Superforecasting: The art and science of prediction (Philip Tetlock)

Abstract

This article by Phil Tetlock discusses the research conducted by the Good Judgment Project (GJP) that sought to identify individuals who were skilled at forecasting the outcomes of geopolitical events. The researchers conducted multiple forecasting tournaments, and one result was that there was substantial variation in forecasting skill among participants. Superforecasters, or the top 2% of forecasters, were able to consistently outperform chance and demonstrated skill in various domains within political science. These findings suggest that forecasting skill can be learned and improved through practice and feedback. In addition, superforecasters exhibited certain cognitive traits and strategies, such as an ability to think in probabilities, consider alternative scenarios, and resist biases. By understanding the factors that contribute to forecasting skill, the research conducted by the GJP has the potential to improve our ability to predict future events and make better decisions in various domains, including politics, economics, and business. – AI-generated abstract.