Chris--
I'd said:
Remember that it's been shown that every nonrandom method gives some kind of incentive for strategic voting."
You reply:
CB: Yes, but not in the zero-info. case. I don't think that DD(WV)'s 0-info. incentives combined with its vulnerability to Burying and defection is the best bargain.
I reply:
I can't agree or disagree without knowing what DD is. DD stands for direct democracy, but I assume that that isn't what you're referring to.
I suggest that you worry too much about 0-info strategy. Though our elections are now 0-info, they won't remain 0-info if Approval or Condorcet wv replaces Plurality.
You're concerned that wv is vulnerable to burial and defection. I'm not quite sure what you mean by defection, but if it refers to truncation, then I point out that, for the conditions described in the premises of SFC & GSFC, and for the majority referred to in those premises, nothing short of preference-falsification can cause strategy need. In other words: No truncation problem.
But if, by defection, you're referring to the co-operation/defection dilemma that James described, then it's true that wv is subject to that problem--but not if ATLO is available.
"The equal-ranking incentive that you describe is gotten rid of by AERLO."
I am strongly of the view that as far as possible, the result of the election should be determined purely by the voters, who are nominally (or "in theory") voting sincerely.
I reply:
But in methods that don't meet the defensive strategy criteria, it would be unrealistic to expect people to vote sincerely.
Therefore I'm opposed to explicit strategy devices
I reply:
Ignoring strategy need won't make it go away.
I'd said:
"Still, it would be a nice luxury for everyone to have no incentive to do other than sincerely rank those better-set candidates."
You continue:
I am sure that SCRIRVE, and doubtless some other methods, fill that bill.
I reply:
Doubtless some do. Margins does. Maybe SCRIRVE does too. The problem is the bills that they _don't_ fill. When they avoid that mild strategy incentive, while causing serious strategy dilemmas in which majorities are strategically forced to reverse their preferences or bury their favorite in order to protect the win of a CW or to protect majority rule, then I suggest that those methods are losing more than they're gaining for the voter who wants to reduce strategy need.
You continue:
I think the "Definite Majority Choice" idea recently floated here of having the voters enter an approval cutoff in their rankings and then the method eliminates all candidates who are pairwise beaten by more approved candidates; is promising.
I reply:
Maybe some of those proposals meet Condorcet's Criterion. Probably their use of Approval brings some additional benefit. Probably some of them meet WDSC, and would therefore meet SDSC if they had AERLO. But I doubt that they meet SFC.
If you're going to propose a rank method, propose one that meets at least WDSC and SFC. Why settle for less, when the simple and modest PC meets those two criteria, and when SD, SSD, BeatpathWinner/CSSD and RP additionally meet GSFC and SDSC?
Of course it would be premature to say for sure that they don't meet SFC. But I don't doubt that the new methods have advantages, and I'd be interested in hearing more about their advantages.
Mike Ossipoff
_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
---- Election-methods mailing list - see http://electorama.com/em for list info
