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

[Craig Carey]

This to the Election Methods list

Would all PaP subscribers please consider erasing the previous
message titled "CVD wants +16% to make loss: CVD is bad...".

                                 ^\---o---\v


If you spot errors here, please be free to note the errors.
At the bottom of the message I write about my Sheepskin Gumboot
Hat awards and I am undecided. Is it free to get advice from the
CVD?.

The topic of the message is the CVD website and the false information
on it.

To be able to follow this message, it may be necessary to read
this message (on the CVD's siding to advocate a stunningly bad
and arbitrary to boot, preferential voting method, which the CVD calls
IRV:

                   PaP message no. 68, on the CVD website
Topic:
CVD; Councillors sheared apart; Article 25 ICCPR rights

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

Shift 1/8-th of the vote to C, from AB, and almost when the full
1/8-th is shifted towards C, candidate C changes from a winner to
a loser. The UK Electoral Reform Society and the CVD either know
that or they are dangerously uninformed of the nature of the
Alternative Vote when there are only 3 papers. The CVD seems to
be holding back this data from their website. Mr Loring is the
webmaster over there. If any has the data on the reasoning behind
keeping the public much in the dark, I'd be interested to get
that data. I guess the CVD keeps the politicians in the dark.
Some politicians have well done websites thoroughly within their
own areas of competence. The CVD has opinions and pointillistic
data that could end up to be really of no value at shifting
opinions from bad to better (and away from the IRV method).

In 15 Oct 1999 I wrote to Mr Loring and I got not reply. Mail
failure perhaps. I told him some of his ideas were not worth
me knowing about and my P1 was something something he should
be aware of. The whole matter can be understood with formula.
The webmaster has not put mathematics onto the website. What
do they think that preferential voting is?. A mathematician
knows that a large number of opinions can add up to little.
New Zealand is fortunate in not having any Centre for Democracy
and Voting. I used this mailing address: LoringRbt at aol.com.

-------------------------------------------------------------------


The CVD is a group that promotes THE arbitrary preferential voting
method named the Alternative Vote . The CVD renamed the method to
"Instant Runoff Vote". That shortens to IRV.

The CVD has a base in Washington, apparently.

http://www.fairvote.org/about_us/index.html

The CVD is has got a website that says in many places the
"IRV" preferential voting method is fair. This is what this message
is about.

Here is one page with "CVD" authorship, that has the CVD claiming
that the "IRV" method is "fair":

http://www.fairvote.org/irv/a_fairer_way.htm

CVD's IRV pages: http://www.fairvote.org/irv/index.html


Note that the title is that IRV is "A Fairer Way", not that it is
fair. My comments below are made with the assumption that the CVD
would call IRV fair. It is not fair since it is in places worse
than the ideal (the ideal bans (AB{C+)-(C{B+)), and it is also in
places worse than FPTP. FPTP also does not exhibit (AB{C+)-(C{B+)).

[See PaP #68 URL, CVD promotes idea that 1/6th vote gain should
(i.e. does) cause a defeat, in the Alternative Vote (='IRV')].

I think that the CVD is providing false information website by
writing that IRV is fair.

The CVD website says that IRV is fair when it is not. The CVD
does not say exactly for regions and paper counts, IRV is fair.
Without such qualification people would interpret the truth of the
statement to be the validity of the statement, i.e. the claim
of fairness is false if false anywhere.

If the CVD wants to say that IRV is overall fair, then how could
that be executed?.

If I had to defence the CVD I would admit error and repair the
website and change the statements that IRV is fair to statements
that IRV is not fair. At least that would get the CVD website
in agreement with the true facts.

The CVD might be in a mood to say that the "IRV" method (which
is a multidimensional Boolean function) can be said to be fair
overall. A problem with that is that the idea of "overall-ness"
is an attempt to call false true. It doesn't seem right when the
method is defined by a multidimensional Boolean function. The
results do not contain real numbers or probability.

The CVD may have no defence if it claims that it made a true
and fair statement when saying IRV is fair, if its defence is
just that fairness is undefined. Unfortunately for the CVD if this
defence were made, IRV is indeed unfair at some points. See my
comment at the 'Politicians and Polytopes' list.

The CVD is describing an arbitrary fixable and badly designed
method as being fair.

Why was the name changed from Alternative Vote to IRV?. But that
is a distraction from my topic of the CVD putting online a rather
misleading website. It could actually make people believe that
IRV is the best method. The method will flip a candidate from
being a winner, into a loser if the candidate gains 16.6% of the
votes from other candidates. Why?. The CVD didn't seem to have
a section on that.

Has the CVD ever analyzed the 3 paper problem?. Surely they
have known about the defects in the example I gave and they
quite possibly have some way to tell this list how to not just
lose 16.5% of the vote, but also how to negate its influence.
Why does the CVD promote that ?. If a judge ruled that one side
had won and then told the wrong side to pay the damages and
court costs, it would seem unfair. I am undecided on whether or
not the CVD has some knowledge of the IRV method; i.e. whether
or not it was aware of the 16.6% problems. I didn't see proof
of an awareness of the precise nature of the defects in the
Alternative Vote, at the CVD website.

The IRV method has a problem which is that a vote for a candidate
makes a candidate lose.

Perhaps the CVD could answer this simple question: is it wrong
to alter STV so as to increase the wastage of votes transferred
away from a loser and reduce the wastage of votes transferred
away from a winner.
It seems to be reasonable to do that. Then, since the IRV method
is 1-winner STV, we'd have increased vote wastage for winners,
and the very minimum wastage that would fix the CVD's own
non-monotoncity defects. There is a disquieting aspect which is
that the CVD is stating its own beliefs (which a touch false)
and the CVD could be a candidate for the NZ mathematicians' and
preferential voting theorists' Sheepskin Gumboot Hat award (for
1999). But first I need to find out if the CVD wants to keep on
promoting the arbitrary Alternative Vote method online.

I haven't yet seen any data on the CVD success. There is a
problem with how to define a success, when the event is the
attempted conversion of a local government authority, from FPTP
to the Alternative Vote. One method is (locally/infinitesimally)
fair and the other isn't.

I am collecting pictures of woolly gumboots. I can distort them
to a seeming conical shape to match the CVD's notorious
phrenological endowments.