Ordeshook's strategy concerns

[Mike Ossipoff]

Ordeshook's comment seems to be one of those that often come from
the academics, when they complain about the possibility of strategy
in Pairwise methods, without actually studying the strategy comparison
of the best Pairwise methods with other methods.

Samuel Merrill, it seems to me, in his book _Making Multicandidate
Elecitons More Democratic_, voiced a similar concern about
Pairwise method & strategy, again without any effort to 
actually compare their strategy situation with other methods.

Objections like those of Ordeshook & Merrill may be valid 
for the worse Pairwise methods. They may be valid for Copeland.
But they aren't valid for Condorcet, which has less need for
defensive strategy than any other 1-balloting method I've
heard proposed. (Unless some close relatives like Cheatproof
, a close relative of Condorcet, could challenge it--I have to
say that to cover my bets).

We have some academics saying Copeland is the best Pairwise
method. We have other academics saying that Pairwise methods
have a strategy problem and aren't really significantly better
than, say, Approval. They can point to Copeland, the academics'
favorite Pairwise method to back up their claim. If weren't
so un-suspicious :-) I'd suggest that these folks were an efficient
team.


Mike


--