Shmi comments on Bayesian exercise