Over the past decade, it seems that operations research and management science have become preoccupied with the refinement of sophisticated techniques as they apply to well-structured problems. The tendency has been to adjust reality to fit the model. This has led to the practice of defining problems in terms of the solution technique and, consequently, to the recommendation of solutions to the wrong problems. This paper reports an experiment which utilized the technique of operational gaming to facilitate the identification of a problem before developing a solution.