Title: Discovery reach for generic supersymmetry at the LHC: MT2 versus missing transverse momentum selections for pMSSM searches

Abstract: Different search strategies for supersymmetry have been employed by the LHC general-purpose experiments using early data. As proven by their early results, these strategies are promising, but raise the question of how well they will generalize for the future. We address this question by studying two thousand phenomenological minimal supersymmetric standard model parameter space points that come from a fit to indirect and cosmological data. We examine the 5 sigma discoverability of the points employing a typical ATLAS-type search based on missing transverse momentum (MET), a search based on an optimised MT2 cut and a combination of the two, taking into account standard model backgrounds. The discovery reach of the strategies can depend strongly on the systematic uncertainty in the background, subject to the stringency of the cuts and the details of the background simulation. By combining the MET and MT2 based strategies, with an integrated luminosity of 1 fb^-1 (10 fb^-1) at 7 TeV, 4-8% (42%) of the points are discoverable, depending on the systematic uncertainty on the background. At 14 TeV and with 10 fb^-1, 96% of the points are discoverable. While the majority of points can be discovered by both strategies at sqrt{s} = 14 TeV and with 1 fb^-1, there are some that are left undiscovered by a MET search strategy, but which are discovered by the MT2 strategy, and vice versa, therefore it is essential that one performs both in parallel. We discuss some of the factors that can make points more difficult to observe.