Re: Update

[Dave Crocker <dhc@dcrocker.net>]

Joel,

First, I certainly agree that there is nothing risky about making the
change.  And I share the subjective sense of comfort about it.

My point is that we have no empirical basis for believing that it
makes any real difference.

As an example hypothesis, note that there could rather easily be an
experience cliff after attendance of 1 meeting.  That is, if you
attend more than one meeting within a reasonable short time, them
perhaps you are likely to attend many.

So upping from 2 to 3 would make no real difference.

To repeat:  This is just an example of a pattern that would render the
proposed change utterly meaningless.  Without looking at the data more
carefully, we simply do not know anything about the meaningfulness of
the change, other than that it keeps the pool of nominees roughly the
same size.

And, of course, this all continues to ignore the small matter of validating
the IETF process knowledge base of that pool.

d/

ps.  Given that a few weeks ago, my own use of subjective assessment
for similar analysis was rather thoroughly dismissed, it is odd to see
it being applied here.

Thursday, January 23, 2003, 11:05:18 AM, you wrote:
JMH> Actually Dave, we do have one piece of evidence that the experience base is
JMH> different.  The old rule guaranteed that the person had been to two
JMH> meetings.  The new rule guarantees that the person has been to three meetings.
JMH> (I don't care whether in the most recent data sample all of the 2 out of 3
JMH> people had actually attended 5 meetings.  The rules change still shifts the
JMH> base assumptions one can make about a nomcom participant.)

JMH> Most of us believe that this helps the chances of the nomcom participant
JMH> being better.
JMH> Yours,
JMH> Joel M. Halpern

JMH> At 10:47 AM 1/23/2003 -0600, Dave Crocker wrote:
>>Thursday, January 23, 2003, 10:39:15 AM, you wrote:
>> >> In contrast with the theoretical analysis, the empirical numbers we
>> >> have been given show no serious difference.
>>
>>HTA> No serious difference in the *number* of eligible people.
>>HTA> Which addressed the concern that some had that the pool would shrink too
>>HTA> much, but tell us nothing about who those people are.
>>
>>ok.  at the least that means that we do not have any basis that the
>>experience base is signficantly different.  hence it means that there
>>is no basis that the pool of nomcom participants is better (or worse).



JMH> ----------------------------------------------------------------
JMH> To subscribe or unsubscribe from this elist use the subscription
JMH> manager: <http://lists.elistx.com/subscribe>



d/
-- 
 Dave <mailto:dhc@dcrocker.net>
 Brandenburg InternetWorking <http://www.brandenburg.com>
 t +1.408.246.8253; f +1.408.850.1850