[EM] Strategy-free Condorcet resolution methods

[Forest Simmons]

Is there some non-deterministic resolution of Condorcet that would
completely discourage insincere ranking?

Neither random ballot nor random candidate does the job, but how about
some probabilities based on how close a candidate is to being a CW, or
something like that?

If we found a non-deterministic resolution that worked well, then we could
simulate it with deterministic chaos, so that the results would be
repeatable (i.e. checkable).

Forest

On Thu, 20 Mar 2003, [iso-8859-1] Kevin Venzke wrote:

> Could I get some opinions on what resolution methods
> are most free of incentive to rank insincerely, or run
> clones, even if the results are "inferior"?
>
> Smith//RandomBallot seems to encourage (however
> slightly) up-ranking compromise candidates.
>
> Smith//RandomCandidate isn't clone-proof.
>
> Smith//MMPO is something I want to think about.  I was
> also thinking about a kind of Smith//Bucklin, but that
> would suggest up-ranking compromises, too.  The
> results might be decent, though.
>
> What if the winner of the strongest (WV) defeat
> against a Smith member were made the winner?
>
> This example was discussed awhile ago:
> 49 Bush
> 24 Gore
> 27 Nader>Gore
>
> propositions:
> 51-49 Gore>Bush
> 49-27 Bush>Nader
> 27-24 Nader>Gore
>
> RP(margins) elects Bush; RP(wv) elects Gore.  Are
> there other (simpler) methods which necessarily elect
> Gore?  "Most Votes In Any Contest" elects Gore.  I
> believe "Plain Condorcet" elects Gore.
>
> Kevin Venzke
> stepjak at yahoo.fr
>
> ___________________________________________________________
> Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en français !
> Yahoo! Mail : http://fr.mail.yahoo.com
> _______________________________________________
> Election-methods mailing list
> Election-methods at lists.electorama.com
> http://lists.electorama.com/listinfo.cgi/election-methods-electorama.com
>