Introduction

Good Judgment Project beat ARPAP benchmark in the first year and all subsequent years. Other teams were not able to do so for reasons that are less interesting, including mismanagement of how they went about it. The project had a bit of a reputational advantage because of earlier work I did on expert political judgment. We put them together in teams and we gave those teams guidance on how to interact with each other. It turns out that both prediction market, collaborating in teams and competing in prediction markets tend to boost performance.