<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
Bart wrote:
<blockquote type=cite class=cite cite> How about the following:
<p>(projected vote percentages shown; assumed accurate to within +/- 5
<br>percentage points)
<p>45% A B C
<br> 5% B A C
<br> 5% B C A
<br>45% C B A</blockquote>
<blockquote type=cite class=cite cite>Adam Tarr wrote:
<br>> Specifically, there is the remarkable fact that a voter in a winning
<br>> votes-based Condorcet voting system can NEVER be hurt by fully expressing
<br>> their preferences. There are cases where fully voting your
preferences can
<br>> fail to help you, but it can never actually hurt you.
<p>Never is a strong word.</blockquote>
To Bart:
<p>This was the origin of the mail explosion.
<br>I think truncation CAN hurt a Condorcet voting system using winning
votes
<br>WHEN THERE IS NO CONDORCET WINNER.
<br>Your example has one (B).
<br>B 55 > A 45 and B 55 > C 45.
<br>Search somewhere else.
<br>Sorry, I though too you had it right
<br>the first time.
<br>I will think over to the question:
<br>"Should we consider multiple truncation?"
<br>My actual opinion is:
<br>Yes, but maybe it has a lesser probability of occurence.
<p>To Adam:
<p>I agree mostly with your following analysis.
<br>Technically, you should consider too the tie case.
<br>It is improbable but possible.
<br>But if your goal is to show that there is NO case
<br>where truncation could hurt the result from the truncator
<br>point of view and using wv, the tie case has to be treated.
<p>Adam wrote:
<blockquote TYPE=CITE>True, and I lack a rigorous proof, but every empirical
example I have seen supports this. Your example is no different...
<br>
<p>There's no sense in talking about uncertainty and ties; it only confuses
the issue. Let's just assume one or the other camp has more votes,
and see if in that light, either side has an incentive to truncate.
If neither side has such an incentive, then neither side has that incentive
in the toss-up case as well (since they know they either have more votes
or less votes than the other guy). Without loss of generality, I'll
give A the edge, which gives us:
<p>46 ABC
<br>5 BAC
<br>5 BCA
<br>44 CBA
<p>Pairwise votes are:
<p>B 56 > C 44
<br>B 54 > A 46
<br>A 51 > C 49
<p>In this case, B is the Condorcet winner.
<p>If both sides truncate we get
<p>46 A
<br>5 BAC
<br>5 BCA
<br>44 C
<p>Pairwise votes are:
<p>A 51 > C 49
<br>A 46 > B 10
<br>C 44 > B 10
<p>Now A wins the election in either winning votes or margins (don't stop
the presses yet).
<p>Now, if only the A camp truncates:
<p>46 A
<br>5 BAC
<br>5 BCA
<br>44 CBA
<p>Pairwise votes are:
<p>B 54 > A 46
<br>A 51 > C 49
<br>C 44 > B 10
<p>B wins the election in winning votes, and C wins the election in margins.
<p>Finally, if only the C camp truncates:
<p>46 ABC
<br>5 BAC
<br>5 BCA
<br>44 C
<p>Pairwise votes are:
<p>B 56 > C 44
<br>A 51 > C 49
<br>A 46 > B 10
<p>A wins the election in winning votes or in margins.
<p>OK, let's look at the decision matrix. Here is the pairwise matrix
of decisions for each camp, and the candidate elected, for each method:
<p>(I apologize in advance if the tables look lousy. Try cutting
and pasting into a text editor with uniform character spacing if it looks
bad. I used the "terminal" font type if that helps.)
<p>The top row is the tactics of the ABC faction, the left column is the
tactics of the CBA faction. T = truncate, NT = do not truncate.
<p>Margins methods:
<p><font face="Terminal, Monaco"> | T | NT |</font>
<br><font face="Terminal, Monaco">---|---|----|</font>
<br><font face="Terminal, Monaco">T | A | A |</font>
<br><font face="Terminal, Monaco">---|---|----|</font>
<br><font face="Terminal, Monaco">NT | C | B |</font>
<br><font face="Terminal, Monaco">-------------</font>
<p>Winning Votes methods:
<p><font face="Terminal, Monaco"> | T | NT |</font>
<br><font face="Terminal, Monaco">---|---|----|</font>
<br><font face="Terminal, Monaco">T | A | A |</font>
<br><font face="Terminal, Monaco">---|---|----|</font>
<br><font face="Terminal, Monaco">NT | B | B |</font>
<br><font face="Terminal, Monaco">-------------</font>
<p>OK, so what can we conclude from this? If the CBA voters truncate,
they always get A elected in either system. This is a "strictly dominated
strategy" to use the game theory name. There's no way the B voters
should truncate, regardless.
