Re: [Election-Methods] [english 95%] Re: [english 94%] Re: method design challenge +new method AMP
Dear Juho,
this sounds nice -- the crucial point is that we'll have to analyse what
strategic voters will vote under that method! Obviously, it makes no
sense to the A voters to reverse their CB preference since that would
eliminate C instead of B and will result in B winning instead of C...
Did you look deeper into the strategic implications yet?
Yours, Jobst
P.S. It is quite easy to use also other methods than STV since the
combinatorics are not a problem. There are only n different possible
outcomes of the proportional method (if there are n candidates). In
this example it is enough to check which one of the sets {A,B}, {A,C}
and {B,C} gives best proportionality (when looking at the worst
candidates to be eliminated from the race).
Juho
On May 2, 2008, at 23:59 , Juho wrote:
Here's an example on how the proposed method might work.
I'll use your set of votes but only the rankings.
51: ACB
49: BCA
Let's then reverse the votes to see who the voters don't like.
51: BCA
49: ACB
Then we'll use STV (or some other proportional method) to select 2
(=3-1) candidates. STV would elect B and A. B and A are thus the
worst candidates (proportionally determined) that will be eliminated.
Only C remains and is the winner.
- I used only rankings = also worse than 52 point compromise
candidates would be elected
- I didn't use any lotteries = C will be elected with certainty
Juho
On May 2, 2008, at 22:29 , Jobst Heitzig wrote:
Dear Juho,
I'm not sure what you mean by
How about using STV or some other proportional method to select
the n-1 worst candidates and then elect the remaining one?
Could you give an example or show how this would work out in the
situation under consideration?
Yours, Jobst
Juho
On Apr 28, 2008, at 20:58 , Jobst Heitzig wrote:
Hello folks,
over the last months I have again and again tried to find a
solution to
a seemingly simple problem:
The Goal
-
Find a group decision method which will elect C with near
certainty in
the following situation:
- There are three options A,B,C
- There are 51 voters who prefer A to B, and 49 who prefer B to A.
- All voters prefer C to a lottery in which their favourite has 51%
probability and the other faction's favourite has 49% probability.
- Both factions are strategic and may coordinate their voting
behaviour.
Those of you who like cardinal utilities may assume the following:
51: A 100 C 52 B 0
49: B 100 C 52 A 0
Note that Range Voting would meet the goal if the voters would be
assumed to vote honestly instead of strategically. With strategic
voters, however, Range Voting will elect A.
As of now, I know of only one method that will solve the problem
(and
unfortunately that method is not monotonic): it is called AMP
and is
defined below.
*** So, I ask everyone to design some ***
*** method that meets the above goal! ***
Have fun,
Jobst
Method AMP (approval-seeded maximal pairings)
-
Ballot:
a) Each voter marks one option as her favourite option and may
name
any number of offers. An offer is an (ordered) pair of options
(y,z). by offering (y,z) the voter expresses that she is
willing to
transfer her share of the winning probability from her
favourite x to
the compromise z if a second voter transfers his share of the
winning
probability from his favourite y to this compromise z.
(Usually, a voter would agree to this if she prefers z to
tossing a
coin between her favourite and y).
b) Alternatively, a voter may specify cardinal ratings for all
options.
Then the highest-rated option x is considered the voter's
favourite,
and each option-pair (y,z) for with z is higher rated that the mean
rating of x and y is considered an offer by this voter.
c) As another, simpler alternative, a voter may name only a
favourite
option x and any number of also approved options. Then each
option-pair (y,z) for which z but not y is also approved is
considered
an offer by this voter.
Tally:
1. For each option z, the approval score of z is the number of
voters
who offered (y,z) with any y.
2. Start with an empty urn and by considering all voters free for
cooperation.
3. For each option z, in order of descending approval score, do the
following:
3.1. Find the largest set of voters that can be divvied up into
disjoint
voter-pairs {v,w} such that v and w are still free for
cooperation, v
offered (y,z), and w offered (x,z), where x is v's favourite and
y is
w's favourite.
3.2. For each voter v in this largest set, put a ball labelled
with the
compromise option z in the urn and consider v no longer free for
cooperation.
