Some Comments on “Bayes” Estimators