[EM] Comparing CW with "write-in pairwise" CW

[Kristofer Munsterhjelm]

On 10/31/2013 12:20 AM, Kevin Venzke wrote:

> Technically MinMax doesn't satisfy Smith, but violations would be
> extremely rare. If you use margins with Schulze, yes, you'd get the
> same results as RP(margins) given three Smith candidates. Same with
> River(margins) if there is such a thing.

For the people who value 100% criteria compliances, Minmax(margins) also 
has the benefit of passing mono-add-top. And Smith,Minmax only very 
rarely fails it.

It's still an open question whether or not you can make a method that 
passes both Smith and mono-add-top. I lean towards yes; if I remember 
correctly, Kevin leans towards no.