[EM] STV and weighted positional methods

[Terry Bouricius]

What is even more puzzling is Ms. Dopp's continued defense of plurality 
voting. She attacks STV/IRV because a candidate with broad second rankings 
may not win, while under plurality rules, these lower preferences are not 
considered at all. The immediate discussion was about the majority 
favorite criterion (a candidate that is most favored by a majority should 
win -- which IRV and Condorcet meet), but plurality and many other methods 
fail. But an even more troubling failure is the Condorcet-Loser criterion 
(the election of the one candidate that would lose in every one-on-one 
match-up). Again IRV and Condorcet meet this most basic criterion, while 
plurality elections fail it -- not only in theory -- but with disturbing 
frequency in real-world elections.

Terry Bouricius


----- Original Message ----- 
From: "Kristofer Munsterhjelm" <km-elmet at broadpark.no>
To: <kathy.dopp at gmail.com>
Cc: <election-methods at lists.electorama.com>; <jgilmour at globalnet.co.uk>
Sent: Thursday, January 29, 2009 2:33 PM
Subject: Re: [EM] STV and weighted positional methods


Kathy Dopp wrote:
> I disagree with your characterization that in your example C does not
> have majority support because all 100 voters prefer C over at least
> two other candidates,
>
> However, I see your point that the numerical total value of votes
> received by C, is not a majority out of the total numerical value of
> all votes counted.

The point is rather that while a majority prefers C to at least two
other candidates, a majority also prefers A to all the others. In other
words, if the voters were asked "Do you prefer A to all the others?", a
majority would answer yes, while if the voters were asked "Do you prefer
B to all the others?", or "Do you prefer C to all the others?", that
wouldn't be the case.

It seems suspect for a method to fail to elect a candidate when a
majority prefers that candidate to all other candidates.
----
Election-Methods mailing list - see http://electorama.com/em for list info