TDL Book Reviews: Moneyball
There are quite a few books that have stacked up over the years on my winter summer whatever reading lists. It would take me a long time to list out the books I’ve thumbed through in Borders and said “damn, I should get this.” Then I realize that it’s dumb to spend $20 or $30 on a hardcover book that I’ll read once and then turn into a decoration.
About six months ago, someone introduced me to this odd concept. It’s a store, kind of like a video store, but with books instead of movies. It must be like a chain because they have them all over the city. What’s better, they don’t even charge a rental fee… just late fees… and you get three whole weeks when you rent a book. The website, similar to Netflix, allows you to create an account and set up a book queue. You can pick which location you want your book to go to. When you bring it back, you can even request another book to be sent to your location.
Yeah, two years into the New York City experience and I finally bit the bullet and got myself a New York Public Library card. The first book I got was one I’d been putting off for quite some time, but have been fascinated with.
Moneyball by Michael Lewis is the story of the Oakland A’s. It set out to answer the question “how does a team with so little money win so many games.” The premise is relatively simple. In 2001, the Oakland A’s spent $40 million dollars and put together a 102-win season. The New York Yankees spent $130 million dollars and won 95 games. How was it that a team that totally flew in the face of commonly-accepted baseball knowledge (money = success) win so many games? How was it that teams like the Milwaukee Brewers and Kansas City Royals, teams that consistently claimed their impoverished states would never allow them to be competitive, could never be good, yet the A’s were able to win 100 games?
Well, most claimed it was luck.
Funny thing about baseball is that it’s the only sport with a sample set big enough to reveal trends over time. One guy, back in the 1980s, figured this out. He discovered that applying different mathematical formulas to massive amounts of baseball data available would reveal trends. The trends revealed that certain statistics (the batting average, save, win, RBI, ERA) were tremendously overvalued… meaning a team would pay a ton of money guys with gaudy statistics in these categories. On the other hand, there were statistics that were tremendously undervalued at best, and flat-out ignored at worst (On-Base Percentage, Slugging Percentage, WHIP).
What the A’s discovered, and exploited, were that while the Yankees would pay a guy with .310 batting average (AVG) with a .310 on base percentage (OBP) a boatload of money, they’d be less likely to pay for a .220 AVG with a .395 OBP. They discovered that all these mathematical formulas were largely ignored by the rest of baseball. They paid the going rate for a guy with a low On-Base Percentage… and cleaned up. It included the history of how some statistics, like Batting Average, were invented… that is, the original baseball box score was derived from the cricket box score. Cricket does not include the “walk”. Therefore, the “walk” was originally credited as an Error to the pitcher and the average was only meant to include clean at-bats. That is, a plate appearance where the batter either struck out or hit the ball into play. Since baseball is, if anything, throughly resistant to change, this statistic stayed as part of the game and has become important over the years, like the Win for a pitcher, even if it’s horrifically flawed.
Of the 20 or so people who read this blog, I know a few of you are totally-non sports fans. If I haven’t lost you already, this piece will likely make your PhD-rational mind bleed. The batting average measures how good a person is at hitting the ball into play or striking out. On base percentage measures how good you are at getting on base. What the ivy league snobs who worked for the Oakland A’s realized, well before the rest of the league, was that it didn’t really matter how a guy got on base… so long as he got on base. They realized that a walk (that is, getting to first base for free) was just as valuable as getting to first on a hit. So, if you read a box score that says a batter went 0-2 with 2 walks, it means he came to the plate four times, got on base via walks twice and either struck out or hit the ball into play and was put out twice. His batting average for the day is .000. His OBP for the day is .500. The A’s realized that guys who got on base more were better. As straightforward a concept as this seems… it’s resisted in baseball. Even to this day, old baseball men will continue to flaunt ignorant baseball-isms like “you can’t walk home.”
A good part of the beginning of this book are about statistics. Essentially, how a group of stat-geeks got together behind the writing of a man named Bill James and discovered there were a LOT of undervalued statistics in baseball. They discovered different ways to look at the game. As baseball is very resistant to change, very few people in baseball were willing to listen to the advice of guys who weren’t in baseball. Even though you could show them piles and piles of data to prove that sacrificing an out on purpose is a bad idea… they didn’t believe it. Sacrificing was a way to “manufacture runs”. It is not a way to “give up outs”. As I made mention of in another post, the feud between guys who find ways to apply this math vs. the guys who bull-headedly believe there is nothing to it rages on. If Bill James created the movement in the eighties, this book brought it to the Internet generation.
The rest of the book follows the A’s 2002 season. It starts at the beginning of the season, the season in which they had to replace Jason Giambi. Giambi had just taken a huge paycheck from the Yankees. The A’s were faced with the unenviable task of replacing him. The process by which they go about this: determining what Jason Giambi contributed to the team and mathematically figuring out what offensive tools they would need to replace him in the line-up. How they determine that defense is over-rated. It follows Billy Beane through to the trading deadline and describes how he’d become so good at fleecing teams with trades that teams had become suspicious of dealing with him. You get some story about the A’s 20-game winning streak, capped off by a walk-off blast by a guy, Scott Hatteberg, who was let go by the Rockies. The A’s were able to pay him $900k/season because the league does not value walks. Hatteberg was one of their examples of a guy whose OBP was roughly 100 points higher than his batting average, but was let go because he was no longer a good catcher. The A’s picked him up, made him learn first base, and made him a valuable player. It wraps up with the 2002 ALDS and the fallacy of “manufacturing runs” and “being aggressive on the base paths”; old baseball clichees that, when data is analyzed, don’t really work out the way you think. Manufacturing runs, in baseball speak, is giving away outs to try and get runs.
Really, if you haven’t read Moneyball and you think you know what it’s about, you probably don’t. It’s really a study about how baseball, for the most part, is an insanely bull-headed and stubborn business. Most teams still refuse to recognize that there might be some other way to look at baseball than the way it’s “always been done”. Mathematical models can, in fact, be applied to the mounds and mounds of statistics compiled over 130 years to come up with different ways of looking at the game. You can create a statistic called “Batting Average On Balls In Play” to figure out how good a pitcher really is independent of the defense behind him.
For a baseball fan, highly recommended.