The Science Behind Good Forecasting

The art of forecasting is notoriously fickle. It is often underappreciated how much luck or chance is responsible for a specific outcome occurring. To make matters worse, those responsible for making predictions tend to focus on the result, as opposed to the process used to arrive at the decision, when evaluating their skills. It is well documented that “experts” in their respective fields are surprisingly bad when it comes to prediction, as it is much easier to get entrenched about an idea or view when your knowledge on a matter is comprehensive but narrow in scope. This article from The Atlantic explores the work of Philip Tetlock and some of the keys to successful forecasting. Acknowledging that “the world’s most prominent specialists are rarely held accountable for their predictions, so we continue to rely on them even when their track records make clear we should not”. Tetlock explores what traits and characteristics are essential for improving the accuracy of forecasts. Tetlock also describes how top forecasters continuously update their predictions as new information is received, not confining themselves to a narrow focus. Essentially, the top forecasters ended up being effective collaborators who crossed disciplines and viewed their teammates as sources for learning; in short, they were “curious about, well, really everything.”