To test the validity of theoretical models, the predictions they make must be compared with experimental data. Instead of choosing one model out of many to describe mass measurements of zirconium, Bayesian statistics allows the averaging of a variety of models.