MIKE OSSIPOFF wrote: > To avoid having to compare the probabilities of making X or Y the > unique winner, it probably _is_ better to speak of only counting > John's ballot, if that would avoid the problem. So that's my > official definition now. But there's still that other problem mentioned > above, which I don't yet know what to do about.
Counting just the one ballot takes care of both problems, unless we want to talk about a method, such as you suggested, requiring a margin of victory greater than one. You could extend the definition to say that you count that one ballot the minimum number of times needed to get a decisive win. However, the wording may start to get unwieldy. -- Richard
