The Certainty of Non-April Insanity

We find ourselves at the midway point of our annual orgy of predictive failures, which for reasons of copyright we shall call "Non-April Insanity." If we are able to view the NCAA basketball tournament with rational eyes, we know that our efforts at bracket-building are no more reliable than a coin flip. But rationality has nothing to do with Non-April Insanity. This time of year creates a wonderful laboratory "in the wild" for studying and reflecting on blind spots in human decision-making.

The essence of a good prediction is that it actually predicts at better than a 50/50 rate. So bracket-building is not really about prediction. For that you would be better served to check the Las Vegas odds. That's where real money is at stake. Bracket-building is about prospective story-telling and retrospective story-revision.

For example, we tell each other stories about how a particular team is over-rated or under-rated. Our favorite team should certainly have been seeded higher in the bracket by those foolish and ill-informed members of the committee. And that team which displaced our favorite was far too highly regarded. We establish our preferences, and then we build stories to fit with our preferences. Finally, we call those stories predictions.

It's not that the stories have no facts or information. There is a whole season of data on which to draw. It's very satisfying to create a story that will support our preferences. But, as even the casual fan knows, brackets do not predict--at least, not in any sense other than to project our best wishes.

Once the game is completed, we begin our retrospective predictions. If we were correct in our guesses, then we can easily come up with all the reasons we were right. Count the number of times you will hear someone say, "I knew this was how it would turn out." Of course, no one "knew." After that statement will come a revised story. Some of the previous data held up during the game. Some did not. And that incorrect data will be revised with minimal effort or ruthlessly discarded as irrelevant in order to support the new and improved story.

If we were wrong, then we have to accumulate information that will support a different story. "I was never that confident they would win," someone will certainly say. And that will be the same person who was willing to stake life, limb and love on the prediction before the game happened. That person is not lying. That person is simply being human. We humans constantly revise our "memories" in order to fit with the current situation. Tournament postmortems are wonderful laboratories in which we can observe this retrofitting of memories.

We are assisted in this whole process by the professional prognosticators. They have no better predictive powers than anyone else in the long run. They are neither behavioral psychologists nor professional sports bettors. Their job is entertainment. So they are judged by how believable they can make the stories, and how much they can make us laugh and cry in the process. Accuracy is not important. The appearance of absolute conviction is what matters. And it's a real skill to maintain that sense of conviction when you're wrong fifty percent of the time.

I am re-reading Daniel Kahneman's Thinking, Fast and Slow, as the tournament unfolds. And it is an amusing experience. Kahneman talks about our story-based confidence.

"Confidence is a feeling, which reflects the coherence of the information and the cognitive ease of processing it. It is wise to take admissions of uncertainty seriously, but declarations of high confidence mainly tell you that an individual has constructed a coherent story in his mind, not necessarily that the story is true." (page 212)

So enjoy your bracket-building and story-construction! Do it with a smile. After all, it's about entertainment, right?