Superforecasting: The Art and Science of Prediction

Making predictions and forecasting the future are integral parts of everyday life, yet it’s unlikely individuals realize the amount of predictions one makes in a day. The use of heuristics or “rules of thumb” are ingrained in human psychology and help people quickly determine courses of action with limited information at their disposal. Deciding what time to leave the house to start your morning commute might seem mundane, but it’s a form of prediction where individuals likely use heuristics to tackle the question without engaging in serious analysis before they depart the house every morning. It’s also important to note that from a decision-making framework perspective, heuristics are not bothered by satisficing as opposed to determining the optimal course of action. Back in the hunter-gather days, if you saw a lion-shaped shadow by a watering hole you didn’t have the luxury of obtaining a large sample size of data on whether the shadow resulted in danger or not, n =1 might be enough for you to become lunch for a lion. Daniel Kahneman and Amos Tversky pioneered the theory of separating the decision-making process of individuals into two systems: System 1 and System 2, with System 1 being fast and instinctive, whereas System 2 thinking is slower.

In the book Superforecasting, authors Phil Tetlock and Dan Gardner explore the notion that in order to improve your decision making ability, individuals need to take specific steps to fight off inherent biases ingrained in System 1 thinking. Early in his career, Phil Tetlock conducted a study over twenty years, compiling forecasts from 284 experts in a variety of fields and finding that experts were only slightly better than chance in their forecasts. When asked to make predictions over a time period of 3-5 years, they were no better than a dart-throwing chimpanzee. While the findings from Tetlock are shocking when you consider many pundits make a living off peddling forecasts that have little predictive power and zero accountability, Tetlock decided to hone in on particular forecasters that were regularly more accurate than the average, and the skills required to consistently make accurate predictions. The culmination of this analysis is the subject of Tetlock and Gardner’s book, examining what traits and habits are embedded in individuals who are able to consistently score higher on their predictions, dubbing these individuals “Superforecasters.”

One of the interesting parts of the book is how the media so often turn to pundits who take a confident view on their forecast by providing either a 0 percent or 100 percent probability. As a species we find confidence comforting, and prefer definiteness to a “maybe,” despite probabilities usually being somewhere in the middle. While confidence and accuracy are positively correlated, Tetlock argues this correlation is often exaggerated, showcasing that, “people trust more confident financial advisers over those who are less confident even when their track records are identical.” Superforecasting is a fantastic book for anyone that is interested in how to engage their System 2 thinking to arrive at a more accurate distribution of outcomes when making predictions. For financial professionals that take part in some form of prediction generation each day, this is a book that will help uncover the tools needed to sharpen your decision making framework. More generally, a key takeaway is that “beliefs are hypotheses to be tested, not treasures to be guarded.”