Spring 2007
"From Prediction to Learning: Opening Experts' Minds to Unfolding History"
Journal Article, International Security, issue 4, volume 31
By Richard Herrmann and Jong Kun Choi
No expert in the academic or intelligence community can predict the future, but they should at least be able to accurately analyze and quickly update their beliefs to craft effective policy. Too often, experts not only forget what they used to believe, but also see little connection between explaining the past and predicting the future. A two-year case study of fifteen Korean experts examines their initial predictions about security on the Korean Peninsula and demonstrates how a Bayesian approach helped them first to empirically analyze their theories, and then to successfully update them based on events that actually transpired.
