For Scorecasting (by L Jon Wertheim & Tobias Moskowitz), the question is not whether it is a swing or a miss, rather the question is whether it is a called strike on a 0-2 count. If you read the book, you will find out that the odds of a called third strike on an 0-2 count are different than on a 3-0 count. I’m not sure where this metaphor is going, but my short review is that Scorecasting was a very enjoyable read, even for a data nerd like myself.
It should come as a surprise to no one that I would be very interested in a book about the math, statistics, and incentives of sports. At the same time, I feared picking up yet another Freakonomics retread.
It provides you with pages of interesting trivia about the way sports are actually played. It encourages you to yell at referees when you think they make a bad call. It provides you with data to support the claims you are about to make to your buddies over a beer. It will not, however, tell you the details of how the authors established that data.
A typical vignette begins with a general observation about sports, maybe one of those clichés that everyone “knows”, like home field advantage. The authors then proceed to demonstrate whether the truism is true or not and to investigate potential causes. In chapter after chapter, the authors are clear and as comprehensive as they can be on each topic. And, unlike their predecessors in Freakonomics, Wertheim and Moscowitz are careful about correlation and causation when extrapolating from their data.
Occasionally, the authors expand further and explain not only what athletes (and referees) do, but why they do it. This element is usually speculation based on research from other fields and is the major weakness in Scorecasting. While the observed phenomena may be entirely consistent with previously reported human behaviors, that does not mean that we can assign causation. The authors do fine when they limit their conclusions to the scope of their data.
Something that is not a weakness, but more an amusing truism of time’s inexorable march forward is that some of their analyses regarding specific players (e.g., Eli Manning, who the criticise, but was phenomenal in the 2011-2012 NFL season & won the Super Bowl) is a bit faulty with the benefit of 20/20 hindsight.
It is possible that there are additional weaknesses in Scorecasting. The fact that the authors do not really show their work (it’s not that kind of book) makes it impossible to evaluate Scorecasting at that level. That does mean that I don’t think that the non-sports fan data nerd will find enough numerical and methodological meat in the pages of Scorecasting to keep themselves satisfied. If you are a sports fan, not only will you enjoy Scorecasting, but you may start watching your sports a bit differently.
- Maybe we should also praise referees for making what we think of as “good” calls? Works for my dogs and children. More consistently with the dogs.