[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ga] nomination procedures



On Tue, Nov 16, 1999 at 11:36:41PM -0500, Jonathan Weinberg wrote:
> At 07:39 AM 11/16/99 -0800, Kent Crispin wrote (with respect to proposals
> that the GA, instead of sending to the NC *all* names with the support of
> 10 endorsers, adopt a procedure under which it sends over only the X names
> with the most support):
> 
> >On the other side, what practical benefit do we gain by being
> >restrictive? I don't see any.  From my perspective it just seems like
> >further sustenance for the "we vs them" engram...
> 
> It would be desirable, I think, for the GA chair to have strong support in
> the GA.  A problem with the "send over everybody with 10 endorsers"
> approach, it seems to me, is that it doesn't meet that criterion:  It's
> relatively easy for a person to get the minimum number of endorsers without
> regard to the breadth or depth of his support in the GA as a whole.  The
> advantage of the "send over the X names with the most support" approach is
> that everyone on the list will have a somewhat stronger level of support
> within the GA.

That doesn't necessarily follow.  It only follows when you have an
authentic electorate, which we don't have. 

> (The level of support gets weaker, although the likelihood
> that the NC will approve the procedure gets stronger, as X increases.)
> 
> 	John Klensin, for whom I have great respect, urges that we simply can't
> get it together at this point to have voting, given the potential noise as
> to which list members are not actual people, have forfeited the right to
> participate, etc.  I'm disturbed too as to the state of the GA.  But I
> think that we may see much of the same noise even with a "send over
> everybody with N endorsers" approach.

The primary purpose of the "send everybody with N endorsers" approach
is not into reduce noise -- it is to be sure that qualified
candidates don't get lost in a flood of fake support for bad
candidates.  It will, I believe, also have the characteristic of
reducing noise, because it will remove one of the primary incentives
for noise. 

In the extreme, the method you propose would allow Jeff Williams, who
has quietly signed up 50 fake names to the GA list, to carry the day
and get his 5 favorite candidates selected.  (I myself have signed up
a few fake names just in case the votes are needed.)

Given your rule, the NC would have no choice but to elect someone
from the "Jeff Williams 5".  The rule you propose creates an
incentive for such behavior. 

I said "in the extreme", above, and of course, I don't consider that
extreme *too* likely.  But 1) the incentive it creates is bad enough;
and 2) I do think that an election with an unauthenticatable
electorate is a real danger.

> And if the NC is willing to approve
> a procedure under which it can select a chair only from the X names with
> the most support within the GA, I think it's worth it for us to try to find
> those X names, rather than simply letting the disrupters prevail.

You didn't really answer the question: what is the advantage of
restricting the NCs choice? In the last nomination most candidates
got more than the necessary minimum of endorsements, and that excess
was some measure of popularity in the GA.  I agree with you that in
an election for chair of the GA, support of the GA is important, but
I submit that we will get a rough measure of that support, in any
case.

-- 
Kent Crispin                               "Do good, and you'll be
kent@songbird.com                           lonesome." -- Mark Twain