[EM] Replying to Markus

[MIKE OSSIPOFF]

Markus wrote:

you wrote (16 Apr 2001):
>I should add that, of the 4 list-members that I named (5 if you
>count me), all but one also prefer BeatpathWinner to Ranked-Pairs.

It isn't clear what you mean with "BeatpathWinner."

I reply:

I'm sorry that isn't clear to you, because I've defined it on
EM a number of times, and the term has been in use for over a year
here.

For instance, I defined it when I told why it's equivalent to
Cloneproof SSD. I've also defined Cloneproof SSD here.

If you can't find it in the archives, I'll repeat BeatpathWinner's
definition here in a few days.

Markus continues:

Due to the usual
way, in which new terms are created out of old terms in this mailing
list, the "BeatpathWinner" seems to be the winner of the Beatpath
Criterion Method.

I reply:

No, BeatpathWinner is the name of a method, not a candidate.

Markus continues:

This method has been proposed by Steve Eppley in
Feb 2000
(http://groups.yahoo.com/group/election-methods-list/message/5014):

>Here's another good method which is not equivalent to Tideman or Schulze:
>
>    Beatpath Criterion Method (BCM)
>    -------------------------------
>    For all pairs i&j,
>    if Vij > Vji and Vij > Bji then eliminate j.

However, I don't remember that Norman Petry, Rob Lanphier or I have
said that we prefer this method.

I reply:

And I've never heard anyone but you call that method "BeatpathWinner".

Steve and I both began using the term "BeatpathWinner" here around
the same time, to refer to what we then believed you to mean by
"Schulze's method". At that time we didn't know that "Schulze's method"
was really a Beat-&-Tie-Path-Winner, rather than BeatpathWinner.
So, Markus, pretend that Schulze's method doesn't allow ties in its
beatpaths, and then you've got BeatpathWinner, as that term has
been used and defined here for over a year. Ok?

That's the general meaning that's been in use all this time.
More specifically, I use the term for the version of that method that
deals with equal defeats in a way that preserves equivalence with
Cloneproof SSD, when Cloneproof SSD deals with them in the way that
we agree on: If 2 or more defeats equally have lowest defeat-support in
the Schwartz set, then drop the one with greatest defeat opposition.
If 2 or more have equally weakest defeat-support also share equally
greatest defeat-opposition, then drop them all simultaneously.


Markus continues:

I also observe that the term "BeatpathWinner" hasn't been used
before 2000, when Steve Eppley introduced his Beatpath Criterion.
Therefore it seems to me that your usage of the term "BeatpathWinner"
is possibly caused by the fact that you probably mixed up some of
Steve Eppley's concepts with some of my concepts.

I reply:

I have no idea what you're talking about. My usage of the term
"BeatpathWinner" is caused by the fact that the term has been in
use here for over a year, and has been defined here by me a number
of times.

As for mixing up your concepts, yes all of us thought that Schulze's
method was BeatpathWinner, until you pointed out that Schulze's
method is Beat-&-Tie-Path-Winner.

Mike Ossipoff




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