 The paper proposes a technique to reduce the risk of overconfident estimates of model parameters by averaging over multiple models in two examples, selecting the rank of the state in tomography and selecting the model for the fidelity decay curve in randomized benchmarking. This article was authored by Christopher Ferry.