[EM] DAMC meets reasonable FBC

Michael Ossipoff mikeo2106 at msn.com
Wed Mar 14 03:02:14 PDT 2007


The FBC that I’ve been using is the one that requires that there be no way 
of improving one’s outcome by voting someone over one’s favorite. That 
strictest of FBCs is met by Approval, -1,0,1, MDDA, and MAMPO.

It appears that DAMC doesn’t meet that FBC. But to know that voting some 
particular candidate over your favorite will nearly complete a cycle that 
would then be completed by a defeat of your compromise, thereby preventing 
that defeat of your compromise from being kept--that would require pretty 
much complete information about how the count will turn out. After all, 
without that, voting some candidate over your favorite, if doing so will 
partially complete a cycle, so as to keep some other defeat from being 
kept--that could just as well hurt your compromise as help it. It would be 
pure guessing. And, without knowing if that will help or hurt your outcome, 
what you’d be doing to your favorite tips the expectation-balance of that 
favorite burial to the negative side.

So, if we judge by an FBC that requires that a voter won’t improve his/her 
expectation by voting someone over his/her favorite, without improbably 
complete information, then DAMC passes that FBC.

The pair-wise count methods that could have even any FBC problem are methods 
in which voting Compromise over Favorite can prevent a defeat of Compromise. 
It doesn’t take improbable predictive knowledge to know that doing that can 
be expected to help Compromise, much more likely than hurting Compromise. 
The usual Condorcet versions fail that FBC, and so the super-timid voter 
isn’t completely irrational when s/he votes Compromise over Favorite. S/he 
is actually helping Compromise’s chances some. Personally I wouldn’t 
consider that tiny help for Compromise a justification to abandon Favorite. 
I hope that with Condorcet people will be motivated more by hope than by 
fear, and will rank sincerely, because of SFC’s guarantee, even though 
Condorcet almost surely fails even that more reasonable FBC.

Now, if DAMC meets GSFC, then it has an advantage that I value, over MDDA & 
MAMPO--but at great cost in complexity. And it's not dominated by wv 
Condorcet, because it trades Condorcet's Criterion for expectation FBC.

But one thing for sure: DAMC doesn't rival MDDA or MAMPO in their role, 
because extreme simplicity and definition-brevity is part of their 
advantage. As I said, DAMC is in the Condorcet complexity category.

Mike Ossipoff





More information about the Election-Methods mailing list