When the Oakland Athletics used statistical analysis to build their team, they became one of the winningest and most cost-effective teams in professional baseball. Their approach, which inspired the book and later movie “Moneyball,” set the standard for how to build a competitive team by using large data sets to understand player attributes that might otherwise go completely unnoticed, and then harnessing this information to predict success. Any scout or casual fan could tell you a star first baseman’s batting average or a pitcher’s earned-run average. But who was most likely to hit with runners in scoring position? And which runs given up by a pitcher were really his fault, and which would have been prevented with better defense?
Today, data-driven approaches to prediction are practically ubiquitous in business. Netflix uses it to recommend shows or movies based on other shows you like. Amazon uses it when it suggests what you might want to buy based on past purchases. And as you might suspect, the art of prediction has grown in sophistication. With the advent of artificial intelligence, technologies can now spot relationships that humans could never hope to see. These technologies mine data then use an algorithm to analyze historical patterns and predict future outcomes. In fact, Netflix has said that 75% of what people watch stems from their recommendations—all thanks to predictive analytics.
It’s working for sports and for the corporate world. Now, imagine if this approach could be put to work to confront some of our toughest public policy challenges—like identifying more efficient ways to protect our health, safety and environment.