[Election-Methods] RE : Re: Borda-elimination, a Condorcet method for public elections?

[Kevin Venzke]

Ian,

I think a big part of the reason that Condorcet methods aren't adopted is
that from the perspective of a party in power, it's clear how such methods
could undermine their ability to stay in power. IRV has some ability to
disregard weak candidates, that's all; if a party in power is sometimes
stung by third party candidates then IRV could be appealing here. But IRV
will still basically stick to electing one of the FPP contenders.

If people want to use a Condorcet method but they don't want to use a
method so complicated as RP, Schulze, or Jobst's River method, then I
suggest one of these:

1. Elect the Smith set member whose greatest loss is the weakest (according
to winning votes)
2. If there is no Condorcet winner, elect the candidate with the most
rankings (Condorcet//Approval with implicit approval)
3. Elect the Smith set member with the most rankings

I also feel it is ok to "fudge" Condorcet slightly to gain other criteria.
In particular my "Improved Condorcet//Approval" method technically fails
Condorcet but picks up compliance with the favorite betrayal criterion.

I point this out partly because while I like method #2, my simulations
suggested rather frequent occurrence of favorite betrayal incentive under
it...

I could not support a Borda-based method because Borda's mechanism strikes
me as unreasonable. It doesn't make sense for scoring to be dependent upon
the number of candidates representing some policy.

Kevin Venzke


      _____________________________________________________________________________ 
Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! Mail http://mail.yahoo.fr