The German loser (Re: [EM] Two Paramount Criteria
Craig Carey
research at ijs.co.nz
Sat Dec 20 21:46:02 PST 2003
I wrote this for the mailing list.
Currently I am not sending it there.
I have not observed anything good at that mailing list result from
e-mail (except Richard and MrCatchpole seemed to have an ability to
get influenced. I may have missed bits.).
________________________________________________________________________
To: election-methods at electorama.com
Subject: Is this actually true enough Marcus ? (Re: [EM] Two Paramount Criteria
At 2003-12-20 11:05 +0100 Saturday, Markus Schulze wrote:
...
>Craig Carey wrote (20 Dec 2003):
...
>> (1) The number of winners should be correct.
>> indicates that the Schulze method found the wrong number of winners.
...
>In Section 3 of my paper, I prove that my method is well defined:
>http://groups.yahoo.com/group/election-methods-list/files/nmciswem.pdf
>
The words "well-defined" are not defined in the paper.
Why don't you use your neutrality word.
I reject the idea that the arguments in your paper are correct. There are
a lot of serious problems ion your article.
Each of the errors could terminate an interest in you VM17 article.
In particular your ideal of:
* missing out simple important tests
* purporting to consider overly advanced rules, i.e. monotonicity
* avoiding the mathematics of flats which is always avoided by you
but probably never avoided by a person designing a preferential
voting method but aiming to not expose a govt. public to unfairness
if the method would be used.
* having unfollowable arguments
* claiming a pass occurred when that appears to be something that
seems to be suspicious
* failing to define the algorithm of the article
* dropping in some anti-public nearly-ignored doctrines from Condorcet
thinkers writings
In short, a method that won't have a place in history.
I made a possibly false asumption on how to interpret your ambiguity that
you put into the article (presumably because you were trying to censor
out a mention of the ghastly meritless unfair unjustified pairwise comparing
basis of it) and converted it to a a logic form. It was far too complex to
analyze precisely yet ti plain at the end that the number of winners it
found would probably be wrong.
>You wrote (20 Dec 2003):
>> (2) The method should not have some bias. This fails methods that ignrore
>> the votes and pick the first (not best) candidate on a list the method
>> receives.
>>
>> Shulze's believed the count of the papers could be ignored and the number
[i.e. Mr Shulze or Mr Shulze's PDF article]
>> of voters could be counted. That is stupid and some people don't seem to
>> be able to stop or explain themselves without pointing invalid assumptions.
>
>You will have to rephrase this, because I have absolutely no idea what
>you are talking about.
>
In the last message I replied to, Mr Forest Simmons had
votes generated by random numbers. In the Shulze article, "strictly prefer"
has the number of voters counted. So "Strictly prefer" is applying to
voters.
All the claims of goodness must have been a lie, because:
(a) you proof of goodness is done but it used voters (suppose so)
(b) Simmon's random numbers are used;
(c) now your proofs can't conclude anything yet the numbers and the
method is the same
(d) If God was going to hit you with a lightening bolt if you had of
used the word "Vote" instead of "Voter" then please do tell me
about it. In the absence of similar, I'll assume you can't even
think with a truthfulness that is up to the task of correctly
distinguishing between a cat, rat, dog, photo of a man, and bit of
paper.
(e) I will say that I do not understand. First I must ask if the
audience wanted the false claim that voters voted. However the
arguments and proofs are not all true. I am not studying the
text, since your purpose is to advance unfairness.
-----------------------------------
At 2003-12-19 06:37 +1300 Friday, Craig Carey wrote:
...
:
: In the grand sweep of Mr Schulzes exposition, he designed the
: algorithm in the paper so the input is never votes. The preprocessing
: stage that proves that the method is stupid, got censored out. It
[using the Condorcet thing]
: is the case that reasoning indicates that Schulze's method should
: be rejected instead of tested.
:
:
: Let's consider the algorithm itself. Here I quote from the VM 17
: PDF file:
:
: ---
: : Suppose that d[X,Y] is the number of voters who
: : strictly prefer candidate X to candidate Y. Then the
: : Smith set is the smallest non-empty set of candidates
: : with d[A,B] > d[B,A] for each candidate B outside this
: : set.
: ---
:
: The English text of Schulze probably seems to imply this
: interpretation:
:
: * Let there be only 3 candidates. Let the ballot papers be these:
:
: 2 (CA)
: 5 (CAB)
: 1 (CBA)
:
: What is Mr Schulze's d[A,B] ?.
:
: We could look at the computer algorithm source code but it takes
: the "d" matrix as an input. The vote counting algorithm can't actually
: accept votes.
