Identifying policies that are most likely to produce a desirable outcome

For some types of stochastic optimization problems, it is possible to directly estimate the location of an optimum on a response surface using an experimental technique based on the notion of simulated hindsight. The idea is to retrospectively solve a related deterministic optimization problem with respect to realizations of the stochastic effects as if the outcomes of all uncertainties were known in advance. We explore issues involved in designing, conducting and analyzing simulation experiments using a variant of the technique that addresses situations where it makes sense to define the best solution as the one that is most likely to produce a desirable outcome.