[EM] MaxMinPA

[Michael Ossipoff]

Jameson--

DA (Disqualify/Approve) isn't without advantages.

But it's bad too many name changes.

The name should briefly describe its main distinction.

Your 1st name for it, DA, does that well. DA is distinguished by votes for
and votes against.

...approving & voting to disqualify.

Or: For/Against (FA).

And there's no need to complicate it. Default should be bottom-rating.

It isn't just that DA fails CD. It fails because, in the usual important
sense, it really lets defectors take advantage of co-operators.

A>B
B
C

(Default is bottom)

B is the only candidate not disqualified.

B's defection succeeded.

Yes, DA has something that Approval doesn't have:

The A voters & the B voters can all co-operate, bottom-rating C, and rating
eachother's candidate middle.

In that way, they combine to disqualify C, while still voting among
eachother.

In general, a co-operative mutual majority top-set of factions can all
combine to majority-disqualify their bottom-set, while using their
top-votes to have an Approval election among eachother.

But Score can achieve that, JC that co-operative & trusting mutual majority
give eachother a near-unity fraction of max points.

Of course Approval can achieve that Probabilistically.

If you object to the need for probabilistic voting, then I remind you that
DA needs it in a chicken dilemma.

So is DA's advantage more important than the use-precedent & greater
simplicity of Approval & Score?

In any case, of course, an electorate who have always used Plurality won't
accept a new method with no use precedent.

So it will have to be Approval, Score or Bucklin anyway.

Later something new could be proposed. MMPO offers Weak CD & wv strategy,
allowing people to rank as many candidates as they want to.

Michael Ossipoff

On Oct 20, 2016 11:12 AM, "Jameson Quinn" <jameson.quinn at gmail.com> wrote:

> So, too, Approval, Score & Bucklin need that plumping. It seems a
>> universal requirement.
>>
> Specifically, in a center-squeeze scenario, Bucklin needs enough plumping
> so that the median voter is plumping.
>
> Note that in MAS, as long as the two sides are reasonably balanced, the
> centrist voters can leave both other candidates blank, and that will be
> enough plumping, even though each side will get some non-0 votes from those
> ballots. If the two sides are perfectly balanced, then most of the
> centrist's voters can give one side a 1 and the other a 0, as long as there
> is some group at the median leaving both sides blank.
>
> This is not a guarantee, but I think it's good enough to help prevent a
> plumping strategy from being overused. As to how often it will be
> under-used so that MAS fails to elect a center-squeezed majority CW... I
> think that insofar as the stakes increase, the intuitive strategy will
> approach the ideal strategy, so that failures will tend to be in relatively
> low-stakes elections only.
>
> (Low-stakes of course refers to both the importance of the office and the
> marginal difference in quality between the frontrunner candidates.)
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20161021/5b33f44e/attachment.htm>