4. For each voter who still remains free for cooperation after
this was
done for all options, put a ball labelled with the favourite
option of
that voter in the urn.
5. Finally, the winning option is determined by drawing a ball
from the
urn.
(In rare cases, some tiebreaker may be needed in step 3 or 3.1.)
Why this meets the goal: In the described
Re: [Election-Methods] [english 94%] Re: method design challenge +new method AMP
Juho wrote: Here's an example on how the proposed method might work. I'll use your set of votes but only the rankings. 51: ACB 49: BCA Let's then reverse the votes to see who the voters don't like. 51: BCA 49: ACB Then we'll use STV (or some other proportional method) to select 2 ? (=3-1) candidates. STV would elect B and A. B and A are thus the ? worst candidates (proportionally determined) that will be eliminated. ? Only C remains and is the winner. This is not? clone independent. 52: ACB 48: BCA B+A 'elected', so C wins However, if it is changed to 26: A1A2CB 26: A2A1CB 48: BCA1A2 Since 3 are now elected, it requires 25% of the vote per candidate elected. the 52 block can 'elect' B and C and the 48 block elects A2. This means that A1 wins as he isn't picked. One possible solution to the clone issue is to scale the number of candidates to first choice vote. In effect, in the STV stage, the quota for each candidate would be equal to the number of first choice votes the candidate received. To be 'elected', the candidate would have to exceed the quota. The first candidate to be eliminated becomes the winner. Reversed votes: 26: BCA1A2 26: BCA2A1 48: A2A1CB Quotas (number of first choices in original ballots): A2: 26 A1: 26 C: 0 B: 48 Round 1: A1: 0 A2: 48 B: 52 C: 0 B exceeds quota by 4 and A2 exceeds quota by 22 Round 2: A1: 22 (-26) A2*: 26 (+26) B*: 48 (-4) C: 4 (+4) C is elected though, so A1 still wins. Note this is clone independent though: Quotas A: 52 B: 48 C: 0 Round 1: A: 48 B: 52 C: 0 B exceeds quota Round 2: A: 48 B: 48 C: 4 C exceeds quota A wins. However, if 5 voters voted C first choice, then C would be eliminated as being on the lowest total. One option for that would be to allow people cast a nominating vote as well as the ranking.? The total number of nominee votes would become the quota for each candidate.? If a reasonable number of people (5%) recognised C as a compromise, then he would win. I am not sure of the tactical issues associated with the 2 votes though. Also, it is majority compliant.? If a majority support a candidate first choice (i.e. first choice and nominate him), then he cannot lose. Another issue is how to actually layout the ballot.? It might be worth having voters enter the reversed ballot order.? In most practical cases, voters would need to enter their lowest ranked candidates, unlike in normal STV where it would be their most ranked. The ballot instructions could be something like: Place an X beside the candidate you wish to nominate in the nominate column In the rank column, rank the candidates in order of your preference giving a rank of 1 to your least favourite, 2 to your next least favourite and so on You do not have to rank all the candidates and any you do not rank will be considered preferred to any ranked candidate Raphfrk Interesting site what if anyone could modify the laws www.wikocracy.com AOL's new homepage has launched. Take a tour at http://info.aol.co.uk/homepage/ now. Election-Methods mailing list - see http://electorama.com/em for list info
[Election-Methods] Measuring satisfaction in a multi winner election
Hi bunch, this mail oriented me toward a nother subject I like: Measuring satisfaction among voters. When comparing the result to the possible candidates, one can determine its level of satisfaction by the proportion of candidates elected compared to the number that received support form the voter. In FPTP, each voter is either fully satisfied or fully unsatisfied. Hence the global satisfaction rate is the average of the ballot fraction received by each elected member. With a multi-seat method, the same technic can be applied for each specific group of voters who agree on the same elected members, counting a fraction of satisfaction proportional to the ratio of elected/wanted representativ. For example, with 3 available seats and ten candidates (A to J), let's analyse the global satisfaction: 45% of voters : A B C J ... 5% of voters: B C A J E ... 30% of voters: G H I A C J ... 