Craig Carey's recent ROXY example was just another case of someone trying to find problems with Approval and coming up short. So you have 10 "R" candidates and 20 voters who find them similar enough, and good enough, to approve of all of them. And you have 10 "O" candidates. They are either so dissimilar in spite of having the same first initial that the voters for a single "O" candidate cannot find another "O" candidate that they like, or the "O" voters are so strategically impaired that they do not know how to vote in an Approval election. And you have a bunch of "X" and "Y" candidates that practically nobody cares about. So it's little surprise that the winner will come from the "R" block. Even if there were only one "R" candidate, it seems unlikely that the winner will come from anywhere else. Perhaps if nine of the "O" candidates dropped out then their supporters would redistribute their votes in a way that could affect the outcome. As it is, it seems the "O" voters aren't very bright in terms of strategy. Maybe Craig is looking for a method that doesn't punish bad strategy. I don't think such a method exists. But one of Approval's strengths is that nobody has to drop out in order to give the voters a fairer choice. Remember the people suggesting Nader should drop out so that Gore would have a better shot? That's not a problem in Approval. Assuming of course that the voters are smarter than Carey or Davison give them credit for. Oh, those "O" voters! Richard
