Hi Paul,
Perot is the clear Condorcet winner, but that cannot be the right result. If you replace those names with A, B, C the result looks ok.
I suspect the issue with your example is that:
45% Bush > Perot 10% Perot 45% Clinton > Perot
is interpreted as:
45% Bush > Perot > Clinton 10% Perot > Bush = Clinton 45% Clinton > Perot > Bush
If people had explicitly marked their ballot as above, would you still consider it "surprising" that Perot won? If so, why? Would any other outcome be less surprising?
The fact that Perot wins in this case is, in part, what I like about Condorcet methods. When the two major factions are evenly split, the Condorcet method will find the compromise option. It would make far less sense to me for either of the larger factions to win in this case when neither has a majority (i.e. 50% + 1) support among the entire electorate.
---- Election-methods mailing list - see http://electorama.com/em for list info
