[Mmwg] Important : procedure for runoff election

Thu Feb 16 10:46:57 GMT 2006

Dear Jacqueline,

I am a bit sad. I made a very simple and argumeted proposal for this second
round, suggesting that we allow for two votes instead of one only.
Justification is reiterated below in the copy of my previous mail.

Unfortunately, you are calling for a second round of elections with the
traditional single vote rule, without any mention of the alternative. I
suppose you and others simply did not notice my mail in the flurry of
exchanges. This is OK . I trust you.

Could I ask that at least the proposal repeated below be briefly discussed
among us and that the arguments I put forward be refuted if people do not
want to go this way ? If there is no objection, why not use it ? It is
simple to implement and will produce a clear outcome.

We are at a moment where innovation in procedures is critical and I hope the
members of this group will accept to devote some time to this important
issue and try to test something that might be useful later.

I repeat below my previous mail :

  From: *Bertrand de La Chapelle <bdelachapelle at gmail.com>* Mailed-By: *
gmail.com* To: *Avri Doria <avri at acm.org>*
Cc: *mmwg at wsis-cs.org*
Date: *Feb 8, 2006 5:12 PM*
Subject: *Electing the second co-chair : a Suggestion*

 (...)
On what to do next for the selection of a second co-chair, I am OK for
endorsing Wolfgang if there is a consensus but would prefer a second round
to be set up.

I only identify a potential problem in the latter case. If we are allowed to
cast only one vote, there is no guarantee that anyone will get 2/3 of the
votes. Votes could be distributed evenly among the three remaining
candidates for instance (each one third). Of course we could count on a
momentum in favor of the leading candidate (in this case Wolfgang) and be
out of trouble, but this is not guaranteed by the mechanism. And we should
try to identify mechanisms that automatically produces a predictable result.

I therefore propose the following modality that will guarantee that we have
at least one candidate reaching 2/3 of the votes. We just need to maintain
the possibility to choose 2 names in the remaining list of 3. As there are
only three possibilites to vote (A and B, A and C and B and C if the three
candidates are A, B and C), the whole group will necessarily split into
three clusters (those voting for A and B, those for A and C, and those for B
and C) and two clusters will necessarily represent more than 2/3 of the
total together. Hence, if the two groups forming this 2/3 majority together
have voted for instance respectively for A-B and B-C, candidate B - and
he/she alone - will mathematically enjoy a 2/3 majority.

The only case where this mechanism would not produce a solution is if :
- the number of people casting a vote is a multiple of 3 (any non multiple
is OK)
- AND the group has split evenly among the candidates (ie : combinations
A-B, A-C and B-C have all gathered 1/3 of the members)

In all other cases, the system produces a clear result. And even in the
above case, a rule could be established to attribute the seat to the
candidate that gathered the most votes in the previous round.

So, in order to make myself clear, I suggest :
- that we keep the three candidates
- that we allow a second round with the possibility to cast two votes
- that in case the number of people voting is a multiple of 3 and the votes
are evenly distributed, Wolfgang (who got the most votes in the first round)
is given the seat.

Apart from producing a clear result in all cases I believe this mechanism
would be interesting to explore further and I would be very happy if it
could be tested here.

We will be confronted in the near future with many occurences for selecting
people as chairs, grooup coordinators or members of various committees. I
believe this opens a way to innovate and find modalities that will designate
people that can be endorsed by the broadest majority, avoiding the too
frequent outcome of traditional majority voting that produces victories with
50,01 % of the votes and a split among constituents ("my candidate won,
yours lost" ...).

I hope this helps.

Do not hesitate to correct me if my calculations are wrong.

Best

Bertrand

__________

 Thank you for your comments.

Best

Bertrand

 that the voting procedure (voting for only one candidate) has been adopted
without even the slightest discussion of the very simple proposal I made :
to allow for two votes.

On 2/16/06, Jacqueline Morris <jam at jacquelinemorris.com> wrote:
>
> Dear all
> The runoff election voting system is set up. We will have voting from Feb
> 16-Feb 19, midnight UTC, if that's OK with all of you.
> Please use the same usernames and passwords as sent to you last time by
> Avri.
> the url is:   is: http://www.votingondemand.com/mmwg/
> Candidate bios are available.
> You can vote for 1 candiate, and the person with the most votes wins.
> That's the simplest way, I think. There are no rules in the Charter for
> this.
> Jacqueline
>
>
> Jacqueline Morris
> www.carnivalondenet.com
> T&T Music and videos online
> _______________________________________________
> mmwg mailing list
> mmwg at wsis-cs.org
> http://mailman-new.greennet.org.uk/mailman/listinfo/mmwg
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman-new.greennet.org.uk/pipermail/mmwg/attachments/20060216/2973ea8e/attachment.html