Condorcet cyclic drop rule

[MIKE OSSIPOFF]

>Mr. Ossipoff wrote-
>
>If no one falsifies a preference, and if a majority of the voters
>prefer the sincere CW to Jones, and if they vote sincerely, then
>Jones won't win. Guaranteed.
>---
>D- Three very big *ifs*.

Not really. As for the sincere voting, the whole idea of SFC is to
guarantee that that majority can defeat Jones merely by voting
sincerely. That's part of the goal, not an "if". When they can defeat
Jones even if they sincerely rank all the candidates, then they're
free of need for strategy for that purpose.

As for their preferring the sincere CW to Jones, that's hardly
iffy: The sincere CW is preferred to every one of the other candidates
by more people than vice-versa. Unless there's lots of indifference
between pairs involving the sincere CW, then, there's a majority
who prefer him to Jones. And if there's so much indifference that
that majority doesn't exist, then how much does it matter if Jones
is defeated?

There's one genuine "if": No one falisifes a preference. You might say
that in a 100,000,000 voter Presidential election we can't guarantee
that there won't be one person who falsifies a preference.

No, but it's vanishingly unlikely that falsification will be practices
by a sufficient number of voters to change the outcome. And if falsification 
doesn't change the outcome from what it would be if there
were no falsification, then, SFC-complying methods still deliver on
their promise in spite of the slight & ineffective falsification.

So I'd say that SFC's "ifs" aren't iffy at all.

Wording it to say that no one falsifies simplifies the wording. And
with methods other than Condorcet, even if no one falsifies, the
gurarantee cannot be made.

>
>With divided majorities, there will be *some* strategy/ insincere game
>playing.*whatever* method is being used.

Yes, I answered that in an earlier paragraph in this message. A few
sporadic falsifications won't change anything, whereas no other method
can make SFC's guarantee even if no one falsifies.

>What did Mr. Arrow observe ???

You tell me. Did he say that people would strategize, or did he just
say that several criteria couldn't be simulataneously met? Gibbard &
Satterthwaite showed that with any method there can be situations that
reward strategy. But they didn't say that there would be so much
falsification in public political elections that it would change the
outcome from what it would be without the falsification.

>
>A and B folks have *some* things in common (such as being conservatives or
>liberals). The Z folks are the *other* bunch of folks (liberals or
>conservatives).
>
>Initial Sincere poll
>
>26 A
>25 B
>49 Z
>100
>
>How far do A and B folks support each other ???

If you're saying that all the A & B voters prefer A & B to Z, then
if the A & B voters vote sincerely then Z will pairwise lose to both
A & B. A or B will win with any Condorcet version.

>
>Do any of the Z folks prefer A(B) ever so slightly (and insincerely) over
>B(A) ???

It won't help Z, if, as you implied, all the A & B voters prefer A & B
to Z, and if they vote sincerely.

>
>Do any of the A(B) folks prefer Z ever so slightly (and insincerely) over
>B(A). ???

When you said the A & B voters were liberals or Conservative and the
Z voters were the opposite, I thought you meant that the A & B
voters all prefer A & B to Z.

If you didn't mean that, then you'd have to give more information about
the preferences in order to establish what kind of an example it will
be.


>
>Who has YES support from a majority of the voters ???

You wrote the example; if anyone would know that, you would.

>
>Let the games begin.
>
>How mind boggled will the public/ legislators get when the issue of
>Ossipoffian reverse-order strategy comes up (as is *sure* to happen by
>enemies of *any* proposed reform) ???

It already has, in your message :-)

If you succeed in boggling the public about it, then we'll just go
with Approval or CR. But it sounds as if you've boggled yourself on
the subject. The only thing that can give anyone reason to strategize
to protect majority wishes or a sincere CW would be if lots of
people falsify offensively, in Condorcet. You said yourself that
strategy is part of every voting system. So why are you so boggled
by the fact that it's possible in Condorcet? In Condorcet, the only
conceivable strategy problem is one that won't happen unless falsification 
takes place on a very improbable scale. Can you show
me a method that does better in terms of allowing sincere voting?
If not, then what's your objection?

The difference isn't that strategy is conceivable in Condorcet. The
difference is that that scenario is so improbable in public elections
as to be ignorable.


Mike Ossipoff


_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com