Hoi Daniele, > Since for B->Dpi the time dependence is flat probably the lepton id > was either more efficient or less pure in the first 8 fb-1 Electron ID (and misid) was stable over the run period, this should be limited to muons. > 1) data-MC comparison is bad for B0 in the depleted sample. In particular > look at the pstar spectrum, nneutral and MX. How do you determine the "recoil" B? At the moment (HEAD of CVS), recoilNtp still uses == 2 to check against overlap with the BRECO candidate; I guess that this is not the right thing to do. This will affect any recoil (and depleted/enhanced, since the decision is based on kaons in the recoil) quantity. Cheers, --U.