[EM] Single Contest Method (aka Maori?)

[Kevin Venzke]

Hi Forest,

We can't call SC "Maori" because the acronym isn't descriptive of the
method.

--- En date de : Jeu 21.7.11, fsimmons at pcc.edu <fsimmons at pcc.edu> a écrit :
> Too bad we cannot discard even more information, like the
> fact of a majority winner that isn't one of the 
> contestants in the contest picked by the approval votes.

Yes, that's disappointing. But apparently it's not just politically
unacceptable to ignore a majority favorite. It appears to wreck the
results, too.

> If it weren't impractical, the ideal mode for this election
> would be an approval vote for the explicit purpose 
> of determining who would be in the runoff, and a later trip
> to the polls with only those two names on the 
> ballot.

No, that wouldn't work. You would not be compelled to vote the runoff
consistently with your approval vote. SC works because if you want to
defeat a particular contest (because your candidate will lose it) via
insincerity, you will have to do this by disapproving your own 
candidate, which is not a promising strategy. It might work in rare
circumstances, but it's not like there is one thing that the entire
faction can agree to do.

> Another thought: make the single contest between the
> approval winner A and the candidate B with the 
> most approval on the ballots that did not approve A. 
> If B's total approval is less than the number of 
> ballots on which A is approved without B, then no runoff is
> needed.

Actually, I mentioned that very method in my post.

Kevin Venzke