In a recent column in ValuePoint™, I referenced the late Frank Knight and his thoughts on “measurable and unmeasurable” uncertainty, which drew thoughtful comments from my friends and colleagues in the decision analysis community. Here are a few of the observations:
Uncertainty is neither measurable nor unmeasurable. It is not a quantity. It is how we describe the world as we experience it. Uncertainty is not a characteristic of objects or processes; it is a consequence of what we know.
Referring to uncertainty as measurable is not so much an oxymoron but a non sequitur (“it does not follow”); since its not being a quantity, measurability does not apply to uncertainty. There is nothing to measure.
“Measure” would, therefore, appear to be an inappropriate term. Perhaps a better term might be to “quantify” what you know and do not know about an uncertain entity such as the size of a new market or the success of an R&D project. So-called “subjective probability” approaches have been used successfully for several decades or more to assist management in making decisions, particularly in R&D, where “research success is difficult to define and impossible to predict with certainty.” (“Quantifying and Forecasting Exploratory Research Success”, Boschi, Balthasar and Menke, Research Management, September, 1979)
Before I get into hot water for using the term “subjective probability,” I will defer to Professor Ron Howard’s suggestion to drop “subjective” and simply refer to “probability.” Says Dr. Howard in his seminal paper “An Assessment of Decision Analysis” (Operations Research, January-February 1980), “… the only meaning of a probability is a particular individual’s quantitative description of uncertainty; no modifier is necessary.”
The bottom line: the future is uncertain. There are ways to deal with uncertainty that can significantly improve your decision making. You can start with tornado diagrams that will help identify how uncertainty around each variable in your business case affects net present value.