Let Xk be a sequence of i.i.d. random variables taking values in a finite set, and consider the problem of estimating the law of X1 in a Bayesian framework. We prove that the sequence of posterior distributions satisfies a large deviation principle, and give an explicit expression for the rate...