Moneyball

The link above is to Michael Lewis' book "Moneyball", a wonderfully readable account of how statistical analysis invaded baseball. Billy Beane, a guy who looked like the next baseball hall of famer in high school, did not pan out in baseball. Eventually, he became a general manager with a problem--he needed to build a playoff bound team with less than $40,000,000 when the Yankees and other teams had $140,000,000. His approach, along with Paul de Podesta and the great insights of Bill James, allowed them to come up with new statistics such as On Base Percentage plus Slugging that were good measures of whether a team would win and throw out archaic measures such as manufacturing runs, bunting and base stealing that were not predictive of winning. No one listened to Beane and Podesta. They did not care. On $40,000,000 they were beating the high income teams routinely.
The big payoff to me, besides it being a really interesting book, is in thinking harder about what things we measure for our success and which things are strongly predictive. The things that are predictive may not be the things that we are paying attention to.