Eric,
--- Eric Gorr <[EMAIL PROTECTED]> a écrit :At 5:01 AM +0100 1/24/04, Kevin Venzke wrote:> >40 A>B>C> >>35 B>C>A >25 C>A>B>A wins. The 35 votes are not counted.
Actually, the 35 votes matter a great deal in an RP election - assuming you are referring to '35 B>C>A'
These 35 votes are a part of the pairwise defeat of B>C being the strongest (has a strength of 75), which in turn eventually causes the pairwise defeat of C>A (with a strength of 60) to be rejected because of the potential creation of a cycle (i.e. ambiguous result).
That's true, but it does the BCA voters no good to lock in B>C if the effective
result is that C>A has to be thrown out. They'll wish they had voted B=C>A or
C>B>A.
So I suggest that, arguably, those votes aren't "counted." They don't do what they were probably intended to.
If the rankings weren't sincere, well, anything concerning the outcome of the election could happen.
The fact remains, these 35 votes, as a whole, changed the outcome of the election. As you pointed out, had these 35 rankings been different, the outcome would have been different.
---- Election-methods mailing list - see http://electorama.com/em for list info
