[EM] PR Condorcet algorithm implemented as experiment

[Adam H Tarr]

>I implemented the PR-enforcing Condorcet algorithm I described in my recent
>mail to this list, as part of the CIVS voting web service.  If you would like
>to try it out (and give me some testing!), visit the following URL and vote on
>the "ice cream assortment" election:
>
>http://www5.cs.cornell.edu/~andru/civs/public_elections.html
>
>You can of course set up your own elections using this system too.

It would be interesting to compare these results to the Condorcet-PR link I 
provided earlier:

http://groups.yahoo.com/group/election-methods-list/message/10308

It would also be nice to see how each of these methods handles borderline cases, 
or cases when the factions are less clear-cut.  Who gets superproportional 
representation, does anyone get under-represented, et cetera.

Unfortunately, Condorcet-PR more or less passes over the boundary into "too 
complicated to work out in most meaningful examples", so a computer 
implementation is basically the only way to compare things.

-Adam