Hi all, Ric showed me the distributions of the LAT variable for neutrals belonging to the reconstructed B in the mes signal box and in the sidebands. http://www.slac.stanford.edu/~daniele/vub/lmomcut.eps - top one is the signal box - bottom one are the mes sidebands Since the two distributions seem to be different we were curious about checking the possiblity to cut on it in order to clean the breco sample. I studied the problem finding the neutral with the smallest LAT among the ones that belong to the reco'ed B and cutting on its LAT value in each event. In http://www.slac.stanford.edu/~daniele/vub/cutlat.ps you find the mes distributions as a function of the cut for events that pass a loose cut on the lepton in the recoil (in order to see the actual improvement in statistcal significance for our measurement). As you see, the gain in purity is small. On the other hand the lost of yield is large. We can conclude that this cut does not help. Daniele