Today on Trend Following Radio Michael Covel interviews Philip Tetlock. Phil is a Canadian American political science writer currently at The Wharton School of the University of Pennsylvania. He is right at the intersection of psychology, political science and organizational behavior. His book, “Superforecasting: The Art and Science of Prediction,” is about probabilistic thinking defined. Phil is also a co-principle investigator of The Good Judgment Project, a study on the art and science of prediction and forecasting.
Michael starts off asking, “Regular folks can beat the experts at their own game?” Phil says essentially that is correct. He started The Good Judgment Project in 2011. It was based around forecasting and was funded by the government. He was shocked by the amount of “regular” people he recruited for his study that were able to compete with, or do a better job predicting than professionals working for agencies such as the NSA.
Michael and Phil move onto discussing the Iraq war. They discuss what the actual probability may have been of Saddam Hussein having weapons of mass destruction. George Bush claimed that it was a “slam dunk” when clearly there was not a 100% probability of weapons of mass destruction being there. Michael asks, “When is society going to adopt more of a probability mindset?” Phil says that soft subjective human judgment is going by the way side. Pundits saying, “Someday this will happen” without any real substance, will come to a stop. As long as a forecaster can say, “This may happen in the future” then they can never really be held accountable for being wrong. Michael brings up the example of Robert Rubin. Robert worked for Goldman Sachs and was under Bill Clinton during his presidency. He was a great probabilistic thinker. Everyone loved him until the 2008 crash. Phil uses him as an example of even the best prediction people getting it wrong.
Bottom line, superforecasters look for aggregated data. They know there is interesting data laying around and they tend to look at crowd indicators heavily. The distinction between superforecasters and regular forecasters is their ability to start with the outside view and move to the inside slowly. Regular forecasters start with the inside view and rarely look at the outside view. Superforecasters also believe in fate less than regular forecasters do. When you highlight all the low probability events surrounding outcomes, such as the lottery, many chose to think the event was decided by “fate” or just “meant to be.” Superforecasters think in a way of “well someone had to win, and they did.”
In this episode of Trend Following Radio:
- What are superforecasters?
- Probabilistic thinking
- Looking at aggregate data
Direct download: 425.mp3
-- posted at: 10:00pm EDT