I regularly read the writing of Keith Law. He has a blog, works as a senior baseball writer for ESPN, and he writes game reviews for Paste magazine. We agree on many things related to baseball (1), games, and science. When he announced last year that he was writing a book focused on baseball statistics and analysis, I knew I would be reading it. At the beginning of the summer I picked up a copy of Smart Baseball.
The book was a delight to read. Admittedly I didn’t get this book to be converted to a new way of thinking about baseball–it was preaching to the choir. I am in the new analytics camp when it comes to baseball, so none of the information was disturbing or startling. It has been a slow change for me, a conversation here and an article there. In 2005 I endorsed many more bunts than I do today (today I value the bunt for attempting to get a hit in some scenarios, but almost never to advance a runner at the cost of an out). In 2007 I valued pitcher wins and saves far more than I do now. In 2010 I would have preferred to see batting average to on-base percentage, but now things have changed. My baseball appreciation continues to grow, and I enjoy thinking about strategy and analysis (2).
Highlights of the book include a discussion on how the save rule results in the reduction of value of players in the closer role, detailing expected runs in relation to bunting/stealing/walking, and the measure of prospects and defensive performance.
In the last section of the book Law taps in to his experiences as an MLB front office statistical analyst and prospect scout to address how players are scouted and quantified. I really liked this glimpse into the front office world.
I think there are two primary groups of people who will appreciate Smart Baseball:
- The baseball fan. If you enjoy baseball this is a good book to read and savor. The mix of baseball talk, real-life examples for points, and clear presentation of logic (and logic breakdowns) is refreshing. I found myself reminiscing frequently when Law used a particular examples, many of which I had direct memories of or I had heard stories about. Even if you have a strong grasp of sabermetrics, this book is still an entertaining read.
- The thinker/statistician. If you enjoy seeing how people embrace logical fallacies despite glaring evidence to change, then this is a worthwhile book. Law lays out the oldest and earliest numerical measures of performance in baseball, and in most cases, why they have limited value. It is a good reminder that metrics that fail to measure something meaningful can survive due to nostalgia of familiarity. It is also a good reminder that in the world of baseball experts, at one time the primary measures of a player’s value were based on luck and the performance of others (and many current “experts” continue to hold these views).
By the way, the title of the book comes from a hashtag Law created to showcase bad baseball decisions (#smrtbaseball), which is derived from this:
The alternative to smrt baseball is smart baseball.
(1) The biggest baseball difference that we have relates to the designated hitter. Law does not believe pitchers should be hitting; he supports the National League adopting the designated hitter rule. I feel the designated hitter is an abomination.
(2) I use baseball stats in my Experimental Techniques and Analysis class. One of the lab exercises involves determining which MLB metrics are the best predictors of team success. The first table In Chapter 1 of Smart Baseball, Law presents correlations of batting average, on-base percentage, slugging percentage, and OPS with team runs per game. These data looked very familiar to me, because I have had my students figuring out these r values for the past few years. It’s a powerful lab exercise (especially for a baseball fan).