[EM] Covering vs wv. wv vs margins. RP vs River.

[Kevin Venzke]

Hi Mike,



----- Mail original -----
>De : Michael Ossipoff <email9648742 at gmail.com>
>
>wv vs margins:
>--------------------
>
>Juho pointed out that margins Condordet (could be RP or River) is
>somewhat more resistant to chicken-dilemma defection, than is wv.
>

As would be Smith//Random Ballot. Really WV methods could not be worse for your
purpose, short of failing MMC.

>
>But, just out of curiosity, does Margins likewise offer more
>resistance (in comparison to wv) to a failure of FBC?
>

Experimentally, definitely not. WV tests as almost perfect with regard to FBC.
My AI sim shows a lot of compression incentive under WV, which under margins is
replaced by defensive order reversal.

In the past I believe you used to point out scenarios where equal-top ranking 
was an effective defense under WV, where margins in the same situation required
reversal.

As a general rule of thumb I would speculate that all methods that violate
SDSC 1. are more resistant to chicken-dilemma defection (or alternatively: have
greater nomination disincentive) and 2. require favorite betrayal more often.

Kevin