[EM] Re: approval strategy (Russ Paielli)

[Alex Small]

Russ-
 
I would think that the strategy is straight-forward:  If the Green is one of the top 2 contenders in pre-election polls then the Green's supporters should vote for the Green and anybody they prefer to the Green.  Of course, strategic adjustments can then change the dynamics of the race, and there's no guarantee of finding an equilibrium.  However, one can show that in races with 3 major candidates the only way to not have an equilibrium (under fairly reasonable assumptions) is when there's a Condorcet cycle, AND EVEN THEN an equilibrium is still possible (but not guaranteed).
 
A method has been proposed to partially alleviate the complicated decision-making, and it's not much more difficult to count than approval voting.  Forest Simmons introduced it on this list, and it's called "Majority Choice Approval":
 
-Voters rate each candidate as Preferred, Acceptable, or Unacceptable.
-If a single candidate is Preferred by a majority of the voters he wins.  If multiple candidates are Preferred by a majority (voters have the option to Prefer more than 1 candidate if they wish to do so) then the candidate Preferred by the greatest number of people wins.
-If no candidate is Preferred by a majority then the candidate with the greatest combined number of Preferred and Acceptable votes wins (or, equivalently, the candidate rated Unacceptable by the fewest people).
 
It doesn't completely eliminate strategy issues, because if the voters know (from polls) that there will be no majority and the race will be decided by also counting the "Acceptable" votes, the Greens still have a dilemma with regard to the Democrat.  However, it at least partially alleviates the problem.
 
Also, I admit some fondness for the {-1,0,1} implementation of CR.  It's easy to implement:  Vote yes, vote no, or abstain for each candidate.  Yes votes count as +1, no votes count as -1, and abstention counts as zero.  Most points wins.  I realize that strategically it may be equivalent to Approval, but for voters who are unsure about strategy it gives them a compromise option.  Not perfect, but it's a way out for those who want it.  I'm not interested in trying to save voters for themselves.  If the availability of a sub-optimal strategy makes voters happy, so be it.
 
Alex Small

election-methods-electorama.com-request at electorama.com wrote:
 
		
---------------------------------
Do you Yahoo!?
 Yahoo! Mail - You care about security. So do we.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20050117/f8560d1e/attachment-0002.htm>