[EM] Schulze Method (Markus Schulze)

[Kristofer Munsterhjelm]

Kathy Dopp wrote:
> Markus,
> 
> Unfortunately I don't have time to study it now, but a quick perusal
> makes it seem written in a clear, easily-understood style of writing.
> 
> Am I to assume that this method solves the problem of irrelevant
> alternatives (the spoiler problem) in all cases unlike both plurality
> and IRV?
> 
> I do not know what "clone-independent" means (perhaps I should by now).

No rank ballot system can meet IIA while electing an unanimous preferred 
candidate and having no dictator. (That doesn't mean Schulze is as bad 
as IRV, though, only that "perfection" cannot be attained.)

Clone-independence means that if one adds or removes candidates from a 
set that all voters vote next to each other (but not necessarily in the 
same order), that shouldn't cause someone in the set to lose if someone 
in the set won before the modification, and nor should it cause someone 
in the set to win if a candidate outside the set won before the 
modification. Intuitively, it means that a party shouldn't lose (or win) 
merely by splitting.