[EM] Declaring victory?

[MIKE OSSIPOFF]

I haven't yet posted my final reply to Russ, and so I'll reply to this. I'd 
thought that this was one of Russ's political off-topic postings, due to its 
subject-line.

Russ says:

Well, if Saddam Hussein could declare victory after the 1991 Gulf War, I
suppose Mike can declare victory here.

I reply:

Thank you, but I haven't declared victory, though I've pointed out a few of 
Russ's usual errors.

Russ continued:

Mike posted a message in reply to James in which he stated, as he has
many times, that a proper definition of the Condorcet criterion should
be based on preferences rather than actual votes.

I reply:

That would be true only if you want CC to be uniformly applicable to all 
methods, without basing passing or failing on a method's rules. I use CC as 
an example, only because it's well-known and often mentioned, not because I 
consider it as important as the defensive strategy criteria.

I told why it doesn't make a lot of sense for a criterion to say "This 
criterion doesn't apply to nonrank methods", or "Nonrank methods don't pass 
this criterion".

I'll repeat it again: Criteria are for comparing methods. Criteria that 
don't compare some methods are less useful. Methods that compare all methods 
are the worthwhile criteria. There's no need for criteria that don't compare 
all methods.

If you'd like to replace Plurality with a better method, then it would be 
nice to have criteria that compare that proposed method to Plurality. That's 
obvious. You presumably would like to tell people why your proposed method 
is better than Plurality. A criterion that says, "Nonrank methods don't pass 
because I say so", doesn't tell anyone why BeatpathWinner, etc. is better 
than Plurality.

Results criteria are the only useful criteria. If your proposal is really 
better than Plurality, then presumably something about Plurality's results 
will bear out that claim. Results criteria will show that. Russ's "Because I 
say so" criteria won't show anything except Russ's confusion.

Russ continues:

Mike then replied to me with another of his typical personally insulting
posts, including this little gem:

"I told James that replying to him is like being a highschool teacher in
highschool for problem students. But replying to Russ isn't like that.
Replying to Russ is more like trying to talk with a babbling, drooling
person whose meaning can't even be determined."

I reply:

I said that for the simple reason that it wasn't possible to determine what 
Russ meant.

Russ continues:

I then found what is perhaps the original definition of the Condorcet
criterion by Kenneth Arrow, who won a Nobel Prize for his work. Arrow's
definition of the CC was, "that a candidate who receives a majority as
against each other candidate should be elected."

I reply:

Is Russ trying to use Kenneth Arrow as an example to confirm my statement 
that most voting system academics have their head all the way up their ass? 
Congratulations then, Russ, you've succeeded.

Plurality passes Arrow's CC, as Arrow defined it in that passage. Russ 
astutely suspects that Arrow didn't intend for Plurality to pass. Maybe, 
maybe not, but what difference does it make? As I said before, we could 
waste many kilobytes debating what the academic authors mean. But what 
difference does it make?

Well, it's relevant for this purpose: I said that I feel that most people 
intend or expect Plurality to fail CC, and my CC was written to comply with 
that expectation or intent.

But Russ keeps crowingly re-asserting that Arrow intended for Plurality to 
fail CC. Yes probably, but so what?

And, if Arrow meant that he wanted Plurality to fail CC, but wasn't very 
good at saying what he meant, and if we want to write a CC  that does what 
we guess that Arrow meant, then Russ would write it by saying "Nonrank 
methods don't pass because I say so." I've commented on that above.

Russ continues:

It said nothing about
preferences, so it clearly agreed with my position and disagreed with
Mike's (arrogantly and rudely asserted) position.

I reply:

No, it agrees with my claim about where the voting system academics have 
their heads.

And I repeat that I've never claimed that Arrow or any other academic author 
wrote or proposed a preference CC.  So how am I proven wrong by proof that 
Arrow didn't mention preference??

Well, I should say that I have run across preference CC definitions that are 
failed by every method.

And the Arrow quote doesn't show that Arrow wanted Plurality to fail CC. 
Maybe he did, maybe not. I won't argue with Russ about that, because I don't 
care. But I've previously said that most people probablly expect or intend 
for Plurality to fail CC, and that that informed my CC definition.

Russ continues:

Rather than gracefully admit his error, Mike now has the audacity to
lump Arrow in with all the sloppy "academics" whose attention he has
been unable get all these years.

I reply:

What is Russ's evidence that I've been trying to get academics' attention 
all these years? This is another example of how what Russ often has 
absolutely no relation to what Russ believes to be true. There are people 
like that. There are so few people completely lacking in honesty that you 
probably don't expect to encounter one. But here you have.

