The Book on the Book

When I am away from home without family, as I have been since the end of April, I typically have a lot of time to catch up on my reading. On this trip, I've managed to read a couple of books of interest to the Baseball Think Factory community, one of which was Bill Felber's version of the Pete Palmer/John Thorn classic The Hidden Game of Baseball. As Felber acknowledges, it was Thorn who encouraged him to proceed with the project, and Palmer (along with Gary Gillette) who advised him throughout the project. While The Book on the Book: A Landmark Inquiry into Which Strategies in the Modern Game Actually Work is not exactly a sequel or update to Hidden Game, it's in the same vein. Felber's intent up front is quite clear, as stated in the Preface to Part I:

...today we often do have the tools to determine which strategies work and which don't. We can tell through computer analyses precisely how frequently a stolen base must be successful in order to help to offense. We can develop statistical models to determine exactly how much money is too much to pay Alex Rodriguez. We can determine exactly - if only in retrospect - how large a budget any team must have in order to be competitive in the pennant race. In other words, we can gauge what works and what doesn't.

Felber goes on to apply analytical methods to a number of questions of interest to baseball fans, including many well-known (and often-discussed) topics. These include:

• How much better is it for a pitcher to be ahead in the count rather than behind in the count?
• How useful are the stolen base, sacrifice, and intentional walk?
• When should a tiring starter be pulled?
• Is modern bullpen strategy optimal?
• What is the best approach to team-building? Should teams focus on hitters or pitchers, or follow a balance approach?
• Are payroll disparities really hindering competitive balance? How much money does a team really need in order to be competitive?
• What is the "true worth" of a player, in terms of his salary? Is A-Rod really overpaid?
• How do you evaluate a GM's performance?
• What do "park effects" really mean, and how should you account for them?

Felber does a pretty good job of framing the questions to be addressed and in describing the approaches that he takes to address them. Even though he is not a "stat geek" - he relied heavily on Palmer, Gillette, and his sister, an actuary, to help him with the analysis - he demonstrates an intuitive grasp of the methods that he is using, enough so that he asks the right questions much of the time. (For example, in discussing Win Shares, he asks the logical question "Shouldn't we have Loss Shares?")

The chapters on assessing the relationship between payroll and team performance and on establishing the worth of a player are the best part of the book, as Felber goes against the sabermetric grain, to some extent - but with a well-grounded analytical approach showing that payroll "does" relate fairly strongly to team performance, that teams with lower payrolls are operating at a disadvantage, and that in general players "are" paid commensurate with performance. Felber is not wedded to a pre-defined answer to a question, but honestly tries to follow the method wherever it leads him. Felber's chapter on managers is also very good - he focuses on three commonly used strategies (SH, hit-and-run, straight steals) and evaluates the managers based on how well they deploy those strategies and how often those strategies succeed.

Unfortunately - perhaps because Felber is not well-versed in statistical analysis himself - he doesn't dig very deeply into the questions that he's trying to answer, and his conclusions are often overstated as a result. In evaluating team-building strategies, for example, Felber uses payroll as the basis for his analysis - he looked at what the average team spent on various areas of the team, for example just under 40% of payroll on pitching - and then characterized teams based on what they spent in those areas - a team that spent more than about 49% of its payroll on pitching was a pitching-oriented team, where a team that spent more than about 34.5% of its payroll on its five starters but less than about 49% of its payroll on all pitching was considered to be a rotation-oriented team - and concluded based on won-lost record that teams that the most successful teams were those that concentrated their resources into both their rotation and their overall pitching. The problem with this analysis is that payroll resources and talent aren’t perfectly related - payroll is also a function of service time as well as talent - and the fact that the payroll was pitching-heavy doesn't mean that the team was "emphasizing" pitching, as Felber puts it. I had similar questions about Felber's analysis of closers (which is not terribly dissimilar from the discussions we've had on this Web site) and starting pitchers. It's not so much that the analysis itself is flawed as that it doesn’t go far enough, and Felber doesn’t realize that it doesn't go far enough.

The Book on the Book does raise a number of interesting questions, and Felber does strive to answer them as best as he can. While he clearly doesn't answer them fully, and in some cases inadequately (and I'm sure many people here can and will point out other areas that Felber could have addressed more fully), I think on balance The Book is a worthwhile read, if for no other reason that to get people thinking about some of the unanswered questions that Felber leaves behind.