A Quick and Dirty Response to the Quick and Dirty Crowd; Particularly to Jack Byrd's “The Value of Queueing Theory”

Queueing theory is a useful, important part of the tool bag of OR/MS, and any half serious professional ought to know it. To charge otherwise in a professional journal, on the basis of sending students to visit a two-clerk bookstore, seems to me absurd or malicious—take your choice. In that same bookstore I could prove that linear programming, inventory theory, network analysis, simulation, geometric programming, indeed virtually all of OR/MS is equally useless.

Now, it is really very weird but over the years I've done similar experiments at Columbia with my first year OR students. Do you believe it, but the results were entirely different? In a pretty wide range of problems my students found queueing theory useful—and one application was even to setting red-green cycles on traffic lights. How do you account for the difference—the experimentor, the subjects, the environment? Maybe, but perhaps it's just that we were not quite so quick, or quite so dirty, or trying to be quite so cute.