“Superforecasting: The Art and Science of Prediction” by Philip Tetlock

Tetlock’s book is an intersection of psychology and the social sciences, focussing on how humans make predictions and how they can make them better. He explains that predicting the future is a skill you can learn and improve on. The book recounts the behavioral fallacy traps that are most common and the best ways to avoid them. Tetlock states that people who are good predictors are well read, adjust their views often and in small increments, tend to judge proportion and magnitudes well, seek out contrary opinion and will be swayed by it, and use good statistical baselines before being swayed by current events. Most also use Bayesian logic (informally), even if they do not know that term. However, the key to judging predictions in the first place is not giving vague statements open to revisionist interpretation, but, instead, concrete facts that can be absolutely proved or disproved within a given timeframe. That is not as easy as it sounds: most media pundits and prognosticators make careers out of weaseling their way out of such falsifiable predictions with caveats and clarifications.

The predictions studied focus primarily on the realms of geopolitics and economics, but the methods can apply to most of everyday life. Importantly, it finds that prediction is a skill that can be improved upon with hard work and that often the highest IQ people are the worst at making predictions. The takeaway- “try, fail, analyze, adjust, try again.” The research techniques gleaned are easy to apply to both make probability predictions and how to better interact with people and the news everyday. This book shows how to make precise predictions about hard facts and shows why so many talking heads in the news media consistently refuse to do so. The greatest strength of this book is the clarity of mind it forces upon you. So you might be interested in whether there will be a Third Intifada or if there will be a bigger Great Recession. But you should try to predict the probability that 30 Israeli citizens will be murdered by Palestinians in the next five months or whether the Dow Jones will be lower than 12,000 at any point in calendar year 2018. Or whether any IDF tanks will enter Gaza in the next two months or whether the S&P 500 will go down 10% in the next four months. The specifics don't matter, as long as they are specific and testable. Those are the questions you really can answer with clarity, putting your reputation where your mouth is.