Simple Procedures for Selecting the Best

In this paper we address the problem of nding the simulated system with the best (maximum or minimum) expected performance when the number of alternatives is nite, but large enough that ranking-and-selection procedures may require too much computation to be practical. Our approach is to use the data provided by the rst stage of sampling in a ranking-and-selection procedure to screen out alternatives that are not competitive and thereby avoid the (typically much larger) second-stage sample for these systems. Our procedures represent a compromise between standard ranking-and-selection procedures|that are easy to implement, but can be computationally ineecient|and fully sequential procedures|that can be statistically eecient, but dif-cult to implement. We present a general theory for constructing combined screening and ranking-and-selection procedures, several speciic procedures and a portion of an extensive empirical evaluation.