James--
I'd said:
For instance, I said, and contnue to say that emphasis on methods' "vulnerability" to strategy completely misses the point.
So does the treatment of what you call "burying strategy" as a separate method problem from the strategy problems of Plurality and IRV.
You replied:
It is a separate problem if it succeeds.
Ii reply:
No! If it succeeds, it's indistinguishable and identical to what happens in Plurality and IRV without anyone using offensive strategy.
I added the exclamation point to "No" because that was the point that I was trying to make in the paragraph that you were replying to. And because it's something that I've been trying to convey to EM, apparently without any success, for quite some time. Paradigms die hard, if a paradigm is a collective assumption that someone has a lot of prestige invested in, or which protects something that someone has a lot of prestige invested in, and which somehow has come to be considered to be not subject to question.
The purpose of defensive strategy, by definition, is to avoid violation of majority wishes or to protect the win of a CW. But majority wishes are routinely violated, and CW's routinely denied victory, in Pluralilty and IRV, without anyone using offensive strategy.
Plurality and IRV violate majority wishes and fail to elect CWs easily and automatically without offensive strategy being used. Offensive strategy in Condorcet wv can violate majority wishes and take victory from a CW. The difference? The only difference is why it happened and what method it happened in. The mechanism details can be different with different methods. The result is identical.
You continued:
It may resemble existing compromising problems more closely if other voters counteract it using a compromising counterstrategy (a defensive strategy, in your terms).
I reply:
If it's allowed to happen, it is identical in Plurality, IRV, Approval, CR, or Condorcet wv.
In any of those methods, defensive strategy can prevent it.
One difference in the various methods is what degree of defensive strategy is needed. In Plurality, defensive order-reversal is needed. LIkewise with IRV. In Approval, CR, and Condorcet wv, voters can prevent that violation by mere equal voting.
Furthermore, in Condorcet wv, that is only necessary if it's expected that someone is going to use offensive strategy. The fact that, in Condorcet wv, the problem needs offensive order-reversal in order for the problem to exist doesn't mean that the result is any different from majority-wishes-violation or CW-loss in Plurality or IRV. What it does mean is that the existence of the problem, and the consequent need for defensive strategy, is rarer, because the problem only exists when offensive order-reversal is threatened.
So the differences involve the degree of defensive strategy needed, and whether the problem happens automaticallly, or only via risky offensive strategy. But the result, the violation of majority wishes, or the loss of a CW, is identical and indistinguishable, whether the method is Plurality, IRV, Approval, CR, or Condorcet wv.
By the way, I didn't meantion that offensive order-reversal can be deterred in Condorcet wv via truncation. Or that AERLO &/or ATLO avoid the need for the voter to actually do the equal ranking or truncation.
And also by the way, about AERLO and ATLO, the discussion changed to a debate of AERLO & ATLO vs Cardinal Pairwise. But that isn't what it started out as. It began about whether Condorcet wv had too much of a strategy problem. Well, if that strategy problem is greatly reduced by AERLO, ATLO, or Cardinal Pairwise, it's still greatly reduced, and that affects the issue of whether or not Condorcet wv, for you, has too much of a strategy problem. AERLO, ATLO, or Cardinal Pairwise get rid of nearly all of what little strategy problem Condorcet wv has.
And I repeat that, even if the offensive strategy isn't countered, and succeeds, the result is no worse, and no different, from what it would be if Plurality or IRV did the same thing without needing any offensive strategy to do so.
So no, offensive order-reversal in wv is not a new problem. It's an old problem that happens a lot less easily, and is more easily countered when it exists, in wv, as compared to Plurality or IRV.
I'd said:
But you misidentify the result of what you call "burying strategy" as something different and unique to Condorcet, you tell us that Condorcet
adds a new problem, a new fault.
You replied:
No, it's not unique to Condorcet methods. It's also present in Borda, Bucklin, approval, and CR. Not in IRV, though.
I reply:
Yes it's present in IRV. And it's present in Plurality. Not offensive order-reversal. We were talking about the violations that offensive order-reversal can cause: The violation of majority wishes, and the loss of a CW, which happen much more easily, and require more drastic defensivse strategy to prevent, in Plurality and IRV.
You continue:
See Blake's page
I reply:
I've heard Blake's opinions, and have checked out his website.
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
