In the 2002 draft, Billy Beane faced a crisis: how to make the Oakland A’s into a viable team with a third of the budget of the New York Yankees? Before he got to know Paul, Billy had concluded that there was no way to solve the crisis: the richest teams in baseball had a huge advantage, because they could buy the best players, while the poorer teams had to settle for the bargain players.
In this passage, Lewis will study the overall structure of Major League Baseball, a organization in which some teams, such as the New York Yankees, have much more money than others. In general, it seems that baseball is unfair because the richest teams can continue buying the best, more expensive athletes (which, it’s often argued, is why the Yankees win more World Series than any other team).
In 1999, Major League Baseball created a commission to examine the inequalities of pro ball: they hired four famous people, including the president of Yale, a U.S. senator, and the ex-chairman of the Federal Reserve, to look into the subject. The commission concluded, predictably, that baseball was wildly unequal, and that, over time, the biggest teams would continue to get bigger. The problem with the commission’s conclusions, voiced by the ex-chairman of the Federal Reserve, Paul Volcker, was simple: if rich teams had such a huge advantage, then why did the Oakland A’s, the second poorest team in the league, win so many games? In 1999, for example, the A’s went 87-75.
Broadly, the commission concluded that rich teams had a major advantage over poor teams. However, the Oakland A’s are a strong counterexample to such a conclusion: based strictly on their funding, one would think that they’d be one of the worst teams in the league. The A’s success suggests that the baseball market is inefficient; put another way, some baseball players are either undervalued or overvalued.
In 1999, Billy Beane had presented to Paul Volcker, arguing that his own team’s success was, in essence, a fluke. He told Volcker that nobody went to A’s games, no matter how many games the A’s won, because there were no famous players on the team, and that, in the long run, the A’s would become less successful. But Billy didn’t believe his own presentation. He prided himself on making stars out of A’s ballplayers, and he knew that the A’s were getting better over time—in both 2000 and 2001, the A’s made the playoffs (where they lost to the Yankees). The A’s were, in short, an embarrassment to conventional wisdom, living proof that you didn’t have to have the most money to build a good team.
In 1999, Billy still spouts the same baseball dogma: he tells Volcker that his team’s success is a random fluke, nothing more. But although Billy doesn’t believe his own presentation, he doesn’t know what, precisely, accounts for his team’s success in recent seasons.
In 2001, the A’s lost three of their best players to higher-paying teams. In 2002, however, the A’s won virtually the same number of games they’d won in 2001. How is this possible? First, consider Jason Isringhausen, who left the Oakland A’s in 2001 and signed a lucrative contract with the Saint Louis Cardinals. Isringhausen had been a minor-league pitcher before Billy acquired him in 1999. Billy reinvented Isringhausen as a “closer”—i.e., a pitcher who can wrap up a close game without the other team scoring a run. However, Billy knew that closers were overrated in baseball. In essence, he bought Isringhausen for cheap and “pumped him up” to become highly sought-after. When Isringhausen left for the Cardinals, Beane won an extra first-round draft pick and a first-round compensation pick (which the Cardinals were required to give to the A’s, due to the rules of Major League Baseball). Beane then used his extra first-round picked to draft Jeremy Brown and Benjamin Fritz.
One important consequence of Billy’s strategy during the 2002 season is that no player is irreplaceable: in other words, no matter how talented a ballplayer might be, one can mimic that player’s statistics by hiring another player, or, more likely, an aggregate of multiple other players. Furthermore, many of the A’s players who, from fans’ perspective, are important to the team’s success aren’t actually that important: like Isringhausen, they’re popular and entertaining, but not always the best overall players.
The second major player that Billy traded away in 2001 was Johnny Damon, a center fielder. In many ways, the decision to trade Damon to another team was representative of the A’s new sabermetric strategy. Damon was a fan favorite, but he wasn’t as talented as fans—and the scouts—believed, especially not as a center field defense or a leadoff hitter. When Paul crunched the numbers, he realized that on-base percentage was a much more important statistic for ballplayers than slugging percentage, which ran contrary to intuition. In other words, a player who always hits the ball but sometimes gets out is far less valuable than a player who sometimes hits the ball but never gets out. Johnny Damon had an abnormally low on-base percentage, and so the A’s traded him.
