One thing that's been dogging us -- for quite some time now -- has been this odd dip in the reconstructed charged track pT spectrum when we use charged hadrons as triggers. See, for instance, the plots here:

https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-05172019-run-9-pp-checking-efficiency-when-using-different-nfit-cut-and-when-usin

There is a large dip in the reconstructed tracks in the region pTmc = 9 - 20 GeV/c. Could it be due to the QA cuts we usually apply to tracks? To investigate this, I plotted the various track quantities we cut on for all matched recoil tracks (blue) vs. matched recoil tracks with pTmc > 8 GeV/c (red). "Matched" means that a given reconstructed track has been successfully matched back to a generated particle.

Nothing here really leaps out at me... Almost all of the ratios are flat, except for maybe the DCA...