[EM] RE : Re: Election methods in student government...

[Kevin Venzke]

Hi,

--- Tim Hull <thully at umich.edu> a écrit :
> Also, what preferential methods exist that satisfy "later no harm"?  I
> think
> that students, more than others, tend to "bullet vote" - and this may be
> a
> consideration.  I know range, Borda, et al don't satisfy it, and IRV/STV
> do
> (they fail monotonicity, though).  Is IRV the best there is for
> single-winner than satisfies later-no-harm?  Is STV for multi-winner? 
> Could
> one look for a Condorcet winner (in single-winner) and then fall back on
> IRV
> if none exists without violating LNH?

To the last question: No, no Condorcet-efficient method can satisfy
LNHarm.

Single-winner methods that satisfy LNHarm:
1 FPP
2 IRV
3 Douglas Woodall's Descending Solid Coalitions method. This is monotonic
but in my estimation much worse than IRV in the area of favorite betrayal
incentive. Also, I'm not sure how best to program it on a computer.
4 MinMax(pairwise opposition). I generally like this method's results, 
but it can give some absurd results also.

Here's a useful example:
49 A
24 B
27 C>B

FPP, IRV, and DSC elect A. I don't think this is a good result, but if
you find it acceptable, I'm not sure why you need anything better than
IRV.

MMPO is a tie between B and C. And there is no way to force it to elect
B without violating LNHarm. (You can force it to elect C, if you don't
want to have a tie here.)

Electing C here may not be so bad if you think LNHarm compliance provides
enough incentive, that the voters will be submitting more complete 
rankings.

Kevin Venzke


	

	
		
___________________________________________________________________________ 
Yahoo! Mail réinvente le mail ! Découvrez le nouveau Yahoo! Mail et son interface révolutionnaire.
http://fr.mail.yahoo.com