Counting clusters, e’s and mu’s (with nLive cut included)

Yesterday, I posted a plot of SNR vs. Signal Events for the three cut sets.  Cut Set 1 was clearly the best of those three, so I added Dan’s nLive cut to that one.  I also removed the requirement that the event have at least one cluster, and ended up with a pretty good S/B in some of the 0-cluster, multi-lepton bins.

As before, I’ve counted the clusters, e+, e-, mu+, and mu- in each event, and added up all the bins in which the S/B was above a particular threshold.  By selecting this threshold (and the nLive cut) correctly, we can end up with approximately 20 signal events with a significance of about 0.45:
best.nlive

I suppose the next step might be to integrate Dan’s recent post (surgical cuts on nLive, nPi0, and nTracks, along with charge correlation of clusters and leptons) with this.

-Scott