I haven't tried to get acacemics' attention. When Russ suggested to me that 
I should try to reach the academics, I pointed out that their goals are 
different from mine, and that there would therefore be no point in 
communicating with them. Is that what Russ calls trying to get the 
academics' attention? :-)

Maybe Russ means that I have the audacity to not share his uncritical 
hero-worship.

Russ continues:

But at least he is apparently no longer
trying to change the definition of the CC by fiat

I reply:

My CC definition is the only one that complies with people's expectations 
and intent about CC, as a results criterion rather than  a rules criterion, 
applying uniformly to all methods without mention of methods' rules.

I'm not trying to change the definition of CC, but I've written a definition 
that does what people expect and intend for CC, and is universally 
seamlessly applicable.

Again, CC is just an example, because it's well-known. This applies to other 
criteria too, such as ICC, MMC, and the majority defensive strategy 
criteria.

Russ continues:

>Now, back on topic:
>
>Summary--Unstated CC assumptions:
>
>Russ says that the academic voting system authors intend something that 
>they don't say--that they aren't saying what they intend. Russ says that as 
>if he thinks that it contradicts something that I've said. Actually it's 
>something that I've been saying for years, and have said a few times since 
>Russ has been on the list: Most voting system academics do a very poor job 
>of saying what they mean. That isn't a new discovery of yours, Russ.

In other words, Mike doesn't consider Kenneth Arrow to be in his league.

I reply:

Must I further clarify and spell-out what I've said about Kenneth Arrow?

But my statement that most voting system academics do a very poor job of 
saying what they mean, and that that isn't a new discovery of yours, and 
that your Arrow quote confirms the uncomplimentary things that I've been 
saying about voting system academics--I hope that you understand those 
things, so that you don't have to keep asking about them.

I'd said:

>So preference criteria aren't really a new and heretical thing. It's just 
>that mine don't share the ridiculous fault of previous ones.

Russ says:

A red herring. I never said that "preference-based" criteria are
"heretical" or anything like it. They are what they are.

I reply:

I'm not saying that it matters, but Russ, as an authoritarian, seemed to 
think that saying that Arrow didn't define preference criteria was a 
devastating and decisive argument against preference criteria :-)

...unless Russ had some other, unknown, reason for proclaiming that Arrow 
didn't write preference criteria.

Russ continues:

That's right, Mike. CC applies only to ordinal voting systems. In other
words, a voting system must be ordinal to pass CC.

I reply:

"CC applies only to ordinal voting systems" means that Plurality doesn't 
fail CC. Is that what Russ means? Does Russ know what he means?

Russ states that rule as if it were the definition of the only right CC. 
I've told why it's a really silly way to define CC, or any criterion. So 
Russ appoints himself to decree what the genuine, correct CC is, and his 
decree says that CC is something extremely silly, muddled,  and lacking in 
usefulness or meaning.

Sometimes Russ seems to be asserting that that's how the academic authors 
intend CC: "Nonrank methods don't pass because I say so". I'm not tring to 
claim to know if that's what he means, but, if it is, then he's claiming two 
things: 1) The academics, including Russ's hero, Kenneth Arrow,  aren't 
doing well at saying what they mean; and 2) What they mean is something that 
is head-up-the-ass.

Russ continues:

Oh wait, isn't that
just another way of saying that ordinal voting is a *requirement* of CC?

I reply:

Yes, Russ, we're all aware that you claim that the only right CC is defined 
in that way. I've told you of the stupidity of your definition. Sometimes 
you seem to be claiming that Kenneth Arrow wants to say that too, but can't 
quite say what he means.

Russ continues:

This isn't rocket science, folks.

I reply:

That's another saying that is often heard from someone who has just shown an 
astounding amount of ignorance about something.

I'd said:

>Now, "Application: Rank Balloting" isn't the same as "Requirement: 
>Rank-balloting". Russ, you won't understand that, so just take my word

Yes it is.

I reply:

Ok Russ, I'm going to slowly and carefully spell it out for you again:

"Application: Rank Balloting" means that the criterion doesn't apply to 
nonrank methods. That means, Russ, that a nonrank method neither passed nor 
fails the criterion.

"Requirement: Rank Balloting" means that the criterion requires rank 
balloting. That means that a nonrank method fails that criterion, because it 
doesn't have what the critrerion requires (rank balloting).

I don't know how to make that any clearer or simpler for you, Russ. But I do 
suggest that, if you're still all confused about that distinction, you 
really shouldn't be posting to EM.

>for it. If Blake had said that rank balloting was a requirement of his 
>voCC, that would mean that nonrank methods fail. But when Blake says that 
>rank balloting is the application of his voCC, that means that his voCC 
>doesn't apply to nonrank methods. This will be obvious to most of you, but 
>Russ can take my word for it.

