Late-Nite Error

[Mike Ossipoff]

Last night I said that, in my standard 40,25,35 example, if the
Dole voters order-reverse, that returns a tie between Nader &
Clinton, when the Simpson-Kramer method is used. Actually, I was
mistaken when I said that; it gives the election to Nader, as does
Condorcet. I don't know how I made that error, except that it was
quite late at night.

But I was right about the truncation example. And it isn't just
that Simpson-Kramer would return a tie, so we'd rely on a tie-breaker
(though it might often in small elections). In public elections
a tie consisting of 2 alternatives with equal Condorcet scores is,
of course unlikely, and that goes for Simpson-Kramer too. But
if there could be some truncation among any or all of the voter
groups, as seems likely in a real election, then the tie that
I showed in the Simpson-Kramer truncation example could go either
way.


Mike


--