[EM] Summary Answers

[MIKE OSSIPOFF]

Kurt--

You wrote:

1) Are there cases where you would consider a candidate outside the
Schwartz set to be the proper winner?

I reply:

Sometimes, when there are pairwise ties, Ranked-Pairs, or MAM, can choose 
outside the Schwartz set.  But that; won't happen in a public election, 
where pairwise ties are vanishingly unlikely.

Whether the possibility of choosing outside the Schwartz set is a problem in 
committee voting is a matter of opinion. It's an aesthetic issue, not a 
strategy problem.

Though RP has that aesthetic disadvantage with respect to SSD or 
BeatpathWinner, it's also true that RP has an aesthetic advantage with 
respect to those methods: With RP, a nullified defeat can't participate in 
the nullification of anothler defeat. It makes sense for it to be like that.

So there are these not-really-important aesthetic differences between RP and 
SSD or BeatpathWinner.

I think those aesthetic differences are insignificant, and one should just 
recommend or propose whichever one is most likely to be accepted. For public 
elections, for a Condorcewt version,  I like SSD, because I've encountered 
better reaction to it than to BeatpathWinner, CSSD or RP, and because SSD 
doesn't mention cycles or need them for justification. If people a really 
simple Condorcet version, then it's PC. I've talked to people on newsgroups 
who rejected anything more complicated.

You continued:

2) Are there cases where you would consider a candidate outside the
Smith set to be the proper winner?

I reply:

I expect that everyone here agrees that it's better if Condorcet methods 
choose from the Smith set. All but PC will always do that.

Of course maybe someone could want PC's defining standard, and might want 
the candidate with the least maximum pairwise vote against him/her.

You continued:

I'm curious about that in terms of criteria/strategy, but I'm also
curious about it in a larger way - because I think I remember Mike
saying that if we could be assured of sincere ballots, he'd prefer the
Borda winner - even if there were a different Condorcet Winner.

I reply:

Well, that's if we're talking about rank-counts. Without that requirement, 
then CR would be ideal if voters didn't care about optimizing their outcome, 
but only cared about voting sincerely, to do their part to maximize SU.

Some people point out that it can be a little problematic for all the voters 
to  rates incerely on the same scale, to mean the same thing for all of 
them.. Methods that choose the CW don't have that problem, and I've recently 
posted about how well the CW maximizes SU.

So I don't know--Condorcet or CR to best maximize SU, when voters are 
completely sincere?

It's a moot point, since voters are inclined to strategize. In actual 
electorates, I'd expect Condorcet to ordinarily do the best job of 
maximizing SU. But some months ago someone pointed out how sometimes CR 
could do better, even with some people strategizing and some rating in 
proportion to utility.

You continued:

I
disagree with that stance because I believe that it is simply more
appropriate to accord equal power to each voter, rather than allow a
passionate minority power over a less passionate majority.

I reply:

Sure, but what if voters aren't interested in power? What if the majority 
don't want majority rule, but want maximum SU?

You continued:

3) In a two candidate race, if 51% mildly preferred A to B, and 49%
passionately preferred B to A, who should win?

I reply:

With existing electorates, under existing conditions, the majority should 
win, because otherwise there will be strategy problems.

In a different, Utopian, world, that would depend on whether there's a 
majority who don't want majority rule.

You continued:

I  am also curious how "common" multi-member Schwartz (or Smith) sets
are.

In all of the rank-balloting polls I've participated in or conducted, there 
hasn't been one circular tie. I doubt that I've even heard of a rank-vote 
that had a circular tie.

Sure, considering the Schwartz set to be a tie, for poll-reporting purposes 
is perfectly reasonable.

Mike Ossipoff

_________________________________________________________________
Don’t just search. Find. Check out the new MSN Search! 
http://search.msn.click-url.com/go/onm00200636ave/direct/01/