Example with contrary half preference votes

[Lucien Saumur]

In an article, seppley at alumni.caltech.edu ("Steve Eppley") writes:

>Lucien S wrote:
>>     I am trying to understand the "least-beaten" concept and I do
>>not understand why Clinton would win rather than Dole, with or
>>without half preferences. 
>-snip-
>
>It looks in your matrices like you're measuring the size of a
>pair-defeat by the margin of defeat (votes for pair-winner minus
>votes for pair-loser).
>
>But in Mike O's definition of Condorcet, the size of a pair-defeat
>is measured simply by the votes for the pair-winner (i.e., the number 
>of voters who prefer the pair-winner over the pai-loser).  
>
>This amount is the number of voters who would be unhappy if the
>pair-loser beat the pair-winner in a 2-candidate election; the
>Condorcet method tries to minimize the number of unhappy voters. 

          I am not clear about the difference between what
I have described and what is being proposed. Could you give
me an example where the two methods would produce different
results?

__________________________________________
          aa447 at FreeNet.Carleton.CA
          http://www.igs.net/~lsaumur/