Earlier, you proposed a criterion which said that B should win in the following situation:
49 C 27 B>A 24 A (>B truncated) I criticized that criterion by saying that B is not the right result if the situation is in fact: 49 C 27 B>A (strategic; honest is B>C) 24 A (>B=C sincere) You replied that my criticism did not apply to your criterion, even though the votes are the same. Now you say: That's because you have no way of determinng someone's > preferences independent of their expression of them. which seems to me to be the same thing I've been arguing: that a criterion can only require certain results for certain (kinds of sets of) ballots, even if those requirements are based on some reasoning about possible underlying preferences. Jameson
---- Election-Methods mailing list - see http://electorama.com/em for list info