:
: The d[A,B] value would be:
:
: (Interpretation 1) the number 7, or
: (Interpretation 2) the number 5, or
: (Interpretation 3) the number 6, or
: (Interpretation 4) the number 4, or
:
>You wrote (20 Dec 2003):
>> (3) When the papers are like STV' the winners ought be insensitive to the
>> presence or absence of the very last preference in one or more papers.
>> That Schulze method seemed to fail this test.
>
>Please give a concrete example.
>
Not very bright response.
A gauntlet for me perhaps. Your paper is so bad that it should
be you alone who does research. It is strange how you have so
very little interest in the actual words that you wrote. I guess
you don't care what you write. If so then there should be a check
to see if you are depressed. Also it would be useful to know if
you get more depressed, e.g. when criticised.
Mike is criticising you now.
I have already written the answer: it is that you didn't actually
define the method. That fact proves that your "well-definedness"
proof is fake. You clearly gave the wrong meaning to the English
words. hat is obviously so, it was actually referring to some
crap from the pairwise comparing religion.
--
I would have produced a computer program and prove that Schulze was
a failure. However there is no algorithm to test.
There are so many different correct criteria for rejecting your
paper. Designing while blind will be one of them. That does not
produce hope but instead it lowers the hope of the algorithm being
improved.
--
The Schulze paper arrived D.O.A (Dead on Arrival) at London --
since it didn't define "strictly prefer" well enough and with only
a reference to ballot paper instead of voters. Some places could
have 1/3 votes for prisoners (or fugitives from computational
symbolic algebra)
The STV community may be slow to absorb Mr Schulzes paper, using
this line of argument:
* Mr Schulze must have been engtangled in complexity
* The thing has to be built up and then it fails fundamental tests.
The designer asks readers to test the thing.
* If it had of passed the tests then the complexity would have been
reduced. The designer maybe even failed to learn how extremely
complex the method was.
* Like I said privately, if a good job can't eb done without
light then a switch can be turned on. But Mr Shulze is out of
this world, for clearly such an avoidance of algebra must
prevent the production of good methods. But tehre seems to be
absolutely no complaints about anything from the world of
algebra. A real mathematician can be blocked by a single
issue in algebra. It is like Mr Schulze can shuffle to the
right but it is impossible to rotate 180 degrees and then
move in the opposing direction. Four years of EM and counting
and still not even vague complaints about algebra.
Mr Schulze is starting to say that he does not understand.
That would be criticised by me over time.
Mr Shulze is setting out a new direction: designing while unable
to see everything, and we already have had years of no
achievement. Certainly outside of the world of pairwise comparing,
change can be quick and rapid.
Here is that definedness test that passed a most undefined method:
: 3) Well-Definedness
:
: On first view, it is not clear whether the Schulze
: method is well defined. It seems to be possible that
: candidates disqualify each other in such a manner that
: there is no candidate A with p[A,B] >= p[B,A] for every
: other candidate B. However, the following proof
: demonstrates that path defeats are transitive. That
: means: When candidate A disqualifies candidate B and
: when candidate B disqualifies candidate C, then also
: candidate A disqualifies candidate C.
...
: Case 2: Suppose
: (9b) p[A,B] < p[B,C].
...
: Therefore, the relation defined by p[A,B] > p[B,A] is transitive.
Condorcet seems to have a monastery feel to it. They seemed to
be packed with woman dying with breast cancer, but that is
of course no about the public interest.
It could be a check that the number of winners are right.
I assume that sometimes the method finds 0 winners or 2 winers.
Who would finish up an argument that the number of winners is
correct using the words "[It] is transitive".
Transitivity is some properly requiring that the method is
too unfair (or too lacking in proportionality) for use in
a government election.
---
What seemed most interesting is that after a few easy lines of
transforming the Schulze algorithm into a polytope format,
is appeared that all Condorcet variants must be rejected for
being affected by the presence or absence of the last
preference. I was constrained by having to guess at what
Mr SChulze's definition of the algorithm was.
For at least a year, when Mr Schulze was not writing privately,
he also did not e-mail the polytope formulation of the
algorithm that I had requested.
Had he responded to the information requested, then rapidly
I could have shown that Condorcet idealism leads to an
unacceptable defect that won't be OK in civil or computer science
department, elections.
If Mr Schulze believes that it takes >17.95% of his entire
life before he can repell off that evil called Condorcet, then
do please tell us, for the rest who use logic, the process
could take minutes or something.
German males [yeah] in the Internet have a technique of saying
that they do not understand. This is the 'gnome in the rock'
mastery of shooting up opponents in online arguments.
It permits them to not lose steadily in an argument when the
expectation is that they must have precise model of what it
is that the public interest is (constrained by the topic).
Craig Carey
politicians-and-polytopes
More information about the Election-Methods
mailing list