Condorcet "true preferences"
Saari at aol.com
Saari at aol.com
Tue May 6 07:55:02 PDT 1997
>The voters are not being "asked" to do anything -- in Condorcet they are
>being allowed the right to vote all of their true preferences,
Perhaps I am reading into things, but it appears that the writer is claiming
that a Condorcet ballot (A > B > C) allows a given voter to express his/her
"true" feelings. I must disagree. While such a ballot indeed allows for
expression of "true preferences" (e.g. A is better than B, B is better than
C, etc), it still also omits additional relevant data.
For example, consider two different voters voting on such a ballot with 3
candidates.
Voter#1 likes A, also likes B (but some degree less) and thoroughly dislikes
C.
Voter#2 likes A, thoroughly dislikes B, and hates C even more.
It would be ludicrous to claim that these two voters have the same opinion of
the candidates, especially candidate B. Yet each would cast an identical A >
B > C ballot under Condorcet or most other "ranked" voting/scoring schemes.
No possible tallying system can produce an accurate result if the input data
is faulty. I claim that "ranked" ballots (A > B > C) are a faulty method of
voting, as such a vote discards data which is relevant to the final result.
There are a number of plausible methods for choosing a winner - one plausible
choice is "Choose the candidate who is liked by the greatest number of
voters." Such a method cannot possibly work, however, with a ranked A > B >
C ballot, since such a ballot effectively hides the relevant data (Does the
voter like or dislike candidate B?).
Mike Saari
More information about the Election-Methods
mailing list