The Seven Deadly Sins of Averaging
Stefan Scholtes and I began corresponding in 1999 when he started using my textbook and spreadsheet software with his MBA students at the Judge Business School at Cambridge University. We then worked together on executive education programs for several clients at Cambridge. This ultimately led to collaboration on an approach to managing uncertainty we now call Probability Management.1
Before teaching together, Stefan and I had individually taught the Flaw of Averages to thousands of students and had done it in very similar ways. Compared to explaining something like calculus, it’s like shooting fish in a barrel. The students walk into class, and you ask a few simple questions about average outcomes, which they all get wrong. Then you explain how the statistician drowned in the river and the drunk got killed on the highway, and a hundred faces light up with comprehension. A few days later the students are running simulations and explaining the stuff to others.
During one of our early executive courses together, we had just gone over a simple example of the Flaw of Averages when suddenly one of the participants got up and excused himself. During the next break, a colleague of the executive who had left explained that our example had exposed the folly of a deal the fellow had gotten into the day before, and he was now on his cell phone trying to get out of it.
After class, Stefan and I went off for our customary postclass discussion ...