Dear all,

as you probably saw, we are seeing some disagreement between data and MC 
in Qtot in the enriched sample. The plots can be found here
http://costard.lbl.gov/~kerstin/vubunf/ana32/dataMC/dataMC_KLcorr.html

along with the plots for other variables, also split up by (reco'd) B 
charge (and the effect seems to be present for both charged and neutral B.
We see an excess of events with total charge of +1 (and maybe some 
depletion of events with charge 0).

It seems to me that the same effect is present in the plots that are in 
BAD 1383 page 33 (although somewhat better hidden): This plot is 
normalized to the same area (mine has the different components normalized 
as they come out of the mX fit), but if you look at the bin for total 
charge +1, you cannot actually tell how many sigma it is off, the error 
bar is too small. If you multiply the chi2 by the number of bins,
1.5*9 = 13.5 and look at the agreement of the other bins, it does look
like this particular one has a big contribution to the chi2.

Has anyone looked at this more closely and tried to figure out where
this is coming from? Or if not that, if there is any impact on the 
analysis? If so, it would be great if you could point us to that!
I don't really have any explanation why we should see this, especially 
since it seems to be present in the enriched, but not the depleted sample,
so I wonder if this could be a huge fluctuation...

Thanks,
Kerstin