These comments will refer to Abd ul in the 3rd person, since Abd ul isn't
speaking to me.
Abd ul said:
The problem is that the definition of
clones and the independence criterion, under that definition, *sound* like
reasonable criteria for a fair election, but they are not.
I comment:
Actually they are. Abd ul objects to the term "clone set" to denote what is
known as a clone set. That's fine, and "clone set" isn't a name that I would
have chosen. Before Tideman's "clone set" term was introduced on EM, I was
calling those sets "co-partisan sets". I consider that a better name than
clone set, because yes, "clone" implies identicalness.
But Abd ul's objection to that term, however valid it may be, in no way
means that the criterion itself is unreasonable. Change the name of clone
sets and ICC, but it's still a reasonable criterion.
Of course I've told why MMPO's failure of ICC isn't important: The failure
requires an unlikely strong cycle among the clone set candidates. Their
voters apparently can muster much greater votes against eachother than
against their actual opponents. Someone said that could happen
strategicallly, by preferrers of candidates outside that set. It was pointed
out by more than one person that it's odd that the opponents are
sophisticated and organized enough to engineer that strong cycle, but the
clone-set voters are too unsophisticated and disorganized to equal rank the
clones.
Additioally, CC, Smith, ICC and MMC are about what happens when voters vote
sincerely. If voters are favorite-burial-prone, and the methd fails FBC,
then you won't need to worry about what happens when voters vote sincerely.
But the fact that ICC isn't a good argument against MMPO doesn't mean that
ICC isn't a valid criterion. Maybe not a necessary criterion. Maybe not one
of the more important criteria.
Abd ul continued:
It is this definition which is the problem. The most important problem with
it is the use of the word "clone." The definition does not adequately
specify the required similarity of voter preference that the word in
ordinary usage implies. Yet the failure of the definition is easily
overlooked
I comment:
No. The inappropriate name "clone set" and "clone criterion" do not mean
that the criterion's definition is a problem, and does not amount to a
failure of the criterion's definition.
Abd ul continued:
, that is, it is easy to think that it is sensible to apply this
definition and the criterion to a fair election method and expect it to
pass.
I comment:
I doubt there's more many on EM who would insist that any method that
doesn't meet ICC is unfair.
Abd ul continued:
And if it fails to pass, there must be some defect in the method.
Logically, however, it is also possible that the defect is with the
definition and criterion.
I comment:
There's no defect with the ICC definition and the ICC criterion. As for a
method that doesn't pass, there's a valid criterion-failure. Whether you
call that a defect in the method depends on how important you consider the
criterion. Myself, I don't call it a defect when a method fails ICC.
Abd ul continued:
Right away, I thought: this is an examination of Approval using information
not expressed in the election.
I comment:
Yes, Approval fails ICC and some other criteria because Approval's ballot
doesn't allow that information to be expressed on the ballot. That's a valid
reason for a method to fail a criterion. Though a criterion shouldn't
mention method rules, including balloting rules, that doesn't mean that a
method's results-failure, whatever it's cause (including ballot limitations)
shouldn't be able to make that method fail a criterion.
Abd ul continued:
Yes, it still seems that the criterion ought
to apply
I comment:
ICC, when worded properly, does apply meaningfullyl to all methods.
Abd ul continued:
, but no election method is going to, in all cases, produce a fair
result that can only be judged as fair based on unexpressed information.
I comment:
The meaning of that independent clause isn't clear, but a method's failure
to let voters express certain information can result in a somewhat "unfair"
result, and some criteria will notice that and
return a failure answer.
Abd ul continued:
The other term is "failure." To fail to satisfy a criterion implies a
defect in a method.
I comment:
That's a strong word. Let's just call it a criterion-failure of the method,
which may or may not be serious enough to call it a defect. If every method
that fails a criterion were defective, then every method would be defective,
because every method fails some criterion or other.
Abd ul continued:
In this case, the "failure" to satisfy the criterion,
if it is to be considered as a defect, might be the defect in the
criterion
I comment:
It "might be", but it isn't.
