Assessing the ultra-central flow puzzle in the Bayesian era

An outstanding problem in heavy-ion collisions is the inability for models to accurately describe ultra-central experimental flow data, despite that being precisely the regime where a hydrodynamic description is most applicable. We reassess the status of this puzzle by computing the flow in ultra-central collisions obtained from multiple recent Bayesian models that were tuned to various observables in different collision systems at typical centralities. While central data can now be described with better accuracy than in previous calculations, tension with experimental observation remains and worsens as one goes to ultra-central collisions. Tuning the model parameters cannot remove this tension without destroying the fit at other centralities. Our results show that the ultra-central flow puzzle cannot be resolved by state-of-the-art simulations, suggesting that modifications are needed in the standard modeling of heavy-ion collisions.