Juho Laatu wrote:
Another approach to offering more
flexibility (maybe not needed) and
more strategy options (maybe not
wanted) is to allow the voter to
fill the pairwise matrix entries
in whatever way. This means that
also cycles can be recorded.
One can interpret the basic
Condorcet rules so that they do
not rule out this option. (The
ballot format is not defined.)
(Are there good examples where
these more flexible approaches
would provide some definite
improvements?)
I think that's too flexible. Allowing a single voter to give multiple
votes (but at fractional power) can be justified by that the voter is
judging the candidates on seperate metrics. The sum matrix is still a
proper tournament matrix. However, letting the voter arrange his
contribution to the Condorcet matrix as he wishes may let him move the
matrix out of what could be reached by ordinary votes, which seems
nonsensical.
If one desires such flexibility, it should at least be phrased in terms
of contests. That is, setting M(A,B) to q means you prefer A to B by
fraction q, or that A won in a "match" of some sort when facing B (such
as with round-robin tournament matrices in sports).
----
Election-Methods mailing list - see http://electorama.com/em for list info
