Supersymmetry With Prejudice: Fitting the Wrong Model to LHC Data

TL;DR

This paper investigates how different supersymmetry models can produce similar signals at the LHC, emphasizing the challenge of correctly identifying the underlying model based solely on certain invariant mass distributions.

Contribution

It demonstrates that alternative supersymmetry models can fit the same LHC signals as the correct model, highlighting the difficulty of model discrimination with limited data.

Findings

mAMSB and LVS can be distinguished from CMSSM with 1 fb^-1 of data at 14 TeV

mGMSB cannot be distinguished from CMSSM using end-point kinematics alone

Best-fit points of mGMSB and CMSSM are very similar, complicating experimental discrimination

Abstract

We critically examine interpretations of hypothetical supersymmetric LHC signals, fitting to alternative wrong models of supersymmetry breaking. The signals we consider are some of the most constraining on the sparticle spectrum: invariant mass distributions with edges and end-points from the golden cascade decay chain \tilde{q}_L -> q \chi_2^0 (-> \tilde{l}^{\pm} l^{\mp} q) -> \chi_1^0 l^+ l^- q. We assume a CMSSM point to be the `correct' one, and fit the signals instead to minimal gauge mediated supersymmetry breaking models (mGMSB) with a neutralino quasi-stable lightest supersymmetric particle, minimal anomaly mediation (mAMSB) and large volume string compactification models (LVS). mAMSB and LVS can be unambiguously discriminated against the CMSSM for the parameter point assumed and 1 inverse femtobarn of LHC data at 14 TeV. However, mGMSB would not be discriminated on the basis of the kinematic end-points alone, and would require further, more detailed investigation. The best-fit points of mGMSB and CMSSM look remarkably similar, making experimental discrimination at the LHC appear unlikely by any means.