[EM] Who can't solve 2 candidate elections

[Markus Schulze]

Dear Craig,

you did not answer my question why --despite of the fact that
FPP satisfies e.g. anonymity, neutrality, non-dictatorship,
Pareto, strategyproofness, monotonicity, participation,
consistency, and resolvability in the 2-candidate 1-winner
case-- FPP does not "solve" the 2-candidate 1-winner case
in your opinion. (Believe me: I know what FPP is.)

If you fear that someone could misuse the above paragraph
to promote FPP, you can replace the term "FPP" with "IFPP",
"Borda", "MinMax", "Ranked Pairs", "Beatpath", "River", "IRV",
"Coombs", "Copeland", "Kemeny-Young", "Top-2 Runoff" or any
other method that is identical to FPP in the 2-candidate
1-winner case.

Markus Schulze

> It is normal for me to describe a mistake that you should
> avoid, and then you simply make that mistake I tried to get
> you to avoid.
>
> > To answer my question why FPP does _not_ "solve" the 2-candidate
> > 1-winner case, you would have to mention a desirable criterion
> > that is _not_ satisfied by FPP in the 2-candidate 1-winner case.
>
> It is untrue to say that you can't infer the best solution until
> I write something.
>
> It is very fishy that you request a flawlessly designed definition
> from me, immediately after you dopped into the communications,
> some intensely-useless ideas like Pareto or non-dictatorial.
>
> You seem to be asking for irrelevant information.
>
> That is just a fact and your refusal to know is something that
> I and others would like to ask about: How could you have written
> so much while never learning what First Past the Post is ?.
>
> If Monte Carlo testing of your answer is done, then we would
> find that your solution FAILS to return an answer, for
> over 99,9% of the probes done of it.
>
> It could be that you didn't use reasoning, or you lacked
> axioms. I certainly didn't see any reasoning so we might as
> well close down this challenge and the finding was that you
> failed.
>
> Let me get some conclusions:
>
>  * you quote dumb rules of economists/etc., and fortunately you
>    did not harass me/us with extremely badly worded English
>    definitions of the things.
>
>  * Of the billions of methods possible, you clubbed your way
>    through to the First Past The Post method.
>
>    I was asking for axioms and reasoning, and I missed out on
>   that in respect of the por-FPP/FPTP decision.
>
>    A Much better method is one that handles (AB) papers but not
>    (BA) papers. That method malfunctions on only 1 paper.
>    Your FPP/FPTP malfunctions on 2 papers.
>
>    I actually asked for inference and you seem to show up with
>    a belief that First Past The Post is a good answer, as if
>    you had spent over 5 years *believing* that FPP/FPTP is an
>    heavy-weight important preferential voting method. I didn't
>    ask for statement of carefully selected beliefs.
>
>    If, similarly, you copy from Mr Kenneth May, then a fail
>    mark would be appropriate: his argument was not great.
>
>  * Too much of the reasoning was secret.
>
>  * You completely ignored your "strictly prefer" summation idea.
>    Maybe it is rubbish when there are 3 candidates. Certainly it
>    is incompatible with monotonicity. Yu seem to have written
>    the only academic article that states a false view on that.
>
>    Now we have 5 papers and you suggest for them, only THREE
>    weights: +1, -1, 0, for counts of the papers (A), (B), and (),
>    respectively. I wait for your clarifications on the objectives
>    of the agenda of not disclosing 5 weighting numbers. Doesn't
>    the 3 candidate case extrapolate down into the 2 candidate
>    case ?, with a result that you can design the method asked for?.
>
>  * You asked for an axiom I had. What about this:
>
>        Axioms the number of winners be correct for all
>         election (points).
>
>    However you didn't provide the answer I asked for; a method
>    that solves elections.
>
>    I did leave it wide open for you to produce a totally wrong
>    answer. Now that you did produce a wrong answer (with no
>    reasoning for it) the challenge can be closed.
>
> All of those Condorcet ideas that were copied from history, 
> were completely ignored by you: it seemed that you only purpose
> was copy -- firstly wrong "criteria" and then after that, to
> copy a completely wrong solution.
>
> There was no use of inference.
>
>   G. A. Craig Carey, Auckland, New Zealand