I'm submitting a ...
Tracks from pandora are commonly flipped; meaning that start point is assigned the end point and vise-versa. This is ostensibly viewed by looking at the true vs. reco costheta direction. This was observed for kaons (docdb 34967) and muons (docdb 41650).
Checklist
Expected Behavior
Costheta 2d histogram should be diagonal, no flipping between +- costeta value.
Current Behavior
Off-diagonal indicative of flipped tracks.
Possible Solution (optional)
Incorperate better charge information or delta direction information into track direction reconstruction.