[EM] Re: Proposal

[Jobst Heitzig]

Dear Rob!

you wrote:
> Well, if all voters use an optimal approval strategy, this is something of
> a knife-edge case.  If the electorate were
> 
> 100:A>D>C>B
> 100:A>B>D>C
> 100:A>C>B>D
> 199:D>C>A>B
> 200:B>D>A>C
> 200:C>B>A>D
> 
> then the approval votes at equilibrium would be
> 
> 100:A>>D>C>B
> 100:A>>B>D>C
> 100:A>>C>B>D
> 199:D>C>A>>B
> 200:B>D>>A>C
> 200:C>B>>A>D
> 
> and A wins.  

I don't understand this. Forest's method would initially give A an
approval of 300, B an approval of 500, and C and D an approval of 499,
hence it would reinforce B as long as B's weight is more than three
times the weight of A, C, and D. From that point on, it would give A an
approval of 300, B an approval of 500, C an approval of 599, and D an
approval of 499, so that it would reinforce C. I don't see at all how A
would arise as "the" winner...