[EM] Reply to Weinstein on Approval in Long Beach

Craig Carey research at ijs.co.nz
Thu Aug 1 12:02:32 PDT 2002

Since APPROVAL was being advocated for Public Elections, I give
a portrayal of an Approval Ballot paper.

Multiwinner Approval Ballot Paper exhibit 1.
You may fill in none or some or all of the checkboxes.

   [ ] Candidate 01
   [ ] Candidate 02
   [ ] Candidate 03
   [ ] Candidate 04
   [ ] Candidate 05
   [ ] Candidate 06
   [ ] Candidate 07
   [ ] Candidate 08
   [ ] Candidate 09
   [ ] Candidate 10

   [ ] Candidate 11
   [ ] Candidate 12
   [ ] Candidate 13
   [ ] Candidate 14
   [ ] Candidate 15
   [ ] Candidate 16
   [ ] Candidate 17
   [ ] Candidate 18
   [ ] Candidate 19
   [ ] Candidate 20

   [ ] Candidate 21
   [ ] Candidate 22
   [ ] Candidate 23
   [ ] Candidate 24
   [ ] Candidate 25
   [ ] Candidate 26
   [ ] Candidate 27
   [ ] Candidate 28
   [ ] Candidate 29
   [ ] Candidate 30

   [ ] Candidate 31
   [ ] Candidate 32
   [ ] Candidate 33
   [ ] Candidate 34
   [ ] Candidate 35
   [ ] Candidate 36
   [ ] Candidate 37
   [ ] Candidate 38
   [ ] Candidate 39
   [ ] Candidate 40

   [ ] Candidate 41
   [ ] Candidate 42
   [ ] Candidate 43
   [ ] Candidate 44
   [ ] Candidate 45
   [ ] Candidate 46
   [ ] Candidate 47
   [ ] Candidate 48
   [ ] Candidate 49
   [ ] Candidate 50

   [ ] Candidate 51
   [ ] Candidate 52
   [ ] Candidate 53
   [ ] Candidate 54
   [ ] Candidate 55
   [ ] Candidate 56
   [ ] Candidate 57
   [ ] Candidate 58
   [ ] Candidate 59
   [ ] Candidate 60

   [ ] Candidate 61
   [ ] Candidate 62
   [ ] Candidate 63
   [ ] Reform League wanting 6 member juries


I need not say that Approval is multiwinner but it is to be assumed
so in the usual style of mathematics that regard things as being
symbols and constraints not existing unless they are actually

Obviously a big election there. An aim would be to use just one
vote if there is a "favourite". Mike Ossipoff always used to say
that a favourite is the man or women that you (the critic) want.
Actually, I can't rememeber the Ossipoff belief well enough to say
if it was the favourite that had to be human, or the critic.

If all goes well, then Joe Weinstein will tell us Hadamard theory.
He will want some time before he starts to write on that topic.
Hadamard theoretic containment of the entropic decay that could result
from undue philosophicalism, and thus save Long Beach voters from
trialling out Approval (when an Alternative is a smorgasboard
of CVD options).

Here is the founder of the society for electing the wrong candidate

At 02\07\31 18:55 -0700 Wednesday, Forest Simmons wrote:
 >On Wed, 31 Jul 2002, Joe Weinstein wrote in part:
 >> Each method has its pros and cons.  These are
 >> being discussed daily in an international internet forum called
 >> 'election-methods-list'.

This could be a list for the public since Mr M was arguing at times,
for mobbish responses (e.g. to claims by me that definitions are
unrepairable [to the last, they are also not repaired, too].

 >That's a good way to put it.
 >One might also mention affiliation with the Election Methods Advisory
 >Council either at this point or later in the letter, perhaps after the
 >(excellent) introduction to Approval.

The Election Methods List is a mailing list that can figure out less
in any given 4 year interval than a mathematician can figure out in a


At 2002\07\31 15:05 -0700 Wednesday, Joe Weinstein wrote:
 >Thanks, Alex!  In addition, consider the following points.

