As I said once before, Margins keeps coming back, like some kind of Romero cannibalistic zombie.
I drove the final stake into the Margins monster when I pointed out on EM that, with Margins, sometimes the only outcomes in which a CW is elected at Nash equilibrium are ones in which defensive order-reversal is used. In other words, in some situations, the election of a CW without defensive order-reversal has to be a Nash disequilibrium, an unstable outcome. In contradistinction, with WV Condorcet, and with Approval (and RV), when there's a CW, there is always at least one Nash equilibrium in which the CW wins without any order-reversal. But some Margies are die-hards. Methods that fail in the way described above, I refer to as "falsifying methods. WV Condorcet, Approval, RV, Bucklin, MDDA, MAMPO, and maybe some others are not falsifying methods. Most other methods are falsifying methods. In particiular, Plurality, Instant-Runoff, and Margins Condorcet are falsifying methods. Juho says that Margins does better when voters are sincere. But I've posted, for Juho, examples in which innoncent, nonstrategic truncation can result in a violation of majority rule, and create a defensive strategy problem of a magnitude that doesn't happen in WV. If you want to elect more candidates who have majority opposition against them, than Margins is for you. Juho, read what I posted to you when we last discussed this. Mike Ossipoff ---- Election-Methods mailing list - see http://electorama.com/em for list info
