Re: Robonson wins [...]
At 03\04\21 16:46 +0200 Monday, Matthias Urlichs wrote:
>Hi,
>
>> He's one of /THEM/ and, no, I'm not permitted to go into details.
>
>Well, if we're having THAT kind of discussion, there's a couple of quotes
>which have been censored from the *CENSORED* report which prove that
>they have sucxiqz567$%&
>NO CARRIER
>
>
It would be nice down here if reader expected a method to be fair
even if silent on the existing polytope.
I lost time writing a reply to Mr Urlichs.
Regarding whether Condorcet is monotonic, I have an answer below.
I have drafted response to Mr Urlich's last message and I am not
posting it in here. Perhaps Mr Urlichs would try harder to not
ignore the polytope of the Condorcet method when appearing to
defend it. He suggested a principle of Martin being preferred.
The ballot paper
44444
is a special paper that gives the voter a power equal to 50,000 times
the power of all other ballot papers. Only Mr Urlichs knows that.
The method can be used to elect the leader of the Debian project,
but due to DCMA encryption issues it a widely trusted black box.
(No one wrote in to this list to complain, e.g. since posting up new
research articles on preferential voting to one of my mailing lists
on how to win and be a right winner rather than A WRONG WINNER,
Martin, except that members here have no individual rights. That
social contract is a bit thin, but I guess that upload access
rights are the currency of the day).
Lets have the argument of the Condorcet [liker] that explains why it
is wrong to use a winner-choosing polytope that gives over 40,000
times too much influence to some single voter who is also a
contestant.
Suppose the method is extremely unfair but that is known from a
consideration of its polytope defining where candidate A (say) wins.
Condorcet passes. How irrelevant to this message?. Anyway what was
the test on the polytope that it passed ?.
Mr U said that more persons preferred Martin than others.
That is checking only at one single election point. Since it not
imposing any constraint on a surface of a polytope then it is not
imposing rights and/or power-limitations on the smallest number
of papers or people.
A similar idea is the how it is OK to let the GNU GCC compiler
get slower and less efficient each time an upload occurs.
Here is text on the Debian Social Contract webpage:
+
+ The Debian Free Software Guidelines (DFSG)
+
+ [5] No Discrimination Against Persons or Groups
+
+ The license must not discriminate against any person or group of
+ persons.
+
http://www.debian.org/social_contract
Any instance of non-monotonicity would be a discrimination. The whole
election system now has to be shut since it discrimates unfairly and
it will negative support rises which is significant in ensuring that
fairly often, the winner was the WRONG PERSON.
Condercet advocates do not like polytopes and the allow through the
gates totally corrupt preferential voting method polytopes is not
enough to change an opinion. I have not seen the Debian pro-Condorcet
theorists. I wonder if it is anybody I known?.
I was wondering if there is a culture of unfairness.
|
| Let the papers be these (and there is 1 winner and exactly 3 candidates):
|
| a(AB)
| b(B)
| c(C)
|
| "A<B": a<b
| "B<C": a+b<c
| "C<A": c<a
|
| The undefined 'paradox' region is (a<b)(a+b<c)(c<a). For the Alternative
| Vote, C wins in there and thus the method is not monotonic. But for
| Condorcet, the winner is not defined in the region, and avoiding probing
| in there results in NO finding of Non-monotonicity.
|
| The method can be failed for getting the number of winners wrong.
| Suppose that was to be fixed. Then B loses in the A wins region which is
| (b<a)(c<a). Then cast a B-loses shadow using the monotonicity rule
| saying that B continues to lose when (B) papers are changed into (C)
| papers.
|
| A diagram of that is here: http://www.ijs.co.nz/images/quota-31-2.gif
|
| That pulled the unknown region away from the A-wins region, and
| brought the C-wins region up to touch the A-wins region.
|
| Now the C-wins shadow is cast using the monotonicity rule:
|
| * C keeps winning when (AB) papers are changed into (C) papers.
|
| A diagram of that is here: http://www.ijs.co.nz/images/quota-31-3.gif
|
| Truncation resistance actually says this (but truncation resistance
| is assumed so (CAB) can be changed to (C) since the considerations is
| only about candidate C.
|
| C keeps winning when (AB) papers are changed into (CAB) papers.
|
| No government election would reject truncation resistance.
|
| Condorcet is no better than the CVD's so called IRV method.
| They share the first appearing major problem.
|
| The leader previously won until named with the 1st preference, and
| sometimes loses.
|
| That is this:
| (a) Candidate C wins with the paper (AB), but
| (b) Candidate C can lose with paper (C).
|
| I name that the First Preference Loser (FPL) rule: The tilt of the
| slope is
|
| A fixed Condorcet method will never be fair for all elections
| provided that it respects the Condorcet winner.
|
| The term the "the bible is the word of god" is another phrase that
| is hard to get clarified up as being believed or rejected.
|
| The Condorcet method is never able to fixed.
Is it true that Debian Condorcet can't solve all 0 winner and all 1 candidate
elections. It seems just about impossible it would get more complex
election problems correctly solved. Still I invite readers to all believe
otherwise.
Also another thing about the Social Contract is the refusal to hide
bugs.
What webpage should I browse to if I alone were to be kept up to date with
the new perfectly fair preferential voting method that this Debian project
would be creating ?. I guess there is nothing.
If I had a dog, I am sure it would get a better deal that being offered
the opposite of what it wanted. In the world of the CVD (the CVD IRV
lobbying few), a trick is to make it minimally obvious that that
dropping of the promises at election time is nether spotted nor covered
by any guarantee of quality.
Craig C
Reply to: