Anthony Duff wrote: > > I am interested in the question of the frequency of non-existance of > a sincere CW. I personally do not know that it is probable.
In Merrill, "Making Multicandidate...", in the table on p.24, he shows frequency of sincere CW for 5 candidates under a random society simulation, and for some spatial models using either 2 or 4 dimensions. The spatial models had a CW around 98 or 99 percent of the time, and with the 5-candidate random society model there was a CW 76% of the time. Merrill apparently believes that a realistic scenario would fall somewhere between the random society and spatial models. > I have been under the impression, that it is actually improbable, and > that when it would occur, the cycle of defeats would probably be > quite weak. Further, it is my impression that the root of strategy > problem of condorcet is that while a sincere CW may exist, a losing > faction can, with straightforward strategy, vote strategically to > defeat the sincere CW. That might be the greater concern-- would cycle-creation be an important strategy? If so, then actual frequency of cycles could swamp the sincere figures. Bart ---- Election-methods mailing list - see http://electorama.com/em for list info