Lawrence Berkeley National Laboratory

We discuss how the different estimates of elliptic flow are influenced by flow fluctuations and nonflow effects. It is explained why the event-plane method yields estimates between the two-particle correlation methods and the multiparticle correlation methods. It is argued that nonflow effects and fluctuations cannot be disentangled without other assumptions. However, we provide equations where, with reasonable assumptions about fluctuations and nonflow, all measured values of elliptic flow converge to a unique mean v_2,PP elliptic flow in the participant plane and, with a Gaussian assumption on eccentricity fluctuations, can be converted to the mean v_2,RP in the reaction plane. Thus, the 20percent spread in observed elliptic flow measurements from different analysis methods is no longer mysterious.