This shows the results of matching the two electrons from the Z' to two superclusters. Taking the energy of that supercluster over the energy of the generator level electron. This was done with the hope of finding a more linear relation between resolution and pile up. The data does appear to relate in a more linear fashion than the mass peak widths. The same trend is observed where resolution is positively related to pile up and |h|.
This data is the result of the same techique as above applied to the mass = 2000 GeV simulation. Here the data does not appear linear at all. It follows a constant within the barrel and a second slightly higher constant within the endcap. The last set of Data points corresponding to (2.5 < |h| < 3.0) is to be ignored due to the very low number of electrons which reach this area.