Abstract
This paper presents a procedure for specifying how long a supervisor or monitor of a process should wait between input samples to maximize a given value or payoff function, assuming the resets the controls with each sample as a function of the best information he has. The procedure is based upon Bayesian preposterior information analysis.