During the Sept. 13th SPIN phone meeting, Carl asked how many to how many sigma the Delta R < 0.04 cut corresponded.  The short answer at least 2.  The long answer follows...

Dataset

69 runs of the Pythia Monte Carlo sample were used, approx 20 each of the lower 3 partonic pT bins and 10 of the highest partonic pT bin.  The cuts on the generated pi0s were:

• The event passes the vertex Z cut and the trigger cuts
• The pi0 was within 4 < pT < 16 GeV with 0.8 < eta < 2.0

For the reconstructed pi0s, all the current cuts were used.  Plotted is the delta R distribution between all combinations of reconstructed pi0 candidates and generated pi0s.

Result

So a cut of 0.04 corresponds to mean + 2 RMS, taking into account the whole distribution.  The distribution should actually be a sum of a "correct match" plus a "incorrect match distribution" and so the "correct match" sigma (and possibly the mean also?) is most likely smaller than the quoted values, implying the cut most likely corresonds to more than 2 sigma.  One could use the pi0 gun sample to get another estimate of this, if it were really crucial (which it is most likely not).  The pi0 gun samples have not had the EEmcTrees updated to the latest version.