Stats v. Scouts

Michael Lewis’ novel Moneyball greatly increased the popularity of a theory called sabermetrics, which in short, is the “analysis of baseball through objective evidence.”  As applied, sabermetrics involves teams employing advanced statistics to find undervalued players through free agency and the draft.  Moneyball discusses how the Oakland Athletics used Sabermetrics to successfully compete with teams with much larger payrolls, but Lewis’ reasoning spurned an ongoing debate about whether these non-traditional theories were the best way to build a team. 

The Rise of Sabermetrics
Sabermetrics has a growing base of followers among baseball fans which began even before Moneyball.  These fans often display an intellectual approach leading them to react critically to anything not following the sabermetric approach.  For example, here is a sabermetrics enthusiast’s response (from a random comment board)  to a fan who liked Giants’ prospect Madison Bumgarner because of his “strikeout-inning ratio”:

“footballpimp, it’s actually pretty awesome that you’re trying to use stats to back up your stance. It’s just that you don’t really have a grasp as to how to use them properly yet. You’ll get there, I’m sure. But in the meantime, you need to recognize that the stats you’re throwing out are close to meaningless, especially in the context in which you’re attempting to utilize them…In the meantime, it would probably behoove you to research stats like FIP, tRA, WAR as well as contextual statistics such as park factors and league factors. Those elements, plus Bumgarner’s delivery motion and velocity, paint a much different story than the one you’re asserting. There are many subtle layers to the art of using statistics to extrapolate and predict future performance.”

This sabermetrics poster clearly knows a lot about baseball.  But I was amazed at how dismissive the response was of traditional statistics…my only criticism would have been that username (just kidding). 

The Critics
On the flip side, “Moneyball” has been attacked by a group Lewis refers to as “the Club,” a group of baseball insiders who dislike Sabermetrics.   Lewis is correct when he says: “By the end of the 2003 baseball season I had learned something from publishing Moneyball. I’d learned that if you look long enough for an argument against reason, you will find it.”

Two of the most prominent articles criticizing “Moneyball” come from the an ESPN story and “Friday Night Lights” author Buzz Bissinger. Bissinger criticizes Moneyball for underestimating how the dominating combination of Hudson, Mulder, and Zito played a major part of Oakland’s success.   And both articles cite Oakland’s lack of recent success, along with the termination of Beane disciples J.P. Ricciardi (Blue Jays) and Paul DePodesta (Dodgers), as proof that Moneyball’s theories are overrated.

To me, Bissinger fails to give Beane enough credit for the talent he acquires on a limited payroll.  This economics article from Clemson University does a great job defending Moneyball on economic theory.  The recent criticism of Beane is perhaps somewhat attributable to Lewis’ writing style, which tends to make people out to be greater geniuses than they are.  In reality, Beane is a very good G.M., but nobody can live up to the expectations in Moneyball.

Where I think Moneyball  goes wrong is where statistics are viewed as a replacement to traditional scouting.  For example, in Michael Lewis’ 2004 Sports Illustrated article defending Moneyball, he cites J.P. Ricciardi’s hiring in Toronto as the ultimate implementation of Moneyball philosophies.  Lewis says: “Ricciardi had done what every enlightened G.M. on a budget will eventually do: Fire a lot of scouts, hire someone comfortable with statistical analysis and begin to trade for value.”

To me, this is where Lewis’ case goes too far, and if this was the strategy in Toronto, it epitomizes why things did not work.  It also shows Lewis undervalues the importance of an experienced scout.  But to be fair, Lewis wrote the article without the benefit of hindsight, and in reality, people can argue Ricciardi and DePodesta both went away from the strategy by signing the wrong players to massive contracts.  

The book “Scout’s Honor,” intended to refute Moneyball, discusses the success of the Atlanta Braves through a strategy which was very different than Oakland.  Among other things, the Braves place tremendous emphasis on the “makeup” of a player.  They tend to draft local players because where they can learn a great deal about the player’s work ethic, background, and other factors which a computer cannot quantify. Basically, the Braves believe an 18-year old with the right makeup can develop into a great player with the help of their coaching staff.   Finally, the Braves like to scout in places where other teams are not scouting (sort of a non-statistical “inefficiency”).   For example, the Braves discovered Andruw Jones in the Netherland Antilles while Tom Glavine was found in a part of the Northeast which was not highly-scouted.  In short, as this debate on Baseball America discusses, scouