The Unpredictability of Some Human Choices

Abstract

It is a well known rule of logic that when one tells a story and that story eventually leads to a contradiction then it follows by reductio that the story was an impossible one to begin with. But what element in our story must be rejected as impossible? Is it impossible to place two boxes before a player etc.? Surely not. The most likely candidate for rejection is the assertion which did not amount to the description of an observed fact but which we thought was based on inductive evidence, namely that the predictor is virtually infallible as evidenced by his past performance. (It will turn out in Chapter 15 that it is not the case that the proficiency of the predictor qua predictor could ever be established by any evidence.) We shall deny, then, that our so called predictor has great powers of prediction.