<p>In winning votes methods, truncation for the ABC voters makes no difference
(i.e. does not hurt them, even if it fails to help them). In margins
methods, truncating can prove costly for an ABC voter. This is neither
here nor there for the purposes of my analysis; I can show you a counter-example
where truncation can help in margins cases.
<p>The point is, nowhere here do we get any suggestion that a voter in
a winning votes method can be helped by
<br>truncation.</blockquote>
To Adam, Alex, Mike and all winning votes partisans:
<p>I agree on this. And thank you for showing it
<br>for both winning votes and relative margins.
<br>I will suppose only one side truncates even if it is not all clear
to me yet.
<br>Let us measure the mean gain for those special case.
<br>I will use a +1 gain for any winner ranking improvement, and a -1
<br>for any winner ranking deterioration.
<br>Winning votes for the ABC voters: mean impact: 0.
<br>Winning votes for the CBA voters: mean impact: -1.
<br>Typical impact of truncation using winning votes: -1/2.
<br>Relative margins for the ABC voters: mean impact: -1.
<br>Relative margins for the CBA voters: mean impact: -1.
<br>Typical impact of truncation using winning votes: -1.
<br>If you allow both truncation to occur you obtain the same
<br>typical impacts, only the relative margins intermediates become -1/2
and -3/2.
<p>Of course this is only a sample analysis.
<br>We should consider truncation from only a part of the group as a possibility
<br>and check if multiple truncations are equiprobable,
<br>and finally count gain/deterioration one by one before averaging this.
<br>I will soon start this analysis for small number of voters.
<p>My actual problem was to determine what probabilities to use.
<br>I came to this conclusion:
<br>when one candidate outranks all the others,
<br>he/she will win, margins or winning votes will
<br>not make the difference. It is when the run is tight
<br>between two, three or more candidates that the criteria
<br>is important. But in this case pre-electoral information
<br>is useless because, the error from polls has too much
<br>incidence over the results. Thus, I will consider
<br>any kind of ballot (truncated or not) possible with an equal
<br>probability. When we do not know what to expect,
<br>expecting any case with an equal weight seems the best approach to
me.
<blockquote type=cite class=cite cite> But then I don't see truncation
as necessarily a bad thing. If truncation can defeat a "hated middle" candidate,
it addresses my main misgiving about the Condorcet methods.</blockquote>
<blockquote type=cite class=cite cite>Much in the same way that we can't
differentiate between the indifferent voter and the lazy voter, we cannot
distinguish between the "respected (if unglamorous) compromise middle"
and the "hated (yet still) compromise middle". Smart CBA voters in
an approval election will still approve B, to defeat A, anyway. What
method would actually prevent B from winning when the voters act in a logical
manner? Even plurality and IRV encourage CBA voters to dump C for
B if they have perfect information.</blockquote>
To Adam:
<p>See my previous mail, about methods to mix ranking methods
<br>and Approval...
<p>To Forest:
<p>Thanks for reading about the universal ballot.
<br>I like letters because it has not the grade/rank misunderstanding possibility.
<br>It can make easier to difference CandidateA = CandidateB from CandidateA
? CandidateB.
<br>In the first case we can use the same letter for both candidates, and
in the second,
<br>leave both unmarked.
<br>However, I see no necessity against restricting the number of letter.
<br>If there is 10 candidates, and if I do have a ranking for all of them,
<br>please let me express it. It will not stop you from using only 2, 3,
5 or 7 levels
<br>as you wish and still being compared fairly because the method is
<br>pairwise comparison based.
<p>To Matt:
<p>Sorry but your error mail confused me a lot.
<br>I do not think M. Ossipoff and Adam want to treat
<br>unexpressed comparisons obtained by truncation as 1/2 votes.
<p>To all:
<br>I have a lot of fun and I am going to breakfast.
<p>Steph.</html>