20% of voters: J A B ... Outcome A, B and C elected. (this is a typical STV outcome) Thus the first 50% (45%+5%) are fully satisfied: 100% The 30% are not satisfied: 0% The 20% got 2 elected representative among their 3 first candidates: 66.6% of satisfaction. Global satisfaction: 50% x 100% + 20% x 66.6% = 63.3% Using another method (approbation with the cut-off at ... for example): Outcome A (100%), J (100%), C(80%). Establishing satisfaction at this point is more complex I admit. Some will argue that with approbation philosophy 100% satisfaction is reached when the 3 elected are among the approved candidates. Ranking partisans will argue that the preference ordering still exist and only the first 3 preference should be considered for a fair comparison between systems. I am an in-betweener: sometimes we don't have the detail of the preferences (a real approval ballot does not give this information) but on the other way approving 10 person and having my 8th, 9th and 10th picks elected would not fully satisfy me. Thus I split equally satisfaction among each apporved candidates: 4 approved candidates = 25% satisfaction each. The average reflects my understanding of the measurement. Satisfaction of the 45% group: 75% Satisfaction of the 5% group: 60% Satisfaction of the 30% group: 50% Satisfaction of the 20% group: 66.6% Global satisfaction: 45% x 75% + 5% x 60% + 30% x 50% + 20% x 66.6% = 65.1% The best electoral system should maximize global satisfaction. In this particular example: Approval gives a better result than STV. I invite you to measure satisfaction with your preferred multi-winner method. SPPA (my favorite) produces high satisfaction levels. Stéphane Kevin Venzke a écrit : Hi, --- Howard Swerdfeger [EMAIL PROTECTED] a écrit : but most of these reforms fail to recognize that that Seats do not equal power. So we are still still stuck with a similar problem (votes != power) I was looking into 2 methods of measuring power in a weighted voting system. http://en.wikipedia.org/wiki/Banzhaf_Power_Index http://en.wikipedia.org/wiki/Shapley-Shubik_power_index I was wondering first if there are any methods of measuring power in a legislature that I am unaware of? Secondly if anybody has tried to design a generic system where by votes are kept proportional to power, via allocation of seats? I find this question very interesting... But I am guessing that you don't have many allocation possibilities, especially with a small number of factions. Another thing: I guess it wouldn't be cloneproof. Say there's normally only three parties and everybody votes for a party list. I guess a party could gain an advantage by running two lists instead of one. Kevin Venzke _ Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! Mail 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: [Election-Methods] RE : Re: Is the Condorcet winner always the best?
Satisfaction analysis should help answer your question Diego Santos a écrit : I was not enough clear when i wrote my previous email. The '' is not a real approval mark on the ballot, it was only a satisfaction limit from each voter. I am arguing that not always the Condorcet winner is the one that maximizes happiness of the people, as Jonathan pointed. A approval quorum rule will avoid low utility CW to win. And, opposit to Jonanthan argument, an approval cuttoff does not add too much complexity: it is like a hypothetical candidate NOTB (none of the below). 2007/12/11, Dave Ketchum [EMAIL PROTECTED] mailto:[EMAIL PROTECTED]: On Tue, 11 Dec 2007 12:20:49 -0800 Jonathan Lundell wrote: On Dec 11, 2007, at 6:05 AM, Kevin Venzke wrote: Jonathan, --- Jonathan Lundell [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] a écrit : ...should choose B as a good compromise, with the A voters saying A is good, B OK, C very bad. But Diego's profile suggests to me that the A voters are saying something like A is good, B is bad, C is very bad. Not that they can express it in a normal linear ballot, just that we're being told a little more about their opinions. In my opinion, to the extent that the effect of a badverybad vote is disregarded, the point of letting voters indicate such preferences is undermined anyway. I'm not advocating it as a ballot option, only as a meta-notation shorthand to give us kibitzers a little more information about the voters' utility functions. In my example, the effect of a later-no-harm voting rule is evident. In Diego's, a rule (such as STV) that elects A doesn't seem unreasonable to me. The problem is that with an ordinary linear ballot (no ''), we can't distinguish between the cases. Not that I'm arguing that we should employ ''; offhand, that strikes me as a complication to be avoided. In one sense I don't agree. If is allowed then apparently it's safe to vote badverybad. If isn't allowed then voters will probably be more cautious, since the method could very well take them as serious if they say that bad is better than verybad. I tend to think that if B doesn't win in Diego's scenario, the method is second-guessing the voters. It either disbelieves the C voters' preference for B over A, or finds that there's something more important than majority rule. There's a reasonable argument to be made (hardly originally by me) on either side of the question of whether a compromise candidate is sometimes (or always) better to the candidate of one faction in a close election. If the vote were: 53 A 47 C ...we'd shrug and call it a fairly close election, or at least no landslide, and forget about it, even if all 100 voters strongly disapproved of the opposing candidate. If we introduce