[EM] CVD wants Alt.V to be fairer but it isn't: misleading website

[Craig Carey]

Sorry this message is a mechanistic reply to Schulze


Now is a good time to ask that major question:

Who that can still write, regards truncation resistance as being
of sufficient importance that it should be strictly held.

Mr Ossipoff recently sort of said that monotonicity was of major
importance. I'd disagree and say that these 3 properties are
more important:

(1) The method finds the required number of winners

(2) When the papers are like STV's, the outcome can be reproduced by
     spreading that papers into FPTP papers ('limitation of power')

(3) Truncation resistance: for each candidate, altering preferences
     after (and not including) the candidate's preference makes no
     difference to that candidate's win-lose status.

Then another aim is monotonicity. But if it is to be imposed at the
same time as truncation resistance, it is plainer and better to
impose only P1 (which is similar to (3), except that the candidate's
own preference can be altered and the "no difference" has to be
changed into "can't change from a loser into a winner").



At 22:16 01.10.00 +0200 Sunday, Markus Schulze wrote:
 >Dear Craig,
 >
 >you wrote (2 Oct 2000):
 >> The IRV method has a problem which is that a vote for
 >> a candidate makes a candidate lose.
 >
 >I thought that IRV was your favorite method. What is
 >currently your favorite method?
 >

I won't call the Alternative Vote, "IRV". There is a still a lot of
people inside of the United States of America that call the
Alternative Vote the Alternative Vote. It is well know that the
English language will be corrupted inside the USA as faster rate
than occurs elsewhere.

I did have an opinon at the end of 1999 which was that I would
opt for the Alternative Vote and STV on the grounds that the set
of adequately explicitly defined better methods was empty.

(I'd ignore pairwise fixes to STV and I'd probably regard Hare-STV vs
Droop-STV as dispute with no clear winner. Certainly it ought be
worthwhile to make the transfer value the smaller of 1 and the
surplus divided by the smaller denominator (I have forgotten the
details) [European Irish elections].)



 >By the way: I don't promote IRV. But in so far as every

You promote some worse method. I don't recall there being a lot
of interest. If it is a Condorcet variant then it is either
not truncation resistant, and/or, not monotonic. Last I read,
those vital tests were to be ignored and some beat path test was
called for.


 >election method has some problems, it makes no sense to
 >criticize an election method simply because it has some


The CVD was aiming to mislead the public. It is simple enough:
the CVD is promoting a method that multiplies people's votes by a
negative weight. The reason for that was not online.
That fact of that was not mentioned online. Each vote reversal
corresponds to a possible court case leading to a constitutional
issue. The CVD advocating the reduction of humans. Again no
reasoning. The can produce reasoning and put it online.
Politicians do not take advice from war criminals in distant
lands so why should the CVD be respected when it is time to
hint a plan to negate the weights of voters' votes.




 >problems. The unique senseful way to criticize an
 >election method is to propose another election method
 >and to explain why this other election method is better.
 >

Fixing the Alternative Vote results in a simple method. If
I am handling the maths then it might not seem so simple.

The British Electoral Reform Society could fast track
solutions using computer aided guessing and have the Alt.Vote
method tweaked.


You talk about senseful criticism. I had sensible criticism:
I set out a proposition that FPTP is better than the
Alternative Vote because it passed more of the listed more
essential rules than the Alternative Vote. They both passed
the 3 above, but only FPTP passes the P1 test which is the
4th in this small sequence.

Writers to this list seem to suggest that the air is packed
with rules and we can all call them "criteria". If a method
is failed by a criteria, etc., etc.


 >By the way: You haven't yet defined your IFPP method
 >for every possible situation. Therefore it is


The explicit definition for sure, but it is not obvious that
the implicit definition (axioms) is actually incomplete, is it
Mr Schulze. I suppose it is.
The definitions are online at: http://www.ijs.co.nz/ifpp.htm


...

So Mr Schulze is uninterested in the CVD website?.

The method the CVD wants to promote is fixable: let them tell
us why it has not been fixed. Instead Mr Schulze tried to say
it was my problem. The data indicates that the CVD is possibly
unaware of a problem, or hiding it. Perhaps the topic of the
CVD could be abandoned and we could write about Great Britain?.

What is the use of the word "properties"?. What about my
duality rule?. It is not quite proven. The Approval Vote and
the FPTP method get past that. I suppose SSD/Skhulze  doesn't/

Why does Mr Markus Schulze uphold the rejection of
monotonicity or monotonicity and truncation resistance, ?.

He too wuld have voters and their lifes altered adversely at the
whim or error of a bad method that was not monotonic?.
I don't suppose anybody wants to know why that is Mr Schulze's
stance or view.



 >Markus Schulze
 >schulze at sol.physik.tu-berlin.de
 >schulze at math.tu-berlin.de
 >markusschulze at planet-interkom.de