The document's argument gave a vision of Approval being used in very small
elections. As the elections get bigger, until as large as public elections,
then testing (using no good criteria) may be unsatisfactory and better
may be to reject the Approval method.

 >(1) Here's perhaps the FOREMOST point to bear in mind and address.
 >Unfortunately, supporters of IRV often misleadingly present their method as
 >the ONLY alternative method that anyone has ever conceived.
 >open discussion and clear description of election problems, and realistic
 >local testing of leading alternative methods which actually solve the

Testing is not realistic if it is not done on paper or a computer since
too few points are involved. Surely at least 1,000 points are needed.
It could be tricky to blindly test the so called IRV method (with 3
candidates) and find a grave defect using <=1,000 sample points.

Why should the results from the test be "local" (of a region in USA) while
there was no theoretical testing?

 >(2)  What systematic election method - or at any rate evaluation method - is
 >now most used in the USA?  Arguably, it's what we may call Academic Voting.
 >It's used annually  - by millions of voters (namely teachers) applied to
 >millions of candidates (namely students):  teacher assigns student one of
 >the grades 0,1,2,3,4 (=F,D,C,B,A).  An Academic 'election' results whenever
 >- in some group of students - we seek a student with highest grade average.
 >Academic is one 'grading' (also called 'cardinal ratings') method.  Other
 >grading methods used include the simplest possible - Approval or Pass-Fail
 >(grades 0 = 'fail' and 1 = 'pass'); and Olympic (grades 0 = 'total fail'
 >through 10 = 'perfect').

Does that make it safe for me to give a grade on Joe's statements of
principles that compar Approval with all of the many methods that
correctly keep the power of a paper in the range, 0 to 1 ?.

What about this: an analogy to Olympic Approval

    Each university marker passes or fails a paper, with a pass awarding
    the number +10/(20-[th mark received so far]), and a fail awarding
    the negative of that.
    Each paper can be marked by upto 10 markers, but the markers can
    only be those that were voted for by the student using an
    Approval Ballot ballot paper.

That is a double use of Approval & a modified Approval.
It could be harder to argue away since the externals provide data more
than once.

 >(3) In particular - and here there is widespread agreement among many
 >students and scholars of election methods - Approval is the one true reform
 >method which is far and away simplest to implement using our familiar
 >ballots and available computers.

Still on Approval in schools, coded up in [TRS 80?] BASIC, perhaps.

Approval can be written in just a single line of J or APL.
At 2000\10\20 08:19 +1300 Friday, Craig Carey wrote:
 >http://www.jsoftware.com/  [fixed URL]
 >A reference manual for J's operators/vocabulary:

The "its simple" arguments is in use already by the CVD (even in Long

Joes arguments are not about large elections and those details (what,
the whole message of Joe?), should and will be set aside as being

Also, there may be no agreement at all amongst scholars that says
that Approval is good enough to use. Joe wrote this:

    "here there is widespread agreement among many students and
     scholars of election methods"

Approval is a topic probably not even regarded highly by the abstract-
higher-luminaries that edit journals and books, on recent advances in
social decision theory. A list of some citations showing Approval
is well regarded by living people, might be asked for, except that
the use of the word "here" suggests that Long Beach was region that
help the academics.


Instead the topic can be regarded as being like grammar: polticians
have rules and largely they do not know what they are: the rules
can be identified and then written down. Doing that (might be
easy) and it easily could lead to arguments that a public would find
more persuasive.


More thoughts from Richard:

At 2002\07\31 22:51 -0700 Wednesday, Richard Moore wrote:
 >That's a strange defense. It was after all Craig who made the
 >accusation that small numbers of candidates were being used in
 >examples to cover up some problem with approval.

I give an example of a large Approval ballot paper, above in this
message. You wrote on how to manipulate a multiwinner Approval
election but seemed to miss the topic, since you did not write on
how a group would manipulate the election where the number of
front runners was not small.

Craig Carey 

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

More information about the Election-Methods mailing list