[EM] Proper use of better-than-expectation Approval strategy

[MIKE OSSIPOFF]

Say there's a progressive who lacks honesty on an important litmus-test 
issue, but is still in most respects a genuine progressive.

And say you have to decide whether to vote for him/her in Approval.

Well, because s/she's a genuine progressive for the most part, and we know 
what's likely to happen in the election, you could say that questionably 
honest progressive is better than your expectation.

In a particular actual case, for someone who was a candidate in our EM 
presidential poll, that didn't seem right.

And here's why:

The only time it would matter whether or not we voted for that candidate 
would be if s/he were one of 2 candidates between whoml your ballot could 
make or break a tie. But if s/he accomplishes that, then that says something 
about the electorate. That's information that changes your expectation in 
the election. Not only is Nader better than that questionably honest 
progressive, but Nader is much more famous and popular. If that candidate 
can be in the tie, isn't Nader, or at least some other progressisve, likely 
to be the other candidate in the tie? Isn't your expectation much greater 
when we assume that that progressive is one of 2 candidates between whom 
your ballot can make or break a tie?

So, that candidate is no longer better than your expectation, and the 
better-than-expectation strategy no longer tells you to vote for him. That's 
better.

How the Pij are affected by the assumption that that candidate will be one 
of 2 candidates between whom your ballot can make or break a tie is 
something to take into account with other Approval strategies too, in 
deciding whether or not to vote for that candidate in Approval.

Mike Ossipoff

_________________________________________________________________
Express yourself with the new version of MSN Messenger! Download today - 
it's FREE! http://messenger.msn.com/go/onm00200471ave/direct/01/