Merit ranking of methods, for the Green scenario: 1. Woodall 2. Benham 3. AIRV (defined below) 4. IRV 5. Beatpath, RP, Approval, Score
AIRV (Approval-IRV): Same as IRV, except allows equal ranking (at least for 1st place), and all the candidates currently sharing top position in a ranking are regarded as topping that ranking. [end of AIRV definition] As for why Woodall and Benham are ranked highest, and why Woodall is above Benham, it's for reasons stated in my previous post, and others before that. AIRV is the next best thing. Though IRV's failure of the Condorcet Criterion (CC) can make a strategy situation, sincere voting can still be justified. The chicken dilemma, to which Beatpath & RP are subject, can't be ignored. When it exists, strategy is needed. The chicken dilemma is worse than IRV's CC failure. Hence Beatpath's and RP's rank-postion below that of IRV. Approval, Score, Beatpath & RP are ranked equally at the bottom of that list of methods (but above those not ranked), because FBC, though not needed in the Green scenario, never really stops being somewhat desirable. If there were still a favorite-burial incentive for Beatpath, then Approval would be better than Beatpath. If there were no chicken dilemma, Beatpath & RP would probably be the very best. ...but chicken dilemma will be common in public political elections. I once wrote to some Beatpath-using organizations, offering Woodall and Benham (but especially Schwartz-Woodall), because of freedom from chicken dilemma. I was told that their organizations didn't have chicken dilemma. That's good. If they don't have the chicken dilemma, then Beatpath/CSSD is probably the best choice for those organizations. That gives me reason to be glad that I introduced wv Condorcet, pointed out some advantages of it, and offered Beatpath/CSSD to organizations. Comparing Beatpath, SSD, and RP: ------------------------------------------------ Beatpath and CSSD are equivalent. In public political elections, with many voters and no pairwise ties, Beatpath, CSSD and SSD are equivalent. In comparison to Ranked-Pairs (RP), Beatpath is easier to program, and somewhat faster to compute (though computation time for Beatpath or RP will be negligible with modern computers). But RP is incomparably more briefly-defined, and therefore easier to propose. And its rule-justification is clearer and more obvious. As a proposal for public elecions, the considerations in the paragraph before this one are more important. But, for many organizations, the considerations in the paragraph before that are more important, explaining why it is Beatpath/CSSD that those organizations use. A detail: It is said that CSSD stands for Cloneproof SSD. And it now does, and that's fine. But when I named CSSD, I meant "Committee SSD", because in small committees, where there can be pairwise-ties, that's where CSSD and SSD can differ, and where CSSD has an advantage over SSD. Anyway, Beatpath/CSSD, SSD, and RP, are disqualified, by the chicken dilemma, from public political elections, though they're probably the best for organizations that are sure that they don't have chicken dilemma. ---- Election-Methods mailing list - see http://electorama.com/em for list info