In the poll-topics poll, 2 pairwise-count methods were designated: Cloneproof SSD, and Ranked Pairs (Tideman's margins version). I typed Cloneproof SSD into the computer today. I would have used my new computer, but it's fitted only with a new programming language. I wanted to get the poll result right away, and so I didn't want to choose this time to start using a new programming language. So I used my old computer, which is set up to use a language that I'm familiar with. When I used the old computer, I hoped that it wouldn't fail, as it sometimes does, in a way that loses the program that's been typed in. It doesn't do it often, but its occasional failures were one reason why I got a new computer. Well, the old comuter _did_ fail in that way today, and the time that it chose for failing was after I'd typed the fairly long Cloneproof SSD program into the computer, but was still fixing a few syntax typos in the program as entered. Of course the program is longer because it was necessary to deal with situations where there are 2 equal defeats. So I lost a lot of time. I won't waste any more time with the old computer. Maybe it's time to throw it out. So tomorrow I'll do the Cloneproof SSD and Ranked Pairs(margins) counts on the new computer. Of course if someone else wants to count the ballots, especially the Cloneproof SSD and Ranked Pairs(margins) counts, I wouldn't be offended. But I emphasize that that isn't necessary, since I'll be doing the count tomorrow on my new computer. If someone else does a count, they'll need Martin's ballot, which was sent to me directly. Martin mentioned the possibility of offensive strategy based on previously-posted ballots. With Condorcet versions, including the 2 that were designated, offensive order-reversal can be thwarted by the counterstrategy of defensive truncation. In the subsequent poll, I'd like to have a 2-day period, after the balloting deadline, during which anyone can truncate the ballots that they've voted. In that way they can thwart offensive order-reversal if they detect it. And it might be detectable, since we have a fair idea of list-members' preferences among voting systems. But really, mostly we're on an honor-system, that no one is going to use offensive order-reversal based on observing previously posted ballots. Sure, one solution would be to have all the ballots e-mailed directly to me rather than posted. But, for one thing, there's probably a (mis)perception that such a procedure would be fraud-prone. Actually, since I'd be expected to post the ballots, and whom they're from, when reporting the results, there's no way I could modify, add, or leave out a ballot without someone noticing that. Still, if it's felt by some that verification isn't as good as with posted ballots, then posted ballots should be used. Aside from that, I just prefer posted ballots. In the poll, Voting Systems does so well in the Approval and CR counts that it seems likely to win in the Cloneproof SSD and Ranked Pairs(margins) counts too, even though, when all the ballots are counted, Multiwinner Methods pair-ties Voting Systems. (Of course, not having a computer result yet, I haven't counted the other 154 pairwise vote totals). If Voting Systems wins by all 4 of the designated methods, then it will win the Voter's Choice count. If it loses the Cloneproof SSD and Ranked Pairs(margins) counts to Multiwinner Methods, it could still win, because 4 of the 7 voters designated Approval or CR. Sorry about the delay. That's the last time I use that old computer. Mike Ossipoff _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com
