From cousins@physics.ucla.edu Wed Apr 12 16:19:24 2000
Date: Fri, 31 Mar 2000 09:22:05 -0800 (PST)
From: Bob Cousins 310-825-1928 <cousins@physics.ucla.edu>
To: Giulio D'Agostini <Giulio.Dagostini@roma1.infn.it>
Cc:  [See mailing list in a separate file]
Subject: Re: Conclusions of the CERN CLW

On Fri, 31 Mar 2000, Giulio D'Agostini wrote:

>   A point on which there was a quite general agreement was that F-C is 
>   a mess (see e.g. PRIVATE conclusions by Giunti in hep-ex/0003001)
> 
>   A point on which there was also explicit consensus was about 
>   presenting likelihoods (see discussion session). 

Dear Giulio and others,

I invite you and the others to read the end of my CERN CLW talk writeup
before commenting further on whether or not you think there was general
consensus on my "ten" points.  You will not find your first point above
on my list, nor will you find a point saying there was a consensus in
favor of F-C.  You will find your second point; it was already on my
list I prepared before the workshop, and in fact is in my AJP article of
1995. 

The list which Bill Murray just helpfully broadcast is basically the
list I gave in my oral presentation: what I *hoped* we could agree on.
The list in my writeup is what I believe we *did* have a consensus on.
The list omits the point which Bill mentions reservations about: that
"The problem of upper limits cannot be considered solved without having
a consistent method for two-sided intervals."  Bayesians agree on this,
I would assume (it just means using the same prior for both
calculations), but there was some concern by some others.  So I did not
claim consensus.

The point of my list was twofold:
1) For people actively pushing a point of view, we can stipulate to many
points and therefore stop arguing about them. 
2) For people unfamiliar with many of these issues, it is useful to have
such a list as a starting point.  For example, they will learn and
benefit from the fact that the likelihood by itself is not a pdf in the
unknown parameters. 

Keeping in mind that "consensus" does not imply absolutely unanimous
consent, I do believe that all the points had a consensus.  In fact, the
only point for which I have evidence that the agreement was not
"unanimous" is the first point on my list. 

Best regards,
Bob Cousins