Re: SUMMARY: publish list of nominees?

[Theodore Tso <tytso@mit.edu>]

On Tue, Apr 02, 2002 at 06:58:26PM +0700, Robert Elz wrote:
> The right way for all of this to be done, is that if you (as a member of
> the community) feel that person X would make a good Y, then you ask X if
> they'd be interested (just as a courtesy, and to save everyone unnecessary
> work), and then if they say yes, or even perhaps, you send a message to the
> nomcom and suggest that X should be made a Y.   Then the nomcom can take
> the set of X's they are given, get rid of any they consider hopeless
> (based upon other feedback, etc), ask whatever is left if they'd be
> willing to be considered, do some arm twisting to convince some of them
> they they really should allow themselves to be considered, and then set
> out making a decision based upon what they have been told in the various
> messages from all the different people who thought X would be a good choice,
> and any other input they may have been able to collect from the people they
> do decide to talk to.

The problem with this is that nomcom may not necessary hear about why
X may be a bad choice.  

For example, suppose we have a case where X has managed to antagonise
50% of the working group chairs in a particular area, but the other
25% of the working group chairs thing that X is wonderful, and nomcom
gets glowing, positive reviews about X.  (Assume that 25% have no opinion.)

Now let us assume that the randomly selected nomcom members don't have
much expertise in that particular area, or at least not much
first-hand experience with the recent politics and personality
conflicts in that particular area, so they don't know about these
issues.  How is nomcom supposed to find out?

One approach might be to send out queries to all of the working group
chairs in the area, plus other significant people.  But if you're
going to do that, you come close to making the list public anyway.  So
why not making it public at that point?

Another approach might be to wait until the confirming body gets the
list (and hopefully someone at the confirming body level will know
about the problem about why X shouldn't be selected).  At that point,
the confirming body, since it is not privy to the discussions made at
the nomcom level, can start asking questions of the form: "Did you
consult X?  Did you consult Y?"  But at this point, it's *very* late
in the process, and by the time the slate is submitted to the
confirming body, the selected candidates know this fact, since it's at
this point that we ask them for permission from their management.  So
between the time factor, and the embarassment factor of telling the
selected candidate, "oops, sorry, we changed our minds", by the time
it gets to the confirming body level, it really is very, very, very
difficult to make a change --- even in a case where nomcom were to
find out some significant new information which might have changed the
outcome if they had known it earlier in the process.

So that's why I believe publishing the list is useful.  If the
community doesn't know that nomcom might be considering person X, how
should people who have negative concerns about someone know that they
should be submitting those concerns to nomcom.  In fact, if we only
rely on people who submit candidates to say nice things about them,
I'd argue this is EVEN MORE LIKELY to result in popularity contents!

> While I'm ranting about nomcom practices let me also say that I don't
> much like the practice of the nomcom sedning out questionaires to
> nominees, in which they ask the nominee to explain why they'd be a
> good person to be appointed, etc.    We don't need to be getting
> people to tell the nomcom all their good points, and why they're
> perfect for the job - that should come from those who suggest them
> (other than in the case of a self-nomination of course, but that just
> falls out of the method).   Asking whether the nominee has the time,
> resources, etc, and even (perhaps) what they'd like to do in the role
> is OK, but I don't believe the nomcom should be soliciting self praise.

It's not just asking for the good points, but it's also about trying
to find out information about the candidates philosophy, etc.  That
being said, it is true that if you select only for people who can fill
out questionaires well, you may not end up with people who can
actually do the job well.  So that is a concern.

On the other hand, part of the problem is that the input from the
community is very often not very substantive.  People who currently
submit X for position Y, very often only give a paragraph at most of
explanation.  All other information about candidate X needs to be
actively solicited, either from the candidate him/herself, or from
community members.  

(And again --- free hint to people who are interested in sending
comments to a nomcom.  Details and explanations about why you feel a
particular way is *critical*.  Simply saying "he's a good guy", or
"you must be joking" is not very useful from the point of view of a
nomcom member how is trying to sift through input and trying to
compare multiple candidates.  Even something as short as "doesn't play
well with others", is very hard to evaluate if you don't give
examples.  Some people do take a lot of time and give the nomcom very
thoughtful answers.  But often, nomcom gets little input from the
community unless it is actively solicited, and it is sometimes very
brief and not very useful when we do get that input.)

						- Ted