This is really just semantic quibbling, of course, which Mike spends a
lot of time on. Suppose I define a "Harvard graduate criterion." If you
never went to Harvard, you "fail" the criterion, of course. Does that
mean that you "failed" the curriculum at Harvard? No. It just means that
you fail the criterion. What Mike is doing is analogous to making a big
deal of the semantic difference between "failing" at Harvard and simply
"failing" to have a degree from Harvard.

I reply:

Sloppy. If your Harvard graduate critrerion has, as its requirement, a 
Harvard degree, then yes, a person who never went to Harvard, never failed 
at Harvard, unquestionably fails the Harvard graduate criterion.

You're wasting my time. Yours too, but you don't seem to value your time 
much. I wait a week or more before replying to you, rather than replying 
promptly, as you do, so that you won't be able to waste as much of my time.

Blake, by saying that his criterion doesn't apply to nonrank methods, was 
saying that nonrank methods don't fail his criterion. I accept the fact that 
you'll never understand that distinction. But, that being so, you have no 
right to resent it when I compare this discussion to teaching an elementary 
school class for problem students. I mean no criticism of elementary school 
students--but that isn't really the purpose of EM. That's why I say that 
you're wasting my time.

Though there is an obvious difference between those two things, I don't 
really care which one you're trying to say is what CC says (or means to 
say). Either way, the criterion would be saying "Nonrank methods don't pass 
because I say so."

I'd said:

>So Russ's own criterion example that he posted, Blake's voCC, contradicts 
>Russ's claim. If the people who don't state their intention share Blake's 
>intention, then their intention is that voCC doesn't apply to rank methods. 
>Russ, that is _not_ the same as saying that nonrank methods fail the 
>criterion. I spell that out like that because you've previously indicated 
>that you don't understand that distinction.

Russ says:

No, I understand it completely

I reply:

Oh, ok, that contradicts what you said above, when you said "Yes it is [the 
same thing]".

Russ continues:

, but it's what is known as "a distinction
without a difference." Ever heard that expression?

I reply:

The difference is this: With one wording, Plurality fails the critrerion. 
With the other wording, Plurality neither passes nor fails the criterion.

That isn't the same thing. Sometimes you disagree, sometimes you say you 
understand. Whatever.

No, I didn't say that it's a really important distinction. Etiher way, it's 
a stupid criterion with little usefulness or meaning. If you ever decide 
which one you mean, it won't matter, because both make for a silly 
criterion.


I reply:


>Do some authors intend, without stating it, that nonrank methods by fiat 
>fail their voCC? Probably some do. Many or most, in fact. But we can't

Where does the word "fiat" come in?

I reply:

It means that the criterion says "Nonrank methods fail this criterion", 
which means "Nonrank methods fail because I say so." Nonrank methods fail by 
decree. We can say "decree" instead of "fiat" if you really want to decree 
it.

Russ continues:

Mike is the one who tried to change
the CC by fiat to something that its Nobel Prize winning originator
never intended. No, it's not by "fiat," Mike, it's by *definition*.

I reply:

What do you think "by definition" means? It means "Because I say so". Call 
that "by fiat", "by decree", or "by definition".

Russ, here too, again, is all confused. I don't claim to have changed 
someone else's definition, by fiat or otherwise. However I've written a 
definition that is better. I've told why it's better. It is seamlessly, 
universally applicable. It acts as people expect and intend, without any 
mention of methods' rules.

I can't decree what someone else's CC definition is, but I do point out the 
stupidity of the one that Russ is promoting.

Russ continues:

As I wrote earlier, I'm making some progress with Mike. At least he now
seems to realize that he cannot just change the CC by fiat.

I reply:

Relax, Russ, no one can change you by decree. I've written a better and more 
useful CC.

Russ continues:

He can
define his own version, of course, but he needs to give it a different name.

I reply:

Condorcet expressed the goal that a rank method should elect the BeatsAll 
candidate. Extending that to cover more general methods doesn't require a 
name-change. It's now felt by (nearly) everyone that CC should apply to all 
methods, and be failed by Plurality. I've written such a CC definition. One 
that applies more uniformly than the mess that you call CC.

You said:

For someone who derides "ESP" so often, you seem to have an
extraordinary insight into my "Russ's mind."

I reply:

Just guessing.

>The kind of criteria that Russ advocates say "Nonrank methods fail because 
>I say so." Somehow that isn't as convincing as having a

No, not because "I" say so, Mike, but because common sense says so.

I reply:

You mean "Because I say common sense says so."

Invoking common sense doesn't tell why BeatpathWinner is better than 
Plurality. Surely any idiot could understand the desirability of being able 
to compare Pluralitiy to better methods, to show how Pluralty fails, by its 
results, not because Russ says that common sense says so.

Mike Ossipoff

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/