Consider the following case: 6 EABCD 1 ACBDE 2 BACDE 2 CABDE 6 DABCE In IRV, after 3 rounds, you end up with 6 ED 1 DE 2 DE 2 DE 6 DE so D beats E in the 4th round, 11 to 6. The IRV proponents like to call this a majority victory for D. But what do the majority of voters think of D? You have 6 voters who ranked D last, and 5 who ranked D next to last. So 11 of 17 voters evidently think D is a poor candidate. Only 6 of 17 ranked D among their top choices. That's how well IRV serves majority rule. The problem with the IRV proponents' claim for majority rule is that they have assembled their majority only by eliminating many of the preferences of that majority. IRV would elect anchovies as a favorite food because a majority of people like anchovies better than liver. (OK, some of you out there may like one or both of those foods, but you get the idea.) A population can be divided into a majority section and a minority section many different ways. Why is the particular majority selected by IRV entitled to be the majority that rules, when there are so many other possible majorities? I acknowledge it's possible that the 11 voters who preferred some permutation of ABCD over E may actually have very close ratings of those four. They could all despise E but just be unable to agree on whom they want instead. But the IRV proponents should similarly acknowledge the opposite possibility, that 5 of the 11 dislike D almost as much as they dislike E. IRV simply doesn't measure the strength of preferences. And D and E could both be extremists. How can the centrist voters in this scenario change their votes to ensure a centrist candidate wins? Even the voters who favor D or E would probably rather not see this race come down to a contest between D and E, because they know that if their choice loses, they will have to endure a win by the most evil candidate on the ballot. I know I would rather think we were choosing between two of the best, rather than the leftovers of some pseudo- random elimination process. -- Richard
