[EM] margin vs winning votes

[Kristofer Munsterhjelm]

On 07/23/2014 09:00 PM, C.Benham wrote:
>
>> voting your favorite sincerely as first choice instead of not voting
>> cannot make him lose while winning votes is the criteria.
>> voting your favorite sincerely as first choice instead of not voting
>> can make him lose while margin is the criteria.
>
> Stephane,
> I'd be interested in seeing that, because I think it's been shown that
> Mono-add-Top (the property you refer to) is incompatible
> with Smith, which is met by Winning Votes using Schulze, Ranked Pairs or
> River.

It's still open as to whether one can make a method that satisfies both 
Mono-add-top and Smith. What Woodall shows is merely that no such method 
can also pass Plurality.

Finding a method that's both Smith and MAT compliant would be very 
interesting. Venzke doesn't think it can be done, if I recall correctly. 
In practical matters, Smith,Minmax does get quite close, but in the 
domain of criterion compliances, close just isn't good enough.