Abd ul continues:
Instead, it is merely a corollary of a known
*characteristic* of the method
I comment:
Callilng it "a corollary of a known *characteristic* of the method"doesn't
excuse that result, and doesn't make it any less desirable, or any less a
valid failure.
Obviously every result of any method is "a corollary of a known
*characteristic* of the method".
Abd ul continued:
Some of the discussion of election methods on this list and elsewhere is
rather divorced from practical considerations and consequences. This is
appropriate to a degree, but I've mentioned before that, pedagogically, it
leaves a great deal to be desired. It is difficult to learn arbitrary
information. If Election Methods experts would hope to become a priesthood
with esoteric knowledge not available to the common person, hoping somehow
that the ignorant public would then consult them, proceeding in this way
will help the cause.
I comment:
No one on EM is trying to keep information from Abd ul. He himself is doing
that. He comes along, new to the subject, and starts asserting where we've
all been wrong all these years. Perhaps if Abd ul weren't so busy asserting,
he'd give himself the opportunity to find that information that he says
is difficult to obtain. If there's information that Abd ul considers
unavailable, then he should be asking questions instead of making
assertions.
Please note that there's nothing wrong with tellling EM how it has been
wrong all these years. I myself often do so. But what I'm saying is that Abd
ul is doing so carelessly. And that the information that he says is
difficult to get would better be gotten by asking questions than by making
assertions.
As for the public, we often talk about the great importance of saying things
clearly for the public. Brief definitions, free of terms that most people
don't know or won't like.
Abd ul said:
Applying
this to the present situation, if the language is confusing, no matter how
accurate it may be to a specialist, it makes it much more difficult to get
that elusive "sense of rightness" that typically accompanies a correct
understanding, leaving the reader with a sense of confusion *even if the
text has actually been understood correctly.*
I comment:
Quite so. For the public,we try to word things briefly and clearly. For EM
we aren't always so brief, because sometimes precision conflicts with
brevity.
Abd ul continued:
As defined, it [Approval] does indeed fail [ICC]l, but, again, this is not
a
"failure" in the ordinary sense.
I comment:
Yes, it is. It's a failure in the ordinary sense.
The IICC that I defined in my reply to James, in reply to James' suggestion
for such a criterion, is apparently a valid criterion, and maybe a useful
one. ICC is more demanding. A criterion can be more demanding without being
"defective" or lacking validity, meaning, or usefulness.
Approval's failure of ICC says something about Approval, a desirable
property that Approval lacks. I still think Approval is one of the best
methods (along with CR, MMPO and wv).
Abd ul continued:
It would be more neutral to say something
like "Approval disregards the ICC criterion
I comment:
Approval doesn't "disregard" ICC. Approval fails ICC, for meaningful and
valid reasons involving Approval's results.
Abd ul continues:
, which was designed for ranked
ballots
I comment:
The ICC that we've been defining, the ICC that Abd ul is referring to, was
not designed for ranked ballots. It was designed to apply to and compare all
methods, which it does. Approval fails that criterion.
Abd ul continues:
, but satisfies a more specific criterion using definitions arguably
more appropriate to approval ballots."
I comment:
Criteria with different definitions for different methods aren't useful at
all. If a method fails a criterion, that doesn't mean that that criterion
isn't appropriate for that method, and that a less demanding criterion would
be more appropriate for that method.
Different criteria make different distinctions. Approval fails ICC, and that
distinguishes Approval from BeatpathWinner. But criteria that Approval
meets, though not more appropriate because Approval meets them, are useful
for making other comparisons and distinctions. IICC distinguishes Approval
from Plurality, as ICC distinguishes BeatpathWinner from Approval.
Of course Approval meets other criteria, such as FBC. FBC is not met by
BeatpathWinner. Approval also meets WDSC, which distinguishes Approval from
Plurality, IRV, and Borda.
Mike Ossipoff
_________________________________________________________________
On the road to retirement? Check out MSN Life Events for advice on how to
get there! http://lifeevents.msn.com/category.aspx?cid=Retirement
----
Election-methods mailing list - see http://electorama.com/em for list info
