Figure 1.

Comparison of model-choice accuracy. Model-choice accuracy for models (A–D)M_msBayes, (E–H)M_Ushaped, (I–L)M_Uniform, and (M–P)M_DPP when analyzing data generated under models (A, E, I, and M)M_msBayes, (B, F, J, and N)M_Ushaped, (C, G, K, and O)M_Uniform, and (D, H, L, and P)M_DPP. The unadjusted posterior probability of a single divergence event, based on |τ|=1, from 50,000 posterior estimates are assigned to bins of width 0.05 and plotted against the proportion of replicates in each bin where the truth is |τ|=1.