a third candidate whom the A and C voters despise only slightly less than C and A respectively, and end up with something like Diego's profile, we have 100 (or 90 in that profile) unhappy voters instead of 47. A and C agree that B is better than their standard enemy. C voters will be happy to help install B, since this is better than installing A. A voters may be a bit unhappy, but they at least avoided installing C. Probably A supporters will be too unhappy, because their favorite candidate would win if B was not nominated. I'm not saying that it's unarguable, nor that the voting system should somehow anticipate the situation (through the use of '', for example). I think it's a fuzzy case with no perfect answer, and that we don't really want to make the ballot more complex, or add to the possibilities for manipulation that such a rule would entail. I'm just saying that it's not obvious that, in all cases, the best rule is the one that lets B win. Choices can be hard. Get far enough from a tie and A or C will win. If we manage a cycle we can debate the results of that. -- [EMAIL PROTECTED] mailto:[EMAIL PROTECTED]people.clarityconnect.com/webpages3/davek http://people.clarityconnect.com/webpages3/davek Dave Ketchum 108 Halstead Ave, Owego, NY 13827-1708 607-687-5026 Do to no one what you would not want done to you. If you want peace, work for justice. Election-Methods mailing list - see http://electorama.com/em for list info -- Diego Santos Election-Methods mailing list - see http://electorama.com/em for list info Election-Methods mailing list - see
Re: [Election-Methods] IRV ballot is at least as fair as FPTP ballot
Again satisfaction analysis can be used to objectively determine which of IRV and FTP produces the best outcome. Using enough election data, one could even measure how often IRV may elect the candidate not favored by most voters. My humble estimation is rarely (1/50 times). In comparison I estimate FPTP outcomes to be deficient (1/5 times) and condorcet methods (1/200 times). I qualify a method to be deficient when another outcome would produce a better global satisfaction. Stéphane Kathy Dopp a écrit : Date: Wed, 26 Dec 2007 19:33:27 +0100 (CET) From: Kevin Venzke [EMAIL PROTECTED] Subject: [Election-Methods] RE : Re: IRV ballot is at least as fair as FPTPballot Yes some voters have second-choice considered but they are all still treated equally. So, you define equal as your ballot's first and second choices count, but someone else's first and second choices do not. I and many others may disagree with your definition of equal treatment and I hope that the courts do too since IRV may often elect the candidate not favored by most voters. But it does this according to what it believes each voter wants. If your Well I don't want some voters' second choices given consideration but not all voters' second choices given consideration. Will it take what I believe into consideration? No. first preference wins the election, then you don't want your second preference to be counted. If your first preference is a very weak candidate, then you want him to be eliminated so that your second preference can be counted. One problem is that my second choice candidate may be eliminated in the first round and my first choice candidate not have success either - despite the fact that my second choice candidate is the most popular among all voters. For instance, this example, which is one of countably infinite examples where IRV elects the candidate not supported by most voters: Republican Libertarian Progressive Democrat 1st choice 4 3 3 2 2nd choice 1 2 1 7 3rd choice 1 1 6 1 I.e. in this example with 12 voters, the Democrat loses in the first round, even though the most number of persons supported the Democrat overall - letting the Republican win, even though the Republican (in this example) is not as widely supported as to other candidates. I created a list of the 12 voters and their choices in a spreadsheet. I suggest you do a little experimenting on your own because I do not have time to do the analysis for you because I am too busy to spend my time disabusing you of a fiction you hold. I am working on other more critical matters. Please do the analyses yourself with a spreadsheet so you can see how trivially easy it is to make IRV put the wrong candidate, not supported by most voters into office. For example in Florida, what if the true first choice of most Democratic voters had been Nader in 2000, then IRV would have immediately knocked the Democrat out of the race, enabling the Republican to win. (BTW, Nader had virtually nothing to do with Gore losing Florida in 2000 - there was 1. electronic fraud in one county that robbed thousands of votes from Gore temporarily, 2. tons of illegal undated military ballots sent into election offices during the election contest and when the Dems tried to challenge the illegal ballots, they were intimidated by being told they were not patriotic, and 3. Katherine Harris removed thousands of legal voters off