July 21, 2004
The Numbers GameI really don't do product reviews or endorsements in this space. Back in another life, when I was the managing editor for Law Office Computing, that's more or less all I did. While the technology was occasionally interesting, the entire process of doing a magazine built around product reviews left me a bit cold. So when I do decide that a product is worth writing about for Baseball Prospectus, that's something you should take seriously.
Alan Schwarz's new book, The Numbers Game, is excellent. Schwarz takes a trip through the history of baseball statistics, from the pre-history of the 1850s to the performance analysis being done today. He covers Henry Chadwick through Nate Silver, stopping along the way at Al Munro Elias, Earnshaw Cook, and some bearded guy from Kansas. The Numbers Game is comprehensive in its coverage of the history, and thorough in its descriptions of the work being done today.
I'm fascinated by beginnings and endings, so the chapters in this book that detail the earliest days of the game, and how the record-keeping developed, are page-turners for me. The way in which the decisions to track particular events in a game reflected the personal beliefs of those doing the collecting was a new concept for me, as was the descriptions of how people 120 years ago gobbled up statistics the same way rabid seamheads and fantasy players do today. There are quotes in the first 30 pages of the book that could easily have shown up on our pages today. For example:
'The best player in a nine is he who makes the most good plays in a match' [Chadwick] wrote, 'not the one who commits the fewest errors.' In other words Chadwick preferred range--the ability to field more balls overall--to avoiding the occasional error.
Somewhere, Jose Valentin is smiling.
The first 50 or so pages of this book are filled with this kind of great information that explains just how record-keeping developed, why a flawed statistic such as runs batted in became a core element in box scores and, subsequently, player evaluation, even as people recognized the problems with it.
Beginnings are a strong point in this book. Schwarz describes how the Elias Sports Bureau got its start, how the first Baseball Encyclopedia came into existence, and how Bill James found his way into answering questions about baseball while working as a security guard. Some of these stories have been told before, but none have been told in one book that ties them all together. By covering so many different people and telling their stories, Schwarz has managed to write a book that is both a great read and an important work of reference, the kind of book you'll blow through in an afternoon in the yard and then keep on your shelf next to BP and The Baseball Encyclopedia.
No book about baseball statistics can be complete without an examination of the work being done in the modern era, and how that work is making inroads into the game's highest offices. Active sabermetricians such as Keith Woolner, Clay Davenport, Voros McCracken and Ron Antinoja are all in here, as Schwarz describes their methods and how they've advanced our knowledge of the game.
Given the level of detail in the early part of the book, the chapters that deal with the modern game feel rushed, but it's understandable. There's almost certainly an entire book just in examining baseball statistics since the information revolution, and to cover everything in the same level of detail as you would, say, the development of the box score, would require doubling the book's size. Given the mandate of the book, Schwarz does a good job in hitting the high notes, focusing on the increased emphasis on OBP, the role of randomness in performance evaluation, and how the availability of pitch-by-pitch data--and even more granular information--is beginning to have an impact on both the coverage and the play of baseball.