Re: [81attendees] diverse meeting locations (was: are we getting complacent? Good job!)

[John C Klensin <[e-mail]>]

--On Monday, August 08, 2011 20:41 +0000 Curtis Villamizar
<[e-mail]> wrote:

> There are many things for IAOC to consider.  It is worth
> noting that attendance does not seem to be affected by
> location:  Quebec 1127, Prague 1229, Beijing 1207, Maastricht
> 1192, Anaheim 1248, Hiroshima 1152, Stockholm 1124, San
> Francisco 1186, Minneapolis 962, Dublin 1183, Philadelphia
> 1174, Vancouver 1128, Chicago 1175, Prague 1193, San Diego
> 1245, Montreal 1257, ...

Curtis,

Just in the interest of good statistical practice, the only
thing total attendance figures are good for is the P/L
statement.  That is legitimate and important to the IAOC, but
has little to do with effective IETF functioning (and that
assumes that there are not significant number of day passes or
other cheap admissions in those totals).    Otherwise, one needs
to look at attendance from outside a particular area, the
long-term average attendance of people who happen to be from
that area, and the "attending because it is close or a
curiosity" numbers separately and think about the effects on
IETF effectiveness.  I am not, for the record, claiming that any
of those groups are unimportant... they are, however, different.

To illustrate how that works, suppose Anaheim included 250
locals who hadn't been at an IETF meeting before or since and
Beijing including mostly hard-core repeat attendees, whether
from China or not.  That would make Beijing a relatively high
attendance meeting and Anaheim a relatively low attendance one
as far as getting IETF work done is concerned, despite the
apparent relationships of the figures above.  It would also make
Minneapolis look close to the others in attendance if there
wasn't a large local contingent who don't otherwise attend
there. Note that I'm not claiming those thing happened with any
of the three meetings -- I haven't seen, much less thought
about, the numbers and have no reason to believe any of those
hypotheses are true.   But it is the sort of thing we need to
consider if we are trying to evaluate numbers.

It is also difficult to calibrate the meaning of all of the
usual whining when long-term projections are concerned.  There
are at least some number of people who are sufficiently
committed to their work in the IETF that, if a very unattractive
(to them or their organizations) location is picked, they will
complain and attend (not necessarily in that order), so we count
them in the attendance figures and decide the meeting was an
attendance-success.  But, if we provide them with several
unattractive locations in a row, maybe they (or their
organizations) reconsider participation in the IETF.  The odds
of figuring out that situation is occurring from either
attendance figures or quickly-assembled surveys is very low; by
the time the effects are actually felt, it is too late.

Unless your point is that almost nothing we do is likely to make
more than about 300 or 400 difference in total attendees no
matter what we do and where we go.  If so, I'd tend to agree
with that, at least within very broad limits, but note that a
drop of a few hundred attendees for several meetings in a row
could have really nasty effects on the budget.

    john