the rolls primarily in Black Democratic districts. IRV would have made sure that Gore didn't win, even without all those other election frauds or voter disenfranchisements. IRV is factually just not capable of doing what it claims to be able to do. Kathy For this reason it's not obvious which voters are put at a disadvantage. IRV not only treats voters' ballots very differently, it ensures that there are numerous ways that a candidate is declared a winner who is supported by fewer voters overall than a candidate who loses in the first round. This fact is irrefutable, obvious and simple. Just try some scenarios out in any spreadsheet. It is a valid criticism that IRV can elect a candidate with less support than some other candidate. This makes me wonder what election methods you do like, since first-preference plurality voting already has the same issue. Just because the plurality ballot doesn't ask for support doesn't mean the concept doesn't exist. In any case, people on this list who dislike IRV still would not want to argue that it should be illegal, since this could set a precedent that prevents other, arguably better methods from being adopted. IRV would only be fair and treat all voters equally if all first AND second choices of all voters were tabulated, with the second choices being given some weight less than the first - ONLY then would IRV not routinely allow numerous ways to
Re: [Election-Methods] IRV unconstitutional? (replies)
Yes those flaws exist. But their FPTP equivalent (vote-splitting) happens very more often than the sum of occurence of the previously cited. Warren Smith a écrit : St.Rouillon: IRV defendors should aim at showing that IRV flaws are smaller than FPTP flaws, thus FPTP should be declared anti-constitutional if IRV is. WDS replies: IRV has some flaws which FPTP does not have. For example, non-monotonicity, no-show paradoxes, non-additivity. I personally think IRV *is* better than FPTP (at least if we can ignore issues of simplicity and fraud worries) but not in these respects. For real-world examples of non-monotonic IRV elections featuring no-show paradoxes, see http://rangevoting.org/Ireland1990.html http://rangevoting.org/LizVwiz.html DonCathy Hoffard: any time a new candidate X entering the race swings the winner from Y to Z, that benefits somebody (namely Z, here) This is not true in most if not all of the [IRV] the General Elections. 90-99% of the General elections involve two major candidates and some minor candidates. The winner will be one of the major candidates. In IRV the minor candidates votes are drop and their votes are now cast for one of the major candidates. You are right in some cases where you have 3 equal candidates. WDS replies: I gave constructed examples before of IRV elections where a candidate by entering race swings the winner. E.g. http://rangevoting.org/CoreSupp.html . So yes, I am right. For a real-world example, in the Louisiana 1991 governor race, see http://rangevoting.org/LizVwiz.html Duke by entering the race caused Edwards to win, whereas otherwise Roemer would have won. So your if not all is wrong - there is at least one counterexample. Another is Peru 2006: http://www.rangevoting.org/Peru06.html . So there are at least 2 counterexamples now. Indeed, theso-called center squeeze effect in IRV is where it is Leftist vs Centrist vs Rightist. Centrist is the Condorcet beats all winner, but is eliminated by IRV because the left rightists squeeze him into too small a regio of top-place support. In EVERY such situation, one extremist, by enteringthe race, swung it to the other whereas without him,Centrist would have won. THis is quite common: in 1 dimensional politics, this happens 1/3 (33%) of the time to IRV. How we know that: see http://rangevoting.org/IrvPathologySurvey.html#csqueeze So you are quite wrong. This is not rare if at all. It is common. The error in your analysis was to only consider the minor guy as entering, and to neglect the major guy as an entrant. Oops. When you only consider some possibilities you naturally get a lesser count than if you consider them all. The underlyign reason for your error was your USA-2007-centric thinking, failing to even consider the possibility that a so-called minor candidate might actually be a Condorcet winner. IRV leads to 2-party domination (a flaw it shares with FPTP) which somewhat justifies your error, but that is another problem. :) -- Warren D. Smith http://RangeVoting.org -- add your endorsement (by clicking endorse as 1st step) and math.temple.edu/~wds/homepage/works.html 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: [Election-Methods] IRV ballot is at least as fair as FPTP ballot
