Comparison of extracted parameters for the blinded synthetic data test. All methods are blurred to match the ground truth GRMHD, blurred to 20 μas before the parameters are estimated. The markers show the results for each method. For the two Bayesian methods, Comrade and THEMIS, the markers denote the median, and the error bars the 95% credible interval. Since the synthetic data only consider a single frequency, the non-Bayesian methods only produce a single image. The dashed black line shows the ground truth, estimated from the true on-sky image blurred with a 20 μas Gaussian beam.