The runs with the fix are on the left, without on the right. This is without the cut on chi square; I am only changing the actual chi square values and amplitude error values by some scalar amount. My guess is that the chi square values somewhat will alter the way things work because they alter whether you do pileUp or not. The fact that there is no chi square cut for clustering tells me it must be this. I will do a run without altering the amplitude error change (to check that this isn't changing anything as I expect) and another with the chi cut. You can evaluate whether the tracks we are losing matter; if you go to the hit on track section the 1 percent of tracks we lose all have really low hit numbers and are all low momentum (almost not even physics scale low momentum)