Michael Lewis 2004. Moneyball. Norton, Kindle edition. With afterword. 337 pp.

Superficially this is a baseball book (with lots of jargon impenetrable to those not thus indoctrinated). However it is also a narrative about bringing a revolution to a field. In this case the field is baseball, which as Michael Lewis tells it, is ruled by unsupported myth-based opinions of a club of old players and the poorly-informed sycophant journalists who write about them. In other words, just like cricket, AFL, tennis, and probably Cane Toad racing.

In this story, the two principal characters are Billy Beane, general manager of the Oakland A’s¹ and Paul DePodesta, assistant general manager². Paul DePodesta is the manager and creator of the statistical data that enable the revolution in this book. It is amusing that even the new wave of evidence-based baseball rationalists trying to turn the game upside down feel the need to invent a new word—“sabermetrics”—for what is just statistics! Although they are mostly outside the “club” of old baseballers and journalists they appear hell bent on creating their own club! Petty criticism aside, an interesting read, and a quick one.

The Oakland As have little money for players so Beane and DePodesta just get the best baseballers (best as judged by what their statistics say about their performance) instead of the best-looking baseballers. Since these players are often “faulty” they come cheap (“faulty” = not photogenic, not marketable, not conforming to the old baseballers concept of what a young champion should look like and so on). Thus the Oakland A’s won a lot of games on a budget a fraction of the big clubs, and ruffled many feathers in so doing. Apparently they still do. Anyone who doubts that Beane and DePodesta were not up for a revolution needs to read an article by Paul DePodesta on The Genesis, Implementation, and Management of New Systems in which he quotes Thomas Kuhn at the end of the piece.

The reader gets to know a few of the characters in the book, mostly players. But what I cried out for was some more detail on the methods Paul DePodesta used to gather, manage and interpret his data. A bit of a search failed to find much, which may mean there are sophisticated systems that no-one wants to talk about because of their commercial value. No doubt that is the case now. But in 2002 when this book was being written, I suspect that Paul DePodesta simply used spreadsheets. There is an article Sabermetrics: The Past, the Present, and the Future written by Jim Albert and a Springer book Curve Ball Baseball, Statistics, and the Role of Chance in the Game edited by Jim Albert and Jay Bennett but nothing like the Bayesian approach of fivethirtyeight

I would have liked to know how Paul DePodesta amassed so much data from so many leagues and games around the USA. Not by himself, certainly. How data gathering on such a scale was funded is never explained.

The book is exhaustively indexed.

¹Oakland A’s is a contraction of Oakland Athletics, so the apostrophe is not spurious as I first thought.

²As of March 2017, Paul DePodesta was with Cleveland Browns NFL club while Billy Beane is still in baseball with Oakland Athletics.