The efficiency calculated in the last post of this study came out to be lower than expected:

https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-03012018-run-9-pp-efficiency-test

We anticipated it level out at about 85~90% for pT > 2 or so (granted, it is for only one bin of partonic pT).  To investigate this further, I required that the only events used in the calculation have a reconstructed vertex satisfying |Vz| < 55 cm and |Vr| < 2 cm.  No requirement was applied to the MC vertex.  It does indeed raise the tracking efficiency, but it still falls short of the expected 85~90%.  I also checked to see how the pTparton = (7, 9) GeV/c bin looked.

Attached are the corresponding plots for without the vertex cut and before applying QA cuts to the reconstructed tracks.