Text this: The d-risk of Bayesian estimation for the probability of success in Bernoulli trials