I think we need to have separate steering files for the 2015 kinematic skims versus those fro the 2016 kinematic skims.  By kinematic skims, I mean the FEE, Moller and V0 skims.  Naiively, we would modify the drivers so that the energy and/or momentum cuts are proportional to the beam energy.   However, this destroys the Moller skim, since the opening angle of the mollers in the 2.3 GeV dataset is so small that the clusters get very close to the edge and most of the clusters are unable to pass the energy sum cut.  There may be other issues similar to this that I don't know about, so we should try to be sensitive to such