How Not to Measure a False QCD Critical Point

Fluctuations of conserved charges are a golden channel for measuring a QCD critical point in relativistic heavy-ion collisions. These fluctuations are quantified by measuring high-order cumulants of baryon-number distributions at a given centrality. Using simulated proton-number cumulants as an example, we discuss how the correlation between particle identification and centrality measurements can distort particle-number distributions. These distortions can easily create large fluctuations in high-order cumulants that might be mistaken for a critical-point signature. We show that certain measurement choices can make the analysis more or less vulnerable to these false signals. We motivate this by considering how the two-dimensional probability space of proton-number versus multiplicity is shaped by analysis choices. We then demonstrate this vulnerability with simulated Au+Au collisions at $\sqrt{s_{NN}}=3.9$ GeV in UrQMD, and two toy models of detector responses to certain classes of events. We explain how an analyzer might observe a false critical signature, and how to avoid doing so, even in a challenging experimental environment.