On Sat, May 3, 2008 at 10:26 AM, Stéphane Rouillon [EMAIL PROTECTED] wrote: Again satisfaction analysis can be used to objectively determine which of IRV and FTP This statement does not make logical sense because measuring feelings like satisfaction is not an objective measure. For example just because most voters have confidence that their invisible electronic ballot are cast and counted accurately, does not make it so. produces the best outcome. Using enough election data, one could even measure how often IRV may elect the candidate not favored by most voters. This statement also does not make logical sense for at least two reasons: 1. because voting behavior and strategy changes depending on the counting methodology, so you can not learn anything about whether how often the elected candidate would be the one most favored by voters from examining other voting schemes, and 2. unless you can mind read and interview every voter and know what strategy they used when voting, and also unless you can review every ballot, not just the summary data you may be able to obtain from some election officials, you may not be able to accurately judge most favored by voters from looking at election data. I'm going to ignore the rest of your comments since your two opening comments are logically invalid. Kathy My humble estimation is rarely (1/50 times). In comparison I estimate FPTP outcomes to be deficient (1/5 times) and condorcet methods (1/200 times). I qualify a method to be deficient when another outcome would produce a better global satisfaction. Stéphane Kathy Dopp a écrit : Date: Wed, 26 Dec 2007 19:33:27 +0100 (CET) From: Kevin Venzke [EMAIL PROTECTED] Subject: [Election-Methods] RE : Re: IRV ballot is at least as fair as FPTP ballot Yes some voters have second-choice considered but they are all still treated equally. So, you define equal as your ballot's first and second choices count, but someone else's first and second choices do not. I and many others may disagree with your definition of equal treatment and I hope that the courts do too since IRV may often elect the candidate not favored by most voters. But it does this according to what it believes each voter wants. If your Well I don't want some voters' second choices given consideration but not all voters' second choices given consideration. Will it take what I believe into consideration? No. first preference wins the election, then you don't want your second preference to be counted. If your first preference is a very weak candidate, then you want him to be eliminated so that your second preference can be counted. One problem is that my second choice candidate may be eliminated in the first round and my first choice candidate not have success either - despite the fact that my second choice candidate is the most popular among all voters. For instance, this example, which is one of countably infinite examples where IRV elects the candidate not supported by most voters: Republican Libertarian Progressive Democrat 1st choice 4 3 3 2 2nd choice 1 2 1 7 3rd choice 1 1 6 1 I.e. in this example with 12 voters, the Democrat loses in the first round, even though the most number of persons supported the Democrat overall - letting the Republican win, even though the Republican (in this example) is not as widely supported as to other candidates. I created a list of the 12 voters and their choices in a spreadsheet. I suggest you do a little experimenting on your own because I do not have time to do the analysis for you because I am too busy to spend my time disabusing you of a fiction you hold. I am working on other more critical matters. Please do the analyses yourself with a spreadsheet so you can see how trivially easy it is to make IRV put the wrong candidate, not supported by most voters into office. For example in Florida, what if the true first choice of most Democratic voters had been Nader in 2000, then IRV would have immediately knocked the Democrat out of the race, enabling the Republican to win. (BTW, Nader had virtually nothing to do with Gore losing Florida in 2000 - there was 1. electronic fraud in one county that robbed thousands of votes from Gore temporarily, 2. tons of illegal undated military ballots sent into election offices during the election contest and when the Dems tried to challenge the illegal ballots, they were intimidated by being told they were not patriotic, and 3. Katherine Harris removed thousands of legal voters off the rolls primarily in Black Democratic districts. IRV would have made sure that Gore didn't win, even without all those other election frauds or voter disenfranchisements. IRV is factually just not capable of doing what it claims to be able to do. Kathy For this reason it's not obvious which voters are put at a disadvantage. IRV not only treats voters' ballots very differently, it
Re: [Election-Methods] IRV ballot is at least as fair as FPTP ballot
