Moneyball or Great Player Development?
Much has been written about the success formula for the Billy Beane–led Oakland A’s. In Michael Lewis’ book, Moneyball, he cites a strategy of finding players with high OPS (on-base percentage + slugging percentage) as a key to the A’s success. These OPS standouts were undervalued by the market, meaning other teams valued them less than their fair value and less than the market valuation for batting average and slugging percentage.
Beane took advantage of this market inefficiency after he brought in sabermetric analysts who showed that OPS had a closer link to winning games than did slugging percentage or batting average. Lewis gives credit to Beane for building his team around high OPS players at discounted prices. It’s indisputable that Beane’s A’s enjoyed great success from 2000 to 2004. In 2000 through 2002, Beane delivered first-quartile wins with a fourth-quartile payroll, while in 2003 the A’s remained in the top quartile in wins but slipped to the third quartile in payroll. (2004 saw the A’s in the second quartile in wins and 16th in overall payroll, a more modest accomplishment when compared to their success of the previous years.)
Studies have shown that Lewis was right about the valuation of high OPS players. In a study by economics professors Jahn K. Hakes and Raymond D. Sauer, “An Economic Evaluation of the Moneyball Hypothesis,” the authors validate the lower valuation from 2000 to 2003 but claim the inefficiency was gone by the 2004 season. While the study reassures us that Beane and the A’s exploited inefficiency in the way players were compensated and as result enjoyed a discount in their cost of wins, it did not address an important follow-up question: Was this exploitation the key enabler for the A’s to deliver three straight years of first-quartile wins at a fourth-quartile payroll level?
My analysis indicates it did help, but had far less financial impact than many readers have inferred from Moneyball. I contend that the mix of players on the A’s roster, by pay grade—restricted, arbitration eligible, and free agents—had far more impact on their ability to deliver top wins/lowest payroll performance.
Focusing on the 2001 season as an example, possibly Beane’s crowning accomplishment to date (the second-most wins in MLB, with the second-lowest payroll), we can analyze the composition of the A’s roster and get to the core of our question.
The roster of the 102-win ball club was heavily skewed to young, captive talent. Of the top 14 win contributors (those with marginal wins of 2.9 or greater, according to Baseball Prospectus’ WARP statistics), eight were restricted players not yet eligible for arbitration, four were arbitration eligible, and two were in the free-agent pay grade. The mix of contribution to wins is 62% for the least expensive, restricted, group, 28% for the arbitration-eligible players, and 10% for free agents. In 2000, 90% of the A’s marginal wins came at a discount to the market price, since MLB’s rules prevent players from shopping their services and achieving a market price through an auction process until they are eligible for free agency.
The group of eight restricted players averaged 5.9 wins each, for a total of 51 marginal wins at a payroll cost of about $2.5 million. If we take our 2005 estimate of free-agent costs discussed in Chapter 7 and discount it back to 2001, we can estimate the restricted, arbitration-eligible, and free-agent rates per marginal win. If the A’s had achieved their 102 wins with the league average mix of 35% restricted, 23% arbitration eligible, and 42% free agents, their payroll would have been in the neighborhood of $104 million.
Their payroll of approximately $34 million represented a savings of $70 million versus the average mix of players by pay grade. Of the 12 high-impact restricted and arbitration-eligible players on the 2001 A’s, seven were originally signed by the A’s organization (five drafted in the annual June amateur draft (Jason Giambi, Eric Chavez, Tim Hudson, Mark Mulder, and Barry Zito) and two signed as international free agents (Miguel Tejada and Ramon Hernandez), while Frank Menechino was acquired through the Rule 5 draft.
According to the marginal win statistics from Baseball Prospectus, had the A’s not had the services of Mulder (8.7 WARP), Hudson (7.9), and Zito (7.3), and instead had replacement players in their roster spots, they might have been a below .500 team. So while the A’s clearly had a keen eye for talent, it may have had much more to do with pitching and less to do with OPS, at least for the 2001 club.
In the first several years of this decade, the A’s defined success as winning efficiently—making the play-offs with a low payroll. At the core of the A’s formula for success was their ability to draft and develop superstar players, not their ability to find an occasional bargain hitter who was undervalued on the market because his keen batting eye allowed him to take more walks. They clearly schooled their up-and-coming hitters—Giambi, Tejada, and Chavez—on the art of plate discipline, and it contributed to their success.
In the end, the cost savings derived from internally growing their talent had far greater financial impact than any inefficiency they exploited in batting statistics. The A’s masterfully exploited the ultimate success formula in MLB: draft and develop enough talent to staff your major league roster internally and reap the discount afforded by captive players as they await their free-agent payday.