact correspondence. I have not, however, found any pages or papers where "mathematically equivalent" is used to mean "virtually identical". The closest I could get was http://www.mathworks.se/help/toolbox/nnet/ref/tansig.html which says, quoting: "[The function] is mathematically equivalent to tanh(N). It differs in that it runs faster than the MATLAB implementation of tanh, but the results can have very small numerical differences". Still, here the text says that while the function does, in a mathematical sense, give identical results to tanh(N), roundoff errors can lead to different results in practice. In other words, "mathematical equivalence does not imply computational equivalence". That differs from the VoteFair case, because in the latter case, the algorithm itself gives a different result, even if you were to run it on rational number arithmetic rather than floating point. Finally, in logic, two statements are equivalent in the case that they mutually imply each other. "X is greater than Y" is equivalent to "Y is less than X". While logic is not mathematics, the latter is relatively close to the former, and it would be more surprising if the definition of equivalence was changed than if it was not. - * If you state that it doesn't count, since there are infinite integer pairs for which g(x, y) will give x+y and for which g(x, y) will give x*y, then the argument works just as well if you restrict f and g to the integers between, say, +1 billion and -1 billion. Election-Methods mailing list - see http://electorama.com/em for list info
[EM] Democracy Chronicles, introductions
As long as everyone is somewhat comfortable with keeping the "mathematically equivalent" wording, I think we can move forward with the article. I put together the more complete text of the article with the interview included and some additions to the intro to remind readers of the group and the series of articles. I am hoping we can finish discussion of the article and publish this week. Please make any suggestions for changes before mid-week and I will go ahead and publish. Don't worry about formatting because I will clean the article up for publication. Mr. Fobes, If you have any images from Votefair I can use on the article or any other images, please let me know. I have been reliant of Flickr and Wikipedia for the entire website. Thanks. The Fobes article: EXCLUSIVE INTERVIEW: Election Reformer Speaks With Democracy Chronicles! Author and Activist Richard Fobes Discusses His Ideas For Election Reform in the US Over the course of a series of articles Democracy Chronicles is presenting the results of the fascinating interviews we have conducted with prominent signers of the group that published the 'Declaration of Election-Method Reform Advocates'. The interviews will cover the opinions of a diverse group of election experts from around the world. These interviews could not have been accomplished without the determined help of author and election reform advocate Richard Fobes. In a small token of our appreciation for his efforts, we are publishing his interview here as the first of the series of interviews exclusively on Democracy Chronicles. Richard Fobes, who has a degree in physics, became involved with election-method reform when he realized, while writing his book titled "The Creative Problem Solver's Toolbox" [link], that most of the world's problems can be solved, but the current voting methods used throughout the world are so primitive that citizens are unable to elect the problem-solving leaders they want. That insight motivated him to spend time over the last two decades developing and writing open-source software for a system of voting methods that he calls "VoteFair ranking." The core of the system is VoteFair popularity ranking, which is mathematically equivalent to the Condorcet-Kemeny method, one of the methods supported by the "Declaration of Election-Method Reform Advocates." At his VoteFair.org [link] website, Fobes offers a free service of calculating VoteFair ranking results, and a number of organizations have used the service to elect their officers. At that site Fobes also hosts an American Idol poll that allows fans of the TV show to rank the show's singers according to who is their favorite, who is their second favorite, and so on down to who they like the least, and the calculations reveal the overall ranking. Based on the results, Fobes writes commentaries that anticipate and explain so-called "surprise" results in terms of important voting concepts, especially vote splitting, vote concentration, and strategic voting. Below, Mr. Fobes answers the questions of Democracy Chronicles' Adrian Tawfik who recently conducted the interview online: Democracy Chronicles: Briefly explain what characteristics you think are most important for a voting method to have? Richard Fobes: To produce fair results, a voting method should look deeply into the voter preferences. The current approach of voters only being allowed to mark a single choice, and then using an overly simplistic counting method (plurality), is a huge failure to look beneath the surface of voter preferences. In contrast, I think a voter should be allowed to rank all the candidates from most preferred to least preferred, and the counting method should fully rank all the choices from most popular and second-most popular down to least popular. If a method correctly identifies the least-popular choice, then voters can better trust that the method also correctly identifies who deserves to win. Democracy Chronicles: What do you think is the most important election reform needed where you live (either locally or nationally)? Why is this reform important? Richard Fobes: I believe that the election reform that is most needed in the United States is to ban the use of single-mark ballots in Congressional elections, including primary elections. This ban would allow us, the majority of voters, to fill Congress with problem-solving leaders instead of special-interest puppets. This reform is more important than reforming Presidential elections because the job of the President is to enforce the laws that Congress writes, and because it would dramatically weaken Congressional lobbyists (who have far more power than Presidential advisers). Democracy Chronicles: What is your opinion on other aspects of election reform such as reforming money's role in politics or redistricting (particularly in the US but very interested as well concerning election reforms intern
