[EM] Wikipedia

[Adam Tarr]

James Gilmour wrote:

>Are you, Adam, suggesting that Condorcet is not a "one person, one vote" 
>system?  No voter can give more than one vote to any one candidate.  When 
>it comes to any decision, ie pair-wise comparison, each voter's vote 
>counts for only one candidate.  That would seem to satisfy any reasonable 
>definition of "one person, one vote".

I would agree, James.  The only commonly discussed method that fails 1P-1V 
in my opinion is the Borda count.  But I don't consider 1P-1V to be a 
particularly meaningful criteria anyway (or even a well-defined one).

>Or did you have in mind the situation when a Condorcet cycle exists, so 
>that the vote of a voter who has marked more than one preference counts 
>for more than one candidate at the same time?  In that particular 
>situation one might also argue that the voter's second preference was 
>counting against the voter's first preference.

This is an interesting question, but I don't see any meaningful way to 
apply the principle of 1P-1V to the consideration of multiple pairwise 
contests.

-Adam