Review: Smart Baseball
Smart Baseball is Keith Law’s introduction to the basics of sabermetric baseball statistics and why fans should care about them. I’ve been hearing about this revolution in baseball statistics for a while, albeit peripherally. Law does a fine job explaining the shortcomings in older statistics, explaining many of the new ones, and pointing out the effects this is having on statistics gathering and other ramifications for baseball going forward.
Honestly, I learned as much about the prevailing attitudes in baseball as I did about new statistics. My conclusion is that if the baseball broadcasting, management, and fan constituency is resisting these ideas this to any extent – and I believe they are – hide-bound does not come close to expressing the status quo. I have enough of a traditionalist in me that I feel no pain in continuing to report saves and RBIs (and yes I prefer appending the ‘s’ for the plural), but as tools for understanding the game they are both baroque and broke.
Law does a fabulous job of laying out the case for adding on-base percentage (OBP), slugging percentage (SLG) and their stable mates to common usage. Law’s approach is reasonable and clear. He never argues that the new statistics are somehow perfect or should be used as matters of religion. All statistics and observational frameworks illuminate and emphasize different aspects of the game. Nothing emphasized that more for me than the slash line (AVG/OBP/SLG). Characterizing a batter by the triple of traditional batting average, an on-base metric, and slugging lets me understand a batter better. I can get the gut feel that a historical batting average gives me, the more sound OBP indication of out avoidance, and the marginal value of each hit (how often the guy goes extra bases). That historical value will keep going down, so eventually the slash line will probably get smaller.
How does one even marshal an argument against that? Ink costs?
As the he explains more data-driven metrics like those used to evaluate fielding and then the overarching wins above replacement (WAR) formalisms, he makes the source of the numbers clear. I should be clear that he explains the process of getting and evaluating the data, not the formulas. Some of the formulas are – surprisingly to me – proprietary anyway. As a result I understand the formulation and limitation of these combination metrics better. And they do have limitations. While I find the simple batting metrics to be fairly objective, WAR and the fielding evaluations include a healthy dose of subjective evaluation. Which ones light up which aspects of the game is trickier to see. If one wants to make those evaluations, Law shows you what to investigate.
Finally he describes how these stats (and the old ones) are used today in scouting and other player evaluations. He also describes Baseball’s new commitment to providing a torrent of detailed data. One assumes that this is to even out data collection’s marginal value to teams as well as amortizing collection costs. The result is an incredible mine of information, though.
Overall a cool book from which I learned a lot.
Recommended.