Re: [council] Rules for Convention-Style Voting

[Amadeu Abril i Abril <Amadeu@nominalia.com>]

Andrew McLaughlin wrote:
> 
ANDREW,

Anadew and NC,

I am afraid I have to propose some increased complexity (but not much)
to your simple, clean, proposal.

I am definitely in favour of this multiple round method, as I have
said, and to keep it as simple as possible, close to the conevention
style. BUT I have to insit that, as in Dennis' and Jviers' models each
NC member be allsoed to cast more than one vote, at least during the
first round)s=.

As even those who are not at all sure to be there for the eletion
(like me) are experienciung, lots of pressure form consituency,
regional groups etc is being put on each NC member. It is very likely
that the NC members will vote, specially in the first round, when
NOTHING· is to be really decided, along strict consituency or regional
deviding rules. Registrars NC reps will be asked to vote for the
registrar nominee (if any) and if there is not, along geobrapical
lines; if there is a nominee supported by the IPC constiuecny then IPC
reps are likely to vote for their candidate, and so on. REven if they
are convinced that, given the NC overall composition their canidate
has no chances on the long run.

This will prevn4t each member of the NC to really put forward her
preferred nominee's slate. And, what is really relevant, this will
most probably come to the fatal disadvantage of those canidates that,
comingf form the GA, have no strong support within one given
consituency (think about Nii, Karl, Alejandro, Kent...) but perhaps
have a good level of acceptance among many consitu4ncies. Thoml ańsp
that within the NC we do have six members frm Norht America, plus two
who live abraod. Five Europeans (and a half, counting Ted(. Thre form
Latin America and two and a half (with Richard form AP. But none form
Afirca. So, again in my supposition, Nii coyuld be first choice form
no oen (espcially is he is not"officially" supported by a consituency)
but alos perhaps, and only perhaps he is the candidate with borader
support outside Euroepeans and Northa Americans... Sorry to sue real
names to illustrate these possiblilites, but I guess you all get the
flavour. Without wasting much time, all NC members would have a clear
idea of the relative supports of each candidate, including the ones
they wsupport, but we would not "lose" upfornt many strong
cross-consituency candidates

[please be aware that our main problem is not preventing automatic
majorities form imposing all three directors, as this is far form
being a bi-patisan eledcto5ate. Qutie the contrary, the most  likely
situation is ultrafragmentation of votes among consituencies and
regions, so we have to devise mechainism that allow efficinet
convergence into strong candidates. Aggregation, in owe word]

This is why I propose to allow NC members to cast multiple votes in
the first round. My idea would be three on the first round, therefore
eliminating all those with zero to two votes (ie, those with no
support or very marginal support.

[to be clear: no cumulative voting on the same candiate: the two or
three votes have to be casted for differnt candiadtes or are lost]

Thereofre:

> 
>      1.  The list of Director nominees is presented to the NC.
>      2.  Each NC member votes for one nominee.

AAA: Two at least; perhaps three [My preferred
e option].

>      3.  The nominees are ranked by the number of votes received, and the NC
> is informed of the result.
>      4.  The lowest vote-getting nominee is eliminated from the list.  [For
> the first round, all nominees receiving zero votes *and* the lowest
> vote-getting nominee are eliminated from the list.]

AAA: all those with zero, one or two votes are eliminated. This menas,
all those failing to at least be one of the three preferred candiadtes
for at at least one consituency (strict minimum).

>      5.  The procedure is repeated with the new (reduced) list.

AAA: we have to make clear that in case that after a given round only
candidates form two regions are running, we will "put again" into the
ballot the candidate form a thrid region which has gathered more votes
up to that moment. If not, we could losse the geographical diversity
requiremn4t half the process (as an alternative, we could simply go
on, and take again all those nominees form regions thre four and five
after electing the first tow  directors).

AAA: If, and only if, there is still a "large· number of candidates
(say: more than two candidates form at least three regions) we could
run a second round, with only "two" vbotes per NC member. And
eliminating those with less votes. But I pressume that this won't be necessary...

AAA: In any case, we should "kkep`" all the nominees pasing thru this
first (or second) round. The reason is geographical diversity. As we
wil voting along Andrew's rpospal (one vote each) in successive
rounds, we are likely to "lose" most, if not all the nominees from 
the minoritarian regions (ie, LA, AP and Africa) as the system will
lead to concetrate on the most likely candidate by rounds.... It could
happen that by the time we elect the first member and eliminate all
those form the same region, we would be not able to find nominees
still standing form the other regions.

>      6.  A winner is declared when a nominee receives 10 or more of the NC's
> 19 votes.

AAA: At tyhis moment, all those nominess who are citizens form the
same region as the elected candidate are eliminated. If if necessary,
candidat4es form other regions who have been eleiminated since the
first (or senond round), the one tiewh multiple choices, are
re-instated into the ballot. The we go for the seond and the third.
Indeed we could face situations where this is not necessary. But if we
don't allow for re-consideration of these candidates, no one wil ever
move his "first "choice, as the candidate would be permanently
eliminated, and we could never get the 10 votes needded. We have to
let them to go back to a certain piint of the election, where nominees
can not be voted during a given round (without being permanently eliminated.
> 
> The first seat to be filled will be the 3-year seat;  then the 2-year seat;
> then the 1-year seat.
> 
> In order to reduce the number of rounds of voting, the NC might eliminate
> the *two* lowest vote-getting nominees after each round (until such time as
> three nominees remain, at which point the *one* lowest vote-getting nominee
> is eliminated).

AAA: as I had proposed, a first round  where all NC members have three
votes with all those with less thant hree votes, and s second round
(wiht ttwo votes per NC member applying your rule where perhaps also
three or even four votes would be required to stand would probalby be
enough. Eliminateing to many candiates upfornt would probalby be
counterproductive on the geographical diversity side.


Amadeu