In 2011, Philip Tetlock founded a multiyear forecasting study called the Good Judgment Project with his partner Barbara Mellers. The GJP is a volunteer-run study that recruits people to forecast the future. The study includes superforecasters like Bill Flack, a Nebraska retiree who consistently makes accurate predictions in areas far outside his expertise, such as the chance of Russia annexing Ukrainian territory in the ensuing three months. After recruiting a cumulative 20,000 volunteers from various walks of life, Tetlock and his team concluded that it is not what people think that is important, nor their professional expertise, but how they think. Superforecasters are intellectually curious and willing to do their research. They also consider conflicting arguments and update pre-existing beliefs. Superforecasters like Bill make up 2% of the GJP’s volunteers, and they often outperform professional bodies in other institutions when it comes to making predictions. The GJP’s initiatives, including its Ten Commandments for Aspiring Superforecasters, improved accuracy in forecasting by 10%, which may seem modest but has a profound effect over time.

The authors argue that we live in a world beset with forecasts. On the news especially, many so-called authorities are called upon to make predictions, many of which are unfounded.