Explore efficacy of recoil electron selection

[oscarplewis]

Issue description

At present, the relevant function ecal::trackProp picks out the recoil electron track to propagate forward into the ECal by simply returning the first valid track state. But this has the potential to erroneously select a track that is not the interacting recoil electron (for instance, the abundance of low pT recoil electrons here), especially in the case of multi-electron events. Moreover, how effective is this existing strategy to accurately find the recoil electron, and how does it compare to the simulation scoring plane selection (which should be the "truth")?

Tasks

• Evaluate efficacy of existing selection function
• Compare to selecting the maximal pT track as the recoil track
• How does discriminating against positive QoP values turn out? Is it worth selecting only negative QoP tracks for picking the electron? (this will require checking the error rate in the tracking algorithm at which the electron tracks are mischaracterized and assigned positive QoP)

Existing material to start off

• ecal::trackProp function at HEAD
• Possible pT-max-finding function with QoP discrimination