[EM] minmax is not a good public election method

[James Green-Armytage]

James replying to Kevin, on the topic of minmax(pairwise opposition)...

James:
>
>> 	If the electorate is ready for something as complicated as that, then
>> beatpath(wv), ranked pairs(wv), and river(wv) will be viable options.
>> Aren't these methods more elegant than CDTT,MMPO?
Kevin:
>
>Why do you think so? 

	I usually find methods inelegant when they combine two logically separate
tally procedures. E.g. Smith//minimax, CDTT,IRV, anything with AERLO, etc.
Beatpath, ranked pairs, and river have an internal logic that makes them
naturally Smith-efficient without the need to specifically find the Smith
set.

>I don't think "elegance" is the main point; I bring up CDTT,MMPO as an
>attempt to preserve MMPO's good qualities, not to compete with Schulze,
>etc.

	Okay. (I don't think that elegance is the main point either, but it is a
factor for me, on the theory that inelegant procedures may often create
awkward strategic dynamics.)
>
>Well, obviously it's unnecessary if you don't think LNHarm is valuable,
>which
>clearly you don't, since you value criteria that avoid rare but really bad
>situations, more than criteria which (according to me) provide guarantees
>that are almost always useful.

	Smith-efficiency and resistance to strategy are two of my primary goals.
There seems to be a difficult short-run tradeoff between these and public
salability. The CWP-defined immune set is a more ambitious goal of mine,
and one that probably conflicts even more with public salability than the
others in themselves.
>
>Well, if you want to insist on Smith, why do you bother to type out 
>"Condorcet," "MMC," and "CL"? Just to have a long list?

	It's quite possible to fail some while passing others. E.g. IRV fails
Condorcet and Smith, but passes MC, MMC and CL. Borda passes CL, but fails
the rest. Minmax only passes MC and Condorcet. MMPO doesn't pass
Condorcet, although it nearly does. 
	So, to say that MMPO fails Smith does not imply that it fails the other
criteria. (Although it is true that a failure of the any of the other
criteria implies a failure of Smith.)
>
James:
>> >> 	Why do you think later-no-harm is so important?
Kevin:
>> >Because it gives voters the ability to safely rank all of their
>> >preferences. 
James:
>> 	It is not always safe for voters to rank all preferences in MMPO. As in
>> WV, truncation (equal ranking) can still be useful (arguably necessary
>in
>> some cases) as a deterrent counterstrategy.
Kevin:
>
>Great... I've already said all of that. In any case, I don't think this
>weakens my point much.

	Well, we disagree there. It seems to directly contradict your point that
later-no-harm is important because it "gives voters the ability to safely
rank all of their preferences." I suspect that deterrent counterstrategies
may need to be widely employed if WV methods (or PO methods) are used in
contentious elections. If so, this means that voters will often not vote
their full rankings, in which case I'm not sure how much practical benefit
MMPO's LNHarm compliance will have.
>
James:
>
>> 	If your favorite candidate (F) really doesn't have enough support to
>be a
>> viable candidate, then s/he is unlikely to have a beat against your
>> compromise candidate (C), in which case C can't be harmed by your
>> expressing an F>C preference (in Smith-efficient methods). (Thus, your
>> preference won't cause the election of a greater evil (E).) 
Kevin:
>I know it's unlikely. 

	Okay. 

>But it would be nice if the voter could be assured
>that he can always rank his favorites at the top.

	Yes, it would be nice, but it would be much nicer if voters could be safe
in ranking their favorites strictly in first place, rather than tied with
lesser preferences (no compromising-compression incentive or
compromising-reversal incentive). Probably no viable election method can
universally guarantee this, but CWP and AWP can bring
compromising-compression and compromising-reversal incentives to a very
minimal level. 
	Both you and Mike have mentioned equal ranking as the counterstrategy
response to a minority's attempt to cause a strong cycle to thwart a
mutual majority. This suggests that MMPO would have an additional
compromising-compression incentive not found in Smith-efficient WV methods
such as SD(wv), beatpath(wv), etc.
>
my best,
James Green-Armytage
http://fc.antioch.edu/~james_green-armytage/voting.htm
>
>