[EM] Scenario where IRV and Asset outperform Condorcet, Range, Bucklin, Approval.

[Jameson Quinn]

2010/5/12 Kristofer Munsterhjelm <km-elmet at broadpark.no>

> Jameson Quinn wrote:
>
>  Condorcet winner is M. But if all the RC>M voters truncate before M, then
>> M does not beat R and L, so there's two cycles M>RC>R,L>M. Most Condorcet
>> tiebreakers, including Schulze and Minimax, would name RC as the winner. (Of
>> course, if the M voters retaliate in kind, then R or L would win Condorcet,
>> or M would win Approval, Range, or Bucklin.)
>>
> >
>
>> There are two ways I can see to avoid this dilemma. One is LNH-type
>> elimination, such as in IRV or certain versions of Asset. Since RC is
>> eliminated before M, no lower votes of the RC voters can help RC win.
>>
>
> Would DH3-resistant Condorcet methods like Smith,IRV or BTR-IRV also resist
> the truncation strategy?
>

Smith-IRV would resist the "radical center" strategy. I'm not familiar with
BTR-IRV but it very well may do so too.

Again, note that I am not saying that resistance to the "radical center"
scenario is the be-all-and-end-all. Despite the scenario, I believe that
Schulze, Minimax, and Bucklin are all far superior to IRV or Smith-IRV.
(Asset is a separate issue; overall, I support it, but it's a more radical
change in most contexts and so a more far-off goal). But this scenario is
real and plausible. All it takes is 2 issue dimensions and something like a
rhombus of candidates. In fact, it is one of the 2 most plausible strategic
scenarios I know of under Condorcet systems (the other one is where Bush
voters vote Bush-Nader-Gore and win if there are enough Gore-Bush-Nader
voters; which takes only 1 dimension and 3 candidates, but, unlike radical
center, assumes a non-normal distribution of voters). So it is worth
pointing out. I personally can't think of any ways to avoid it, besides
asset, that don't make the system worse overall; but there are a lot of
smart people on these lists, and somebody might have a good response.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20100512/1c3e9867/attachment-0004.htm>