SW simplicity & decisiveness (was Re: [EM] Hitler-Stalin-Middle

[Mike Ossipoff]

I've probably misinterpreted & wronged Shugart again. He very likely
_wasn't_ saying that Condorcet's method violates his standard of
choosing decisively from a set of identifiable alternatives (How many
methods could violate that? Plurality doesn't). Most likely he meant
only that Condorcet violates his simplicity standard.

Though I disagree with the claim that Condorcet's method is complicated,
and even more with a claim that it's too complicated, I don't want
to be criticising other people's standards, just because they're
different from mine.

But since the simplicity standard is the one that Shugart said Condorcet
violates, let me address that: My definition of Condorcet's method starts
out with 2 very clear & brief sentences. The part on what to do when
no one beats everyone else is longer, but it is stated in 1 sentence.
Picking the person least beaten is obvious, natural, & simple.

Shugart has said that he likes a combination of Approval with a
vote for 1st choice. But Approval is something that many people
_don't_ understand. Many people will tell you that Approval violates
1-person-1-vote. So how is it, Shugart, that you consider a seriously
misunderstood method to meet your simplicity standard?

Double-Complement is a name that the vast majority of those who hear
it won't understand the meaning of. And the rationale for choosing the
requirement for winning in the primary the way you choose it is another
thing that would require long explanation to justify. You explained
it with equations, and you should know that won't work on ER, much less
with the general public. If I were you I'd just say that the requirement
for winning in the primary is being twice as close to a majority than
anyone else is. But even then you have to explain why that seemingly
arbitrary rule is proposed, and that makes Double-Complement a
complicated method.

***

Your standards are simplicity & decisiveness. Plurality meets those
standards admirably. Is Plurality acceptable to you? Double-Complement
is somewhere partway between Plurality & Runoff, and so it can't be
said to be something different from what we already have.

As I said, though, I don't want to criticize other people's standards
just because they're different from mine. Without criticizing, we can
just say that your standards for single-winner methods are extremely
minimal. 

And the decisiveness standard doesn't discriminate between any of the
methods proposed in this committee, since every one of our proposed
methods is decisive in a public election. Ties with any of these methods
are vanishingly improbable in a public election.

Steve said something that I was going to say. Though all of the methods
proposed in this committee are decisive in public elections, are you
concerned about the dilemmas that they impose on the voter? As long
as the method is decisive it's ok if it forces the voter to deal with
strategy dilemmas in order to decide how to vote. Decisive method 
that imposes difficult strategy decisions on voters?


As for what's simple enough, that's a criticism that's used too freely.
Surely it should be necessary to give very compelling evidence that
a method is too complicated in order to talk people out of proposing
it to the public. Merely calling it complicated isn't enough. As
I said, that should be for the public themselves to decide. None of
us should speak for the public, regarding what they'll accept--unless,
as I said, we give extremely compelling reasons for our claims.

***

This discussion of standards has been productive, because it gets
right at why people propose this or oppose that method. This clarification
will make all the difference when we make our report, collective
recommendation & FAQ. It seems to me that merely by starting to discuss
standards, we've gone most of the way toward completion of the task.

***

Mike Ossipoff





--