Friedman and Gates—Another Look

Abstract

A lack of consensus surrounds probability estimates provided by the two leading bid models, one by Friedman and the other by Gates. Each model relies on essentially the same information but arrives at different probabilities of submitting the winning bid. A synopsis of both Friedman's and Gates's models is provided, followed by a descriptive chronology of literature dealing with this controversy. To date, attempts to resolve this debate have been unsuccessful in formulating a consensus view of which model supplies the correct probability. Using a third bid model proposed by Carr as a base, sensitivity curves visually demonstrate both similarities and differences in the models resulting from hypothetical shifts in the underlying location and scale parameters of the bid problem. It is clear from these curves that both models are simultaneously correct and incorrect. Friedman's model is theoretically correct, yet the bid problem is incorrectly specified. Gates's model is practically correct, yet the formula is incorrectly specified.