[EM]Re: AWP versus AM (and DMC)

[James Green-Armytage]

James Green-Armytage replying to Chris Benham...
>
>Your examples of  AWP resisting strategy better than
>AP have not been very impressive to me (or I
>gather,Ted).

	It's not just the examples, it's the analytical argument. It is
irresponsible for you to say that you find it unimpressive without
actually contradicting any part of it. If you don't understand my
argument, I suggest that you are not applying a lot of effort toward
understanding it. I obviously cannot force you to understand something
that you don't want to understand. If you don't have the time to engage
intelligently in this discussion, that's perfectly fine, but you ought to
admit that this makes you relatively unfit to comment on the issue.
	It is possible that strategic vulnerability would kill AM or DMC as a
viable method for use in contentious elections. I have expressed my
reasons for thinking this as well as I have been able. If you or anyone
else was willing to argue the other side of this in any meaningful way,
then we might be able to advance our understanding to some degree, but so
far you have just been avoiding the issue altogether, saying that you are
unconvinced by my arguments without actually replying to any of them.
>
>"I could just as easily say that AM and DMC have a
>huge cost in that they flout the AWP criterion."
>
>CB: Not with any credibility, since the AWP criterion
>was "obviously designed just to support" the method
>(as you put it); and it doesn't seem to test for
>anything that anyone was worrying about before  the
>method was invented.
>The AM criterion, on the other hand, is the perfectly
>natural putting together of  two obviously fundamental
>ideas: "that pairwise beaten candidates should tend to
>lose" and "that more approved candidates should tend
>to win"!

	You mean the "AP criterion", right? "Don't elect a candidate that is
pairwise beaten by a more approved candidate." If you want to call it the
AM criterion, that's fine, just let me know. But I'm pretty sure that it
was the AP criterion last time.
	Restating the AWP criterion: "A voter expresses a strong preference for X
over Y if the voter approves X but not Y. If A pairwise beats B, and there
is no beatpath from B to A such that each defeat includes as many or more
strong preferences than the A>B preference, then B shouldn't win." The AWP
criterion is the AWP-defined immune set, just as Mike's majority rule
definition is the WV-defined immune set.
	I claim that the AWP criterion is a very good criterion for rank/cutoff
methods, in both a sincere and strategic context. Despite your insults, it
is not a "mad" or cynical criterion, but a natural and useful one. The
basic idea is that defeats that include many strong preferences shouldn't
be overturned unless it's absolutely necessary. (When is it absolutely
necessary? When there is a beatpath of defeats with even more strong
preferences leading back from the loser to the winner.) There is nothing
mad or convoluted about this logic, no matter how persistently you
belittle it and ignore it. 
	If a pairwise defeat exists, and a lot of people strongly agree with it,
then it is important to preserve it if possible. If a pairwise defeat
exists, and very few people strongly agree with it, then it is less
important to preserve, and it should be overruled if it forms a top cycle
with pairwise defeats in the former category. Where is the madness in
this? How does this logic seem so obscure to you?
	What is the logic behind the AP criterion? If A is "more approved" than
B, he's probably better, and if A pairwise beats B, he's probably better,
so if A is "more approved" then a candidate B whom he pairwise beats, he's
definitely better, so why elect B? Is that it? How solid is that
inference? I suggest that it is faulty.
	Just because the AP criterion is simple doesn't mean that it's actually
more valuable than the AWP criterion (and we can't have both, because
they're incompatible). I'm not convinced by the logic that I cynically
re-stated in the preceding paragraph, so I need more justification, but
I'm not receiving any. I think that there are perfectly good reasons to
elect B, e.g. if there is a B-->A beatpath consisting of defeats that
include many strong preferences, while the A>B defeat includes few or no
strong preferences.
	In other words, rather than telling me your criterion and praising its
simplicity, you need to make an argument for _WHY_ it is so danged
important that it outweighs both the sincere and strategic benefits of
AWP. I've made several arguments why I think AWP is important, but all you
seem to say is that AP is simple and obvious and easy to understand.
That's not a good enough reason. You seem to think that two measures from
two separate methods (pairwise and approval) add up to one invincible
measure when they agree... is that really logical? Does the convergence of
the two measures really indicate anything more than the two measures
indicate individually?
>
>I  concede that the AWP method has its own mad logic,
>and  that it succeeds somewhat at what it tries to do.
>(And you can take that as my general reply to all
>those arguments of yours that you complain that
>I  "haven't replied to".)

	Mad logic? You're obviously not trying to discuss this in a serious way
with me. If you can't keep enough of an open mind to avoid calling my
logic "mad", we should probably stop trying to converse on this subject.
Maybe sometime you could write a note to me off-list and let me know
whence came this impulse to needlessly insult me... as I remember we used
to have a rather pleasant and cooperative correspondence.
>
>But I think a lot of voters wouldn't understand the
>AWP method or criterion, and some or many of those
>that
>do wouldn't be greatly reassured by the criterion.
>On the other hand, I think most of the voters would
>understand and appreciate the AP criterion, and
>meeting it means  that  we always have some very 
>simple and effective retort to potential complainers.

	It's necessary to separate the question of merit from the question of
easy mass appeal. You seem to use the mass appeal argument to compensate
for having a weak (or absent?) merit argument. As to the mass appeal
argument, it's difficult to settle without extensive focus group research,
but I suggest that an electorate that is engaged enough to switch to a
Condorcet-efficient method is likely to be thoughtful enough to weigh the
relevant strategic concerns, and to choose AWP over DMC/RAV.

>No, but here the Dean voters can argue for  DMC 
>or  Approval Elimination. 

	Why would they do that? Doesn't that also result in the election of their
last choice (Bush)?
>
>
Sincerely,
James