[EM] CVD IRV elections halve and negate voters' votes.

[Craig Carey]

I guess the hillbillies still use FTPT.


At 22:21 09.10.00 +1300 Monday, Craig Carey wrote:
...
>http://www.fairvote.org/library/brochure/pdf/irv_text.pdf
>
>"Is IRV constitutional?
>
>Yes, IRV is a constitutional voting system that upholds the U.S.
>Constitution's principles of one person - one vote and majority
>rule."
>
>   Fascinating, but false too. IRV is a formula that is failed by

Correction: IRV may pass the -1<=power<=+1 test I give.

>   that test and apparently without the CVD or any following its
>   advice benefiting from that. We need to understand that "one
>   person, one vote" means that the power p, must satisfy 0<=p,
>   rather than -1<=p<=1 (which seems to me to be a right
>   interpretation of those 4 words provided that monotonicity is
>   imposed at the same time).


Calculate the most negative power in the Alternative Vote for the
papers AB, B, C

    AB  a
    B   b
    C   c

Suppose (a <= b <= S/3),  S = a+b+c
Case 1 winner

The power of paper X is able to be approximately defined as k,
where k*eps is the smallest voter change of another paper that
will offset a change of eps in the quantity of the paper in
question.

The surface exhibiting a defect in the Alternative Vote is
part of the line, Z:
   Z = ((b - S/3) = k*(c - S/3)).

That line divides the B and C win regions. Z is clipped by
(a<S/3)(c<S/3).

In FPTP, k = 1, and Z = (b=c). A has too few votes to shift the
winner from B-wins to/from C-wins.


In IFPP, k = 0, and the divide Z is st. Z = (b = S/3)

In STV/AV/IRV, the divide Z is given by Z=(a=b)

   <--> (S-b-c = b)
   <--> (2b = S-c)
   <--> (2b - 2S/3 = S-c - 2S/3)
   <--> (2b - 2S/3 = S/3 - c)
   <--> (b - S/3 = (-1/2)*(c - S/3))

    /-------------------------------------------------------------\
    |                                                             |
    |  The power of a voter in the IRV method promoted by the     |
    |  Council of Victorian Democracy or whatever the acronym     |
    |  stands for, can drop down to minus one half. With First    |
    |  Past the Post (FPTP), the power never dips below plus one. |
    |  QED. As required: FPTP is a better method for modern       |
    |  open libertarian justice seeking individual respecting     |
    |  governments, etc. [Any comments on the algebra's validity. |
    |                                                             |
    \-------------------------------------------------------------/


That certainly suggests that in an STV or Alternative Vote election,
the power of a voter can drop down to minus one half. That is some
way short of minus one, so there is no hint of the old principle of
"one man, one vote" being violated by STV. I presume STV passes the
test for all numbers of candidates and winners, and I certainly
have not proved this.

The 12.5% example (a 16.666% example exists too). The nuclear
winter after the CVD had "success". Candidate C managed to shift
1249 votes and managed to effect a defeat. The moral in this for
the 'green' candidates is that you need to be nice to your
opponents [certainly not often].

At 21:14 01.10.00 +1300 Sunday, Craig Carey wrote [to the P&P list]:
...
 > |             before    after
 > |
 > |       AB     3,749    2,500
 > |       B      2,501    2,501
 >:|       C      3,750    4,999   : total 1249 votes cause loss for C

http://www.egroups.com/message/politicians-and-polytopes/68

Back to the CVD think tank. Africa with holding back before taking
advice from the CVD. How long before they come up with better methods?.
Constitutional?. There goes some court's reputation.

Why doesn't the CVD approach all the councils it advocates its new
Choice Vote method to, and its newly named IRV method to, and just tell
them that it made a mistake.

If you are an important politician around Washington, just be thankful
it is only -0.5: the CVD doesn't seem to actually have any principles
limiting the negativeness of that number. Absence of principle might be
a passing US fashion, but of course once councils get in a wrong method
it might get stuck in there. Does Rob Richie et all want to give to this
mailing list a quick perk-up on why Article 25 ICCPR rights of US
citizens amongst the first causalties after the CVD and the councils have
jointly pulled the loo chain on justice and the protection of
individuals?.

What actually happens when US states and cities phone the CVD asking for
a copy of the computer program that they used to help them shape the
Choice method. I guess the CVD might peep out a comment like "the man
doesn't want to release the software". Some councils may interpret that
as a declaration of complete war. This is a failure of the CVD to have
an algorithm (product) that is accessible to idealistic cities, regions
and councils and whatever. They say fairness is great. I inform the
readers that -0.5 were changed to -0.6, the 3 paper method would be
even fairer in a global sense (and less fair to candidates). Global
fairness is so hard to define right that they are surely using an idea
that cannot be understood.

The first big problem with the Alternative Vote is that a voter with
a candidate that loses and has the first preference, can be made to
win by remarking the paper. That is absolutely not desirable and who
would dispute it. (That statement is not strong enough to make it good
rule.)

How long would it take the CVD website to evidence comprehension that
IRV is flawed?. years?, days?, they already knew?. It can't be the
latter: that would imply the CVD had mathematicians as members. I want
to be quite plain: I just can't believe that the CVD has mathematicians
on board. They would cast overboard in the highly politicised central
command style Mr Ossipoff [I 'aint answering these ever] Ossipoff hinted
at.

This fix for IRV/AV seems to be to just drop in quotas (esp. for losers).
There is a difficulty that people may not be able to successfully guess
how to do it right, i.e. how to make the method pass my P1 rule. The
final formula would be simple.

   "I praised my opponent because he was in the minus half paradox zone."
   "That was the easiest campaign I ever fought and we fought dirty"

Is there anybody else that wants to criticize the CDV?.