Johnny Damon, much like Isringhausen, is charismatic and popular with fans, but he doesn’t contribute as much to the game as certain other players do. Paul’s emphasis on on-base percentage reflects his calm, levelheaded approach to winning baseball games. While traditional coaches would prefer a player who hits the ball a lot, Paul prefers less superficially impressive players who don’t always hit home runs but who rarely get outs for their team, either. Thus, Paul encourages Billy to transfer Johnny, whose on-base percentage is too low.
In order to understand Paul DePodesta’s sabermetric methods, we need to understand a few things about Wall Street. In the 1980s, financial markets became computerized for the first time, and markets began trading options and futures—in short, derivatives. Sometimes, derivatives were highly undervalued, and the people who figured this out, often mathematicians from elite universities, used their knowledge to make huge profits. Two such people, Ken Mauriello and Jack Armbruster, retired from finance and founded AVM Systems, a company that applied derivative methods to baseball.
Another major influence on Paul DePodesta’s management strategy, in addition to Bill James and the sabermetric revolution, was the rise of the derivative market. The link between derivatives and baseball statistics confirms what Lewis has already made clear: baseball is a business, and, as with any other business, there are strategies for controlling the market and using one’s resources efficiently.
Armbruster and Mauriello’s insight with regard to baseball was that too often, players were being rewarded for luck, and these improper rewards reflected misleading statistics. The two men learned how to use math to hold different players responsible for different plays and answer seemingly impossible questions like, “how many doubles does a player have to hit to make up for the fly balls he doesn't catch?” They found a way to assess the value of each player’s actions in terms of the same unit: runs. Everything a player did represented a fraction of a run—in effect, a derivative. When Paul DePodesta interned for the Cleveland Indians, he met Mauriello and Armbruster. In 1998, he convinced Billy Beane to hire AVM Systems to help the A’s calculate their players’ value.
Derivatives reflect tiny fractions of stocks, which can be bundled together to make, in effect, a new stock, which can be sold on the market. By the same token, Paul DePodesta’s baseball statistics reflect tiny fractions of runs, which can be bundled together to analyze a player’s overall contribution to the team. By using a derivative approach, Paul reduces the many different baseball statistics to the same common denominator: getting the most runs, and therefore winning the game.
Another lesson that Paul learned from AVM Systems was to minimize the role of luck in calculating a player’s value. Traditional baseball statistics assume that all runs are equal—that, regardless of the circumstances, one run represents the same amount of success and talent from the hitter. AVM Systems used statistics to calculate the expected outcome of a baseball scenario—how likely, for example, a player was to make a home run, given the inning, the number of people on base, and the record of the opposing team. Paul used AVM’s methods to calculate that the cost of trading Johnny Damon and replacing him with another center fielder, Terrence Long, was about fifteen runs per season.
Paul applies derivative methods to the world of baseball by eliminating the role of luck from statistics. Traditional baseball statistics weigh all hits, runs, and bases equally; Paul’s modified statistical system, however, weighs runs according to their probability. Notice, also, that Paul calculates the expected value of losing Johnny Damon in terms of runs per season—a reminder that Paul thinks globally, and, unlike a coach, isn’t too concerned about the outcomes of specific games.
In short, Paul Depodesta’s methods took much of the randomness out of baseball. While statistics couldn’t predict exactly what a player would do in the future, it could give a decent estimate of what a player was likely to do, assuming they continued playing with the same ability. Thus, the A’s traded Johnny Damon, one of their most popular players, but, contrary to the supposition of the Major League commission, the A’s didn’t become a worse team. Losing Damon and Isringhausen wasn’t a major loss for the A’s. However, the loss of their third major player, Jason Giambi, was another story.
Paul’s sabermetric methods take some, if not all, of the randomness out of baseball. Paul’s methods don’t allow him to calculate what will happen in specific games, though they do allow him to predict the A’s overall season record with a high degree of accuracy.