Lorrie Cranor's paper (Re: Coombs' Method)

[Mike Ossipoff]

The Condorcet "bad-example" from the web page evaluates the 
deservingness of the alternatives according to how many people have
ranked them in 2nd & 3rd place. It scores them according to 
rank position, like Borda's point system. Fine, then that author
would prefer Borda. But it should come as no surprise to anyone
that Condorcet's method will sometimes get a different result
than the result that Borda gets--because it isn't Borda. The question
then is: which do we want, Borda or Condorcet, based on what they
do. WE could compare them according to various things. We could
compare them according to which one picks an alternative with highest
Borda score, though the unfairness of that standard for that comprison
is obvious.

Borda is notorious for its violation of majority rule. It's worse
than other methods in that respect. It's the only method that I
know of that can fail to elect an alternative which a majority of
all the voters have ranked alone in 1st place. Borda, like MPV,
ratains the LO2E problem that Plurality has. Therefore, it means
little when that author criticizes Condorcet for not being Borda,
for not scoring based on how many times each alternative is ranked
at each rank position. 

I just wanted to mention this, to reply to that example, even
though what I'm saying isn't really news.

Mike Ossipoff






--