Approval in Texas: they want it (utility) (was Re: [EM] One man, one vote and Approval: Pragmatic Approach

[Craig Carey]

I reply to Mr Ingles at the bottom.



At 2002\07\27 18:38 -0700 Saturday, Rob LeGrand wrote:



 >Craig Carey wrote:
 >> EXAMPLE showing Approval will get rejected by voters (if politicians
 >> pass it):
 >>
 >>    1. Suppose that there are 45 candidates.
 >>    2. The number of seats to be filled equals 3. (3 winners).
 >>    3. Suppose that the latest poll data indicates that 14-20 of the
 >>        candidates running a close contest.
 >
 >I have never seen Approval recommended for multiwinner elections; I advocate

It can be easily extended. Don't forget that it is a mathematical definition
that does not come and go.

 >Approval only for single-winner elections.  Contrary to popular intuition, the
 >number of seats to fill should have nothing to do with the allowed number of
 >votes per voter.  In fact, rather ironically, when considering election
 >methods that allow at most one vote per candidate per voter (like FPTP,
 >Approval and everything in between), allowing multiple votes per voter (i.e.,
 >Approval) is best in the single-winner case and allowing only one vote per

That idea of "best" was directly being disputed by me, yet it is merely
dropped in. It looks like Mr LeGrand is not replying to my message but just
merely writing.

I wrote a 'matter of fact' way; but in this message, after a build-up using
these words:

      "I advocate"

      "Contrary to the popular intuition"

      "ironically"

then: whack!, every argument I wrote is opposed:

      "(i.e. Approval) is best"

Promptly after that a falsehood over the very centre of the topic:
...
 >Each person has the same available options and thus identical power.  It may

I have corrected that twice and referred people out to a
politicians-and-polytopes messages that shows the automated
calculation of the power of Borda.

I didn't both to check Approval because a person would have to be dim
to be elliptic, to imagine that that passed a one man one vote test.
It would take me as much as 0.5 hours to test and reject Approval and
find the power value.

If members want a question: is the greatest power shown by an Approval
paper in a 1 winner 3 candidate election, 2.0, or not ?. I guess it is
2.0 but I did find that Mr Saari's Borda had a power value equal to
2.0 (under the same constraints) and it certainly seems the case that
Approval is worse. By his Borda I mean no more than that the
(A) paper has weights (1,0,0) instead of (2,1,0).

You can all check it yourself if you want.

But to keep lying to people like me and they quit (before even a
possibility of questioning the owner arises).

How did Mr LeGrand ignore these two corrective statements ?.

------------------------------------------------------------
At 02\07\27 22:28 +1200 Saturday, Craig Carey wrote:
...
 > >From: Joe Weinstein
 > >Subject: RE: What are we all about?
 > >Date: Wed, 24 Jul 2002 12:56:42 -0700
...
 > >Rather, in any reasonable interpretation, the rule MEANS 'one person, one
 > >ballot' (of equal inherent power to each other cast ballot).
 >
 >Approval Votes in general do not have equal inherent power. Being true,
 >it can be hard for an Approval advocate to get the defences looking
 >excellent.
 >
------------------------------------------------------------

A second correction of the equality of powers assertion.
Checking for equality of powers is too tough a test.

------------------------------------------------------------
At 02\07\28 12:55 +1200 Sunday, Craig Carey wrote:
...
 >At 02\07\27 09:24 -0700 Saturday, Alex Small wrote:
 > >It has been argued that, since people who approve m candidates may be
 > >exercising a different amount of voting power than those who approve n=/=m
 > >candidates, Approval fails to give all voters equal power.
 > >
 > >Those making the argument are underwhelmed by my assertion that "one man,
 > >one vote" simply means my approval of candidate X should carry the same
 > >weight as somebody else's approval of candidate X (a criterion flunked by
 > >the electoral college, among other non-proportional institutions).
 > >
...
 > >So, if (for the sake of argument) we assume that "one man, one vote"
 > >requires equal voting power per individual, consider this circumstance:
 >...
 >
 >I just corrected that error and it is hardly a small error since just
 >about every kind of paper has differing powers. For a given election,
 >an upper bound on the power of each 2-preference paper could calculated
 >by finding the smallest number of FPTP papers that can reproduce or
 >outdo, its effect. [Outdo can mean that the desire of a paper is
 >calculated by comparing the winners with binary numbers having digits
 >determined by the preferences shown (multiwinner election)].
 >
 >I.e. numbers showing powers of papers (in a given election) can be
 >arrived at and listed. Thus the human rights definition of equal
 >suffrage is to be rejected.
...
------------------------------------------------------------



 >be smart for a given voter to vote for only two candidates, but deciding
 >beforehand to vote for only two is rather silly.  Each voter is responsible
 >for estimating his utility for each possible vote and choosing the best one,
 >whether that means voting for one, two, four or all but one.
...

Empty ideas. Cliches.
Why not this instead of "utility"
A voter/faction in a large public 1 winner election either:
(a) has their candidate win, or
(b) the candidate loses.

In the 1st case, proportionality used (i.e. an idea behind "1+1+1<1+1+1+1").

In the 2nd case, the method itself can come under the most minute
scrutiny.

Rob LeGran is obviously rejecting the 2nd idea of public credibility since
he writes implying that voters could have even a slight interest in "utility".

As far as I know, no reasoning voter anywhere naturally wants to have an
interest in some utility number. You all ought clean out some of the mess
of false claims about utility. There is sure to never occur a time when
"utility [numbers]" lead to anything worth seeking of having.


...
 >Rob LeGrand
 >robl at aggies.org
 >http://www.aggies.org/robl/
 >for Texas State Representative, District 50


Breathtaking: you say that your method passes a test that either no method
can or that only FPTP can (equality of power). There appears to be not the
minutest understanding of what the power of a paper is, but we could find
that the members/Approvalists also can't focus their minds on simple
examples that sets some lower limits on the power of papers.

What's happening is that pro-Condorcet Lanphier is watching the Approvalists
accrete evidence here that they do not know what power is.




At 02\07\27 19:28 -0700 Saturday, Bart Ingles wrote:
 >
 >It seems clear that Craig has not yet grasped the concept of approval
 >voting, and doesn't understand why others might favor it.  Nobody here
 >who advocates approval voting does so for multi-seat elections.  Since
 >Craig's example is for a 3-seat election, the rest of his argument is
 >moot.
...

Thanks Bart. I understood checkboxy Approval well enough to reject it.

PS. at the moment I am being slowly or rapidly being banned from the
Electoral System Designers mailing list of the Canadian. It is a list that
does not seem to ever arrive at a true result. Currently the list owner
is trying to compress one of the members: me. It seems that this mailing
list is elevated over that (despite what members write here).

My Politicians and Polytopes mailing list on research into preferential
voting methods that are well designed and truncation resistant, is
here:

http://groups.yahoo.com/group/politicians-and-polytopes

http://groups.yahoo.com/group/single-transferable-vote/







G. A. Craig Carey
Auckland



----
For more information about this list (subscribe, unsubscribe, FAQ, etc), 
please see http://www.eskimo.com/~robla/em