Of course, I supposed that the information provided from ballots was sincere And I supposed that the outcome that would have been obtained if all voters had voted like a particular voter would give 100% satisfaction to this particular voter. Stéphane Kathy Dopp a écrit : On Sat, May 3, 2008 at 10:26 AM, Stéphane Rouillon [EMAIL PROTECTED] wrote: Again satisfaction analysis can be used to objectively determine which of IRV and FTP This statement does not make logical sense because measuring feelings like satisfaction is not an objective measure. For example just because most voters have confidence that their invisible electronic ballot are cast and counted accurately, does not make it so. produces the best outcome. Using enough election data, one could even measure how often IRV may elect the candidate not favored by most voters. This statement also does not make logical sense for at least two reasons: 1. because voting behavior and strategy changes depending on the counting methodology, so you can not learn anything about whether how often the elected candidate would be the one most favored by voters from examining other voting schemes, and 2. unless you can mind read and interview every voter and know what strategy they used when voting, and also unless you can review every ballot, not just the summary data you may be able to obtain from some election officials, you may not be able to accurately judge most favored by voters from looking at election data. I'm going to ignore the rest of your comments since your two opening comments are logically invalid. Kathy My humble estimation is rarely (1/50 times). In comparison I estimate FPTP outcomes to be deficient (1/5 times) and condorcet methods (1/200 times). I qualify a method to be deficient when another outcome would produce a better global satisfaction. Stéphane Kathy Dopp a écrit : Date: Wed, 26 Dec 2007 19:33:27 +0100 (CET) From: Kevin Venzke [EMAIL PROTECTED] Subject: [Election-Methods] RE : Re: IRV ballot is at least as fair as FPTP ballot Yes some voters have second-choice considered but they are all still treated equally. So, you define equal as your ballot's first and second choices count, but someone else's first and second choices do not. I and many others may disagree with your definition of equal treatment and I hope that the courts do too since IRV may often elect the candidate not favored by most voters. But it does this according to what it believes each voter wants. If your Well I don't want some voters' second choices given consideration but not all voters' second choices given consideration. Will it take what I believe into consideration? No. first preference wins the election, then you don't want your second preference to be counted. If your first preference is a very weak candidate, then you want him to be eliminated so that your second preference can be counted. One problem is that my second choice candidate may be eliminated in the first round and my first choice candidate not have success either - despite the fact that my second choice candidate is the most popular among all voters. For instance, this example, which is one of countably infinite examples where IRV elects the candidate not supported by most voters: Republican Libertarian Progressive Democrat 1st choice 4 3 3 2 2nd choice 1 2 1 7 3rd choice 1 1 6 1 I.e. in this example with 12 voters, the Democrat loses in the first round, even though the most number of persons supported the Democrat overall - letting the Republican win, even though the Republican (in this example) is not as widely supported as to other candidates. I created a list of the 12 voters and their choices in a spreadsheet. I suggest you do a little experimenting on your own because I do not have time to do the analysis for you because I am too busy to spend my time disabusing you of a fiction you hold. I am working on other more critical matters. Please do the analyses yourself with a spreadsheet so you can see how trivially easy it is to make IRV put the wrong candidate, not supported by most voters into office. For example in Florida, what if the true first choice of most Democratic voters had been Nader in 2000, then IRV would have immediately knocked the Democrat out of the race, enabling the Republican to win. (BTW, Nader had virtually nothing to do with Gore losing Florida in 2000 - there was 1. electronic fraud in one county that robbed thousands of votes from Gore temporarily, 2. tons of illegal undated military ballots sent into election offices during the election contest and when the Dems tried to challenge the illegal ballots, they were intimidated by being told they were not patriotic, and 3. Katherine Harris removed thousands of legal voters off the rolls primarily in Black Democratic districts. IRV would have made sure that Gore didn't win, even without all those other election frauds or voter
[Election-Methods] Comparing multi-winner methods
