They call me “Mr. Decision Tree,” and I use this as a pedagogical style.

If I were teaching a more mathematical course, I could have abstracted it into symbols, and the decision tree wouldn’t have loomed large. But in order to communicate with my students, the first thing I say, “Let’s write up a decision tree.” And I became Mr. Decision Tree, and then it became a fetish.

And then I reflected, I said, “Well, why didn’t I do this at Columbia when I taught?” And I went back to what the problems were, and I couldn’t do what I did at the business school at Columbia because that would have involved me in the use of subjective probabilities. So the very fact that I couldn’t use subjective probabilities in the Columbia days meant that I couldn’t exploit the idea of the decision tree.