[EM] 04/05/02 - I Know the Drill, Adam, I'll be on the same page:

[Adam Tarr]

>Winning this election is only our second objective.  Worst may win
>this election, but that is only a short term gain for them.  We are going
>for the long term gain.  We are going for Number One, but first we need to
>be Number Two.

That's basically a load of bull.  What you are saying is that IRV does not 
prevent a two-party duopoly, but it does make it marginally easier for one party 
to shove another out.  That's not too much better.  Moreover, this whole idea 
that we gain some moral victory by polling well, and losing to the extreme 
candidate on the other side, is pretty bogus.  Gains in elections are made by 
winning elections, and everything else finished far, far behind.

Here's another thought: what if the new third party was the Compromise party?  
IRV now keeps them down indefintely, even though more than half the populace 
would prefer Compromise winning the election to either "Favorite" or "Worst".  
Alex Small implies this was the case in the 1980 presidential election.

>(Note: What Adam does not tell us is that candidate Worst will still win if
>the method being used were Approval Voting.)

No, Worst would not usually win.  Check out the Reagan, Anderson, Carter 
examples I cite in the thread "Equilibrium in Approval Voting".  This is almost 
exactly equivalent to my Favorite, Compromise, Worst scenario if you cast those 
positions as Carter, Anderson, and Reagan respectively (no offhand 
editorializing intended here).  As I point out, the only way that nobody regrets 
the way they voted is if the approval vote results in Anderson (Compromise) 
winning.

-Adam