Even after removing the cut on 'pTreco', the widths of the track resolution distribution still don't look right: the minimum is in the wrong place (~2 GeV/c vs. the expected ~1 GeV/c), and the slope at high pT is way smaller than anticipated.

https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-04292018-run-9-pp-resolution-widths-no-ptreco-cut

So now I've done the calculation without any of our typical QA cuts (nFit, nFit/nPoss, dca, eta, pT).  First, I checked it with requiring 'QaTruth > 50', where 'QaTruth' is the percentage of track fit points due to the best matched MC particle.

And then I did the calculation without a cut on 'QaTruth'.

And here are the width distributions on top of each other:

Any differences between the distributions are negligible, and both still don't look correct...