Hello Kevin, these ratios are guesses I have for real elections. But I am fed up with guesses so the goal is to build an objective method able to determine for a perticular set (method, ballots expressing sincere preferences, outcome) of one electoral data, the most satisfying method in the eye of all voters. Stéphane Kevin Venzke a écrit : Hi Stéphane, --- Stéphane Rouillon [EMAIL PROTECTED] a écrit : Again satisfaction analysis can be used to objectively determine which of IRV and FTP produces the best outcome. Using enough election data, one could even measure how often IRV may elect the candidate not favored by most voters. My humble estimation is rarely (1/50 times). In comparison I estimate FPTP outcomes to be deficient (1/5 times) and condorcet methods (1/200 times). I qualify a method to be deficient when another outcome would produce a better global satisfaction. To do this you would have to be clear and consistent with your assumptions... Is global satisfaction the total utility (on some scale) of the winner? When you guess that IRV fails by this standard 1/50 times, are you considering real life elections, or random ones? If random then how many candidates and is there any underlying policy space? etc. Given real life elections I guess 1/50 may be accurate for IRV, but I don't feel this tells the whole story. Strategic nomination and voting incentives, as well as incentives created by institutions other than the voting rule itself, would not seem to be considered at all by this measure. Given random elections with even 3 candidates and the ability to truncate, I guess IRV is much, much worse than 1/50. Kevin Venzke __ Do You Yahoo!? En finir avec le spam? Yahoo! Mail vous offre la meilleure protection possible contre les messages non sollicités http://mail.yahoo.fr Yahoo! Mail 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: [Election-Methods] IRV ballot is at least as fair as FPTP ballot
On Sat, May 3, 2008 at 5:51 PM, Stéphane Rouillon [EMAIL PROTECTED] wrote: Of course, I supposed that the information provided from ballots was sincere And just what State do you live in where you will have opportunity to review the ballots that have been secured and reconciled to prevent ballot tampering, ballot box stuffing, and ballot substitution? I don't know of one state that provides that. Apparently you misspoke (like I do myself frequently) and you meant to say reviewing the actual ballots rather than election data. OK, If you can review all the secured ballots, then you can determine how voters voted, but not necessarily voter strategy that prevented some savy strategic voters from expressing exactly their preferences. However, even despite some voters using strategy because they realize that IRV fundamentally does not work the way it is intended to, you will undoubtedly find ample number of cases of candidates winning elections who were not preferred by most voters. And I supposed that the outcome that would have been obtained if all voters had voted like a particular voter would give 100% satisfaction to this particular voter. I'm not sure what your above sentence means. It seems to me that you saying that you think all voters would vote exactly alike (like a particular voter) so I doubt that is what you mean to communicate. Cheers, Kathy Election-Methods mailing list - see http://electorama.com/em for list info
Re: [Election-Methods] IRV ballot is at least as fair as FPTP ballot
At 12:29 AM 5/4/2008, Kathy Dopp wrote: However, even despite some voters using strategy because they realize that IRV fundamentally does not work the way it is intended to, you will undoubtedly find ample number of cases of candidates winning elections who were not preferred by most voters. Actually, in the majority of RCV elections that have been held recently in the U.S., elections were won by a candidate who was *opposed* by a majority of voters; that is, they voted for someone else, not for the winner. Generally, IRV is replacing top-two runoff, but, by not allowing (and, for those who show up to vote in the runoff, highly suggesting) a specific choice, it is failing to find majorities, more often than not. I found, studying top-two runoffs in San Francisco and elsewhere prior to the use of RCV, that about one-third of elections were reversing the first round result: the runner-up in the first round beat the first round leader in the runoff. However, with IRV and, what, something like thirty elections, not one example where the ranking changed in the virtual runoff compared to the first round. IRV is essentially implementing plurality. Contrary to how it is being sold. Yes, in theory, it fixes the spoiler effect, at least the first-order effect, the one that takes place in a two party system where the third party candidate can't win. But the IRV nasties show up if the third party gets uppity, it can return with a vengeance. Much, much simpler: just count all the votes. Simple. No changes to equipment. Ballots can stay basically the same, slight change in instructions. Most voters will vote the same way. But third party supporters will not fail to notice that they would now be able to vote for their favorite and, at the same time, vote for a frontrunner. No, it's not a perfect system. But it's Approval Voting, and it is a *very* good system. And to get it, we simply have to stop discarding and disregarding ballots where the voter voted for more than N in an N-winner election. Just count the votes. Bumper sticker? Election-Methods mailing list - see http://electorama.com/em for list info
