“SUPERFORECASTERS”— CAN SOME PEOPLE PREDICT? I have posted several times about Professor Philip Tetlock’s studies of how good forecasters are. I posted here six years ago about the conclusion that Nassim Nicholas Taleb (the Black Swan man) drew from Tetlock’s work. Tetlock had collected some 82,000 forecasts from some three hundred specialists in various fields, including politics, economics, and military, and had found that the predictions barely beat out a random forecast generator. Part 2 of THE BLACK SWAN is entitled “We Just Can’t Predict”.
Now there is a new book out which takes a contrary position and argues that there are some people who can predict. And the book is by Philip Tetlock.
The book reports on the first results of what Tetlock calls the Good Judgment Project. The book is SUPERFORECASTING: THE ART AND SCIENCE OF PREDICTION by Philip Tetlock and Dan Gardner. Jason Zweig had a review of the book in the Wall Street Journal (September 26-27). The new book says that Tetlack has identified a group of forecasters who CAN make predictions that are 30% more accurate than those of experts—or those of a hypothetical random forecast generator.