[EM] Approval-Completed Condorcet Redux

[Adam Tarr]

I've been off the list for a while, and just returned.  I noticed that in a 
recent message, Alex wrote, "...a while ago somebody (Adam?) pointed out 
that standard Approval-completed Condorcet can give strong incentives for 
insincere voting."  This is correct.  For more details, you can look at the 
original message in the archives:

http://groups.yahoo.com/group/election-methods-list/message/9456

I blather on about graded ballots in the early part of the message, but 
later on I get to a concrete example that shows some of the weaknesses of 
approval-completed Condorcet.  I'm not sure exactly how applicable this is 
to the current "Condorcet-enhanced approval", though.

Problems like this are what convinced me that straight winning-votes 
Condorcet (using either ranked pairs or beatpath) was the most 
strategy-free system that gives good results.  Nothing I read in the 
"turkey" threads really convinces me otherwise.  The idea of using graded 
ballots, and attaching some significance to a low "GPA" in the winner (a 
dramatically reduced term length, for instance) seems reasonable to 
me.  But my experience shows me that attempting to add any sort of approval 
cutoff in the actual machinery of the election can do more harm than good, 
due to unintended strategic consequences.

-Adam