by Stephen M. Millett
It has been long argued that all scenarios should represent possible, not likely, futures and should not be assigned any probabilities. One argument is that the scenario team and managers should consider all the scenarios as equally likely and to prepare for each scenario. In addition, the strategic purpose of scenarios is to think of alternative futures and get away from financial and other types of time-series trend forecasts. On the other hand, there is a counter-argument that says that probabilities can and should be used with scenarios. These probabilities, however, are Bayesian in that they explicitly include assumptions, imagination, and expert judgment and that they go beyond trend projections. This paper reviews both arguments and makes recommendations as to when probabilities are appropriate for scenarios and when they are not.