The Good Judgment Project and Geopolitical Forecasting
Over the last two years, two researchers have conducted an experiment to explore whether or not there is a way to identify individuals with exceptional ability to predict geopolitical events. Together, Barbara Mellers and Michael C. Horowitz created the “Good Judgment Project” an online prediction market consisting of volunteers who had a Bachelor’s degree or higher. In all, the experiment yielded over 150,000 forecasts across 200 different questions and established a framework for what makes a “superforecaster.”
While it was primarily a psychological experiment aimed at establishing that aforementioned framework, the Good Judgment Project shows that the concept of prediction markets can be quite powerful in geopolitical contexts. The experiment’s findings show that one of the common pitfalls of supposed “experts” in geopolitical forecasting is the lack of concrete feedback. Predictions are made in accordance with doctrines and schools of thought rather than in any system that keeps score. Psychologically, this is the strength of prediction markets: they allow for instantaneous feedback and a greater sense of accountability. Rather than in the occasionally vague world of geopolitics, prediction markets create a numerical system that rewards participants for shifting forecasts because of past errors. This reward system that is predicated on shifting prediction styles directly confronts the concept of doctrinal thinking. More generally, the Good Judgment Project’s success indicates that prediction markets may very well be a part of geopolitical deliberations in the future.
Source: