[EM] Markus, 0633 GMT, 21 March, '05

[MIKE OSSIPOFF]

Markus says:

Dear participants,

Mike Ossipoff proposed the MinMax(wv) tie-breaking strategy.
Now he claims that (by proposing this tie-breaking strategy)
he proposed wv methods in general.

I reply:

Have I ever said that I proposed wv methods in general by proposing the 
MinMax tiebreaking strategy? For one thing, I've never proposed the MinMax 
tiebreaking strategy, though I have proposed PC, and still advocate PC.

For another thing, though PC and Smith//PC were the only Condorcet methods 
discussed on EM at the time when I introduced wv, my advocacy of wv based on 
completely general considerations, which I've described in recent postings, 
makes it impossible for Markus to convincingly argue that I only advocated 
wv by proposing a method that could use wv.

Completely different answer to Markus's claim:

But (and I haven't yet said this reply to Markus's many-times-repeated 
statement), quite aside from what I've said above in this message, what if I 
had only suggested wv for PC? If I had advocated wv only in connection with 
PC, for use with PC, that wouldn't mean that I didn't introduce wv. The fact 
that Markus got wv from me and incorporated it in a different method 
wouldn't mean that I didn't introduce wv, even if it were true that I'd only 
advocated wv for use with PC. Not that that's true anyway.

Say You invent the horsehoe, and advocate it for use with horses. Now, 
later, mules have been bread, and  Markus puts a horseshoe on a mule, and 
says that you didn't introduce the horseshoe, not for general use, because 
no mules existed when you invented the horseshoe, and so you didn't invent 
the horseshow as a device for general use. Because you didn't advocate its 
use for mules, which didn't exist at the time when you invented the 
horseshoe.

Markus continues:

However, I argue that it cannot be said that Mike proposed wv
methods in general because he didn't propose a general concept
that could also be used for other methods than MinMax.

I reply:

wv is a general concept. It's a way of measuring pairwise defeats. That's 
what I introduced and proposed it as. That's what I advocated it as. I 
advocated it based on completely general considerations not dependent on any 
particular method.

Not only can wv be used for other methods than PC (if that's what Markus 
means, this time, by MinMax), but it _is_ used for methods other than 
MinMax. wv is currently being used with CSSD by the Debian organization. wv 
is used for BeatpathWinner and Ranked-Pairs by various websites that conduct 
and count polls.

Markus continues:

Mike
proposed his concept of "majority-rejected" candidates; with
this concept he proposed his "Generalized Majority Criterion"
(GMC); and he used this criterion to motivate the MinMax(wv)
tie-breaking strategy. However, GMC is satisfied by no other
wv tie-breaking strategy than MinMax(wv). Therefore, Mike
proposed only the MinMax(wv) tie-breaking strategy but not
wv methods in general.

I reply:

I introduced and proposed wv years before I defined GMC. And so it's odd 
that Markus thinks that GMC is somehow relevant to what I meant when I 
introduced wv.

Yes, this discussion has taken place many times during the past week. That's 
typical of Markus. If you are tired of Markus's repetition, then I recommend 
that you e-mail Rob about it, and about Markus's consistent violation of 
EM's main guideline of conduct. There are things that Rob could do, if 
members ask him to do something.

Mike Ossipoff

_________________________________________________________________
FREE pop-up blocking with the new MSN Toolbar – get it now! 
http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/