[EM] Re: Markus, RP vs BeatpathWinner

[MIKE OSSIPOFF]

Markus--

You wrote:

So you say that there is no significant merit difference
between Ranked Pairs and the beat path method, that its
"elegantly simple & brief algorithm & computer program"
make the beat path method a better proposal for committees,
and that its brief definition makes Ranked Pairs a better
proposal for the public.

I reply:

Yes, that's exactly what I said.

You continued:

However, according to Steve Eppley, there is a merit
difference.

I reply:

Is there a disagreement there. Recheck what you quoted me
as saying, and what you quoted Steve as saying, above,
and then tell me if the 2 statements disagree.

You continued:

Steve, who uses the term "MAM" for Ranked Pairs

I reply:

MAM doesn't just refer to Ranked-Pairs. It refers
to Ranked-Pairs(wv), when equally strong unconsidered
defeats are dealt with in a certain way, using a
random ordering.

You continued:

...and the term "PathWinner" for the beat path method, writes:

>MAM may be preferable to PathWinner for a couple of reasons:
>
>1. MAM (but not PathWinner) satisfies immunity from majority
>complaints (IMC), immunity from second-place complaints (I2C)
>and other criteria described in the document Immunity from
>Majority Complaints.

Sure, Ranked-Pairs meets some output-ordering consistency
criteria that BeatpathWinner doesn't meet.

You continued, quoting Steve:

>
>2. Computer simulations using randomly generated profiles of
>voters' orderings suggest the alternative chosen by MAM will
>beat pairwise the alternative chosen by PathWinner more often
>than vice versa, and that over the long run more voters will
>prefer MAM winners over PathWinner winners than vice versa.

I reply:

Sure, and that means something and counts for
Ranked-Pairs in its comparison with BeatpathWinner.

I'm not saying that there's no merit difference, only
that it isn't significant. Maybe Steve doesn't agree with
me on that. So what?

I avoid the BeatpathWinner vs RP merit issue because
it's divisive, and because I advocate both methods.

But if you're going to press me on the issue, then I'll
reluctantly comment: I agree with Steve that, on pure
merit, disregarding convenience, my impression is that
RP _slightly_ beats BeatpathWinner.

For the reasons that you quoted Steve saying. And also
because RP doesn't let a nullified defeat take part in
the nullification of other defeats. There's something
especially solid about sequentially keeping strongest
defeats if they don't contradict already-kept defeats.

That's really just
an aesthetic quibble, consisdering the excellent criterion compliances of 
both methods.

I emphasize that I still prefer BeatpathWinner/CSSD as
a recommendation to organizations & committees.

Mike Ossipoff




_________________________________________________________________
MSN 8 helps eliminate e-mail viruses. Get 2 months FREE*.  
http://join.msn.com/?page=features/virus