>IRV violates your IIAC to be meaningful? Or is your argumentation
>completely arbitrary?
Let me explain it for you again: IRV advocates often cite Arrow, to try to show that
we can't get a good method, so we might as well have IRV. When they do, I point out
that by Arrow's outcome criteria, Approval is better than IRV. In other words, by the
criteria cited by that IRVie, IRV isn't as good as Approval. By those outcome criteria,
in fact, IRV isn't even as good as Plurality.
You ask why I could possibly believe that anyone would consider that meaningful.
The person who cites Arrow presumably considers Arrow's criteria meaningful. But
don't ask me why. Ok, I'll guess: That person believes that Arrow says we should
throw up our hands and give up, and that encourages the IRVie to say "Ok, since we
have to give up, we might as well have IRV." _That_ is why the IRVie considers Arrow
to be meaningful. That's my best answer to your question.
Now, as for _my_ IIAC, it's meaningful in a discussion of Arrow's criteria because
It's the only IIAC available to the discussion. It's the only one that's ever been completely
defined on this list, and it's the only one that has been completely defined for whatever
audience the IRVie is citing Arrow to.
You vaguely suggested another IIAC. I suspect that if you ever defined it completely,
we'd find that no method can meet it. If no method could meet IIAC, it would be pointless
for Arrow's impossibility theorem to mention any other criteria. For that reason it's
unlikely that IIAC is defined in the sort of way that you vaguely suggest.
Markus continues:
>You wrote (14 Dec 2000): > > So, as I said, then, what's questionable is why you think that the > > fact that I can't explain the desirability of a criterion that I say > > is not really desirable has anything to do with the validity of "my" > > "concept" that Plurality doesn't collect or use rankings. > >Plurality met your IIAC only when it couldn't be defined on >preferential ballots. When plurality is defined on preferential
>ballots, then plurality violates your IIAC.
No, Markus, none of us are interested in a criterion-applying system that you can
only apply to one method. Because you're unable to tell how your system applies to
other methods, it's of no interest how you say it applies to Plualitly. Give a complete
definition of your criterion-applying system, with its assumption that all methods collect
rankings, and then we can talk about applying it. Until you tell how it applies to other methods
, it can't legitimately be used at all.
>Therefore your
>"universally accepted" concept that plurality couldn't be defined>on preferential ballots
Let me explain something: "Couldn't" and "is" have different meanings.
I didn't say that there's a universally-accepted concept that Plurality couldn't be defined
in terms of rank ballots. As I & someone else pointed out, you can define anything any
way you want, but when you do, you must understand that your newly-defined Plurality
might not be much like the one that most people have heard of.
What I said is that it's universally accepted, by everyone except you, that Plurality
doee _not_ collect or use rankings. Like Don, you keep repeating that you can
define Plurality in terms of rankings. But it's really silly if, when you apply criteria, you
say you want to make up a fictitious definition of one of the methods, before applying
criteria to it. "Pluralitly fails your IIAC, after I redefine Plurality." :-)
You speak of a system that assumes that all methods collect & use rankings, for the
purpose of applying criteria to them, but you've demonstrated that you can't define
your general system. And you've never told me how how you'd apply your system to
Approval, much less given a general rule for your system. Since you can't do that,
you shouldn't waste our time by continuing to talk about defining Pluralilty based on rankings.
> is a necessary presumption of your
>argumentation that as plurality met your IIAC and IRV violated >your IIAC plurality was better than IRV.You want us to change Plurality's definition from the definition that's used in Plurality's
actual implementations. For applying criteria, you want us to change the definition of
one particular method, Pluralilty, before we apply the criteria to it. Or would you
change other methods too? Who knows? Would you assume that other methods, like
Approval, collect & use rankings too? Who knows--you won't tell. If so, how do we
apply your system to Approval? You don't know. Do we use different rules for different
methods, when assuming that they collect & use rankings? Or do you have a uniform
rule for all methods? You apparently don't know. You don't know what you're saying.
You don't know what you're talking about. You continue to waste our time by saying
we should use a criterion-applying assumption that you're unable to define for us.
Do you see the silliness of that?
Until you're ready to define your system, don't waste our time by saying that we should
define Plurality based on rankings.
Mike Ossipoff
Get your FREE download of MSN Explorer at http://explorer.msn.com
