[EM] If you had of written differently then you could have credibility [Ms G.]

Craig Carey research at ijs.co.nz
Sun Nov 28 15:33:37 PST 2004

At 2004-11-29 06:39  Monday, Diana Galletly wrote:
>On Mon, 29 Nov 2004, Craig Carey wrote:
>newsgroups are not accessible (much) beyond the boundaries of the University;
>but in any case they are ucam.*, not cam.*.  cam.* is for the wider population
>of Cambridge and Cambridgeshire.

Got it.

>> Also I say that Mr Schulze should discard all tests and method
>> checking ideals/rules that are not multiwinner.
>Why?  Different methods can be used for different purposes -- does one
>seek a consensus opinion (voting on a proposal), does one wish to represent
>a wide spectrum of views

As far as I can tell there are only 2 theories but one that is acceptable
to politicians. However they may be a bigger selection of approximate
(partly unfair) algorithms.

>> >>     a0 * (A)  +
>> >>     ab * (AB) +
>> >>     ac * (AC) +
>> >>     and other papers.
>> >>
>> >> Then increasing the "ab" Real number will tend to cause B to lose
>> >> since appearing with a positive weight in the Heitzig-ian "A over B"
>> >> total.
>> >
>> >And as far as I can see it's total rubbish, because increasing *any*
>> >of a0, ab, ac etc. is going to cause an increase in the "A over B" total.
>> >
>> Which will tend to harm candidate B and hence the entire method could

   be wrong.
>Why is this unfair?  Those who vote solely for A prefer A to B.  Those who
>vote for A and B only (in that order) prefer A to B.  Those who vote for
>A and C only prefer both of them to B.  This is, one assumes, the voters'
>prerogative -- that they might be able to express their own preferences as
>they wish!

While Truncation Resistance is an axiom, all the 'a*' terms are coalesced
into a single a * (A) term. Thus I ignored what happened to A.

In face I wrote on candidate 'B'.

It is wrong in general, to let 'B' can have a better chance of losing when
(AB) papers are added. I just wrote on this topic when saying that the
students of the Condorcet variant idea, have to get the faces to be
deleted by the polytope expression simplification software.

>You may have a point -- but if you do, I cannot discern it.  Perhaps you would
>like to elucidate?.

Today I wrote on this new idea of getting Mr G-A to rapidly fail 99 worthless
random garbage methods, just before he had to use the exact same tests to
reject his own method. He has a polytope with wrong faces instead of a need
to consider an N**2 matrix and then fail in the attempt.

All methods are built of polytopes and Condorcet variants are too, T get a
single face wrong opens it to a review by a numerical optimizer.

>be worth taking on board; unfortunately you totally fail to present it
>in a fashion that is understandable by others.

Your e-mail address is not subscribed to Politicians-And-Polytopes, and
also not subscribed to the Single-Transferable-Vote mailing list.

>(whereas you presumably do?); unfortunately one of the realities in
>getting funding to do research (at least, here in the UK) is that one
>be able to present one's results in a form that is easily comprehensible
>by others and thus will contribute to the general public understanding.
>You would fail this hurdle; and I find that a shame.

That comment on credibility seems to be a bad guess. It might be
helpful here for you to provide all of your axioms; and show that all
are tested enough.

>indeed the other questions around the issue).  I'm starting to wonder
>what definition of monotonicity you might be using.

If candidate A moves towards the 1st or jumps onto the end, and if
the weight of the paper is positive, then candidate A never changes from
a winner into a loser. Also the changes can be across more than 1 paper,
and also the ()->(A) change is allowed. Scrambling subsequent preferences
is not.

So it is just the plain old monotonicity receiving STV style ballot papers.

>> I don't know where participation is defined. But the best plan is to
>> fully ignore the idea.
>Which idea?  Participation criterion?  Condorcet methods?  Giving the
>voters any form of assurance?

I am not going to start writing on an unneeded "participation [whatever]",

---                                                        Craig Carey

More information about the Election-Methods mailing list