Hello Everybody,

I have been thinking about ways to enlarge the generic MC sample, 
without doubling the generic production. 

Should we consider a sample of BB events in which
   one B decays generically
   the second B decays semileptonically (all decay modes)

This would allow us to check backgrounds from
 - all other semileptonic B decays
 - from the second B in the event, combinatorics,
   either Breco tagged analyses  or untagged analyses.

For one of our core analyses, B to D tau nu, the statistical errors 
of the MC are 
                     B+-        B0             
      PDF shape     11.5%      4.5%
      Constraints     4.2%      6.1%
      Total MC stat 12.2%      7.6%
compared to total systematic errors of  15.6% and 9.4%.
So, by more than doubling the MC sample, we could reduce this
error by a factor of 1.5, in parallel to improvements in our 
understanding of the background modeling.
                  
Also analyses with semileptonic tags, the other B could serve for
background assessment, for instance in b to s gamma and 
other rare decays. 

With 2x11% prompt semileptonic decays, plus 2x2.5% with tau decays,
we could get a factor 4 for 500 million events generated.

Another issue is, can we get more continuum MC, uds 2x for
muon fake background, and may be also more ccbar since this gives 
real leptons.

We are asked to make requests by tomorrow,
Shouldn't we discuss this?

Ciao
Vera



============================================
Vera Luth
SLAC  - Stanford University
2575 Sandhill Road, Menlo Park, CA 94025 
Phone:  650 926 2702   FAX  650 926 2657
============================================

-----Original Message-----
From: [log in to unmask] [mailto:[log in to unmask]] On Behalf Of Roberto Sacco
Sent: Monday, November 03, 2008 8:34 AM
To: vub-recoil
Subject: generic samples + NO meeting tomorrow

Dear all,

we are not going to meet tomorrow, as we are still working on what was  
decided during the collaboration meeting.

Regarding the generic MC samples, let me remind you that, in our  
published analysis, the uncertainty due to limited MC statistics is  
our second highest, at around 4%, depending on phase space cuts.  
Increasing the generic MC statistics by a factor 4 could bring this  
uncertainty down to about 2%. We expect the new mES fit uncertainty to  
be, too, at the 2% level. Would you agree with requesting a factor 4  
increase? Do we want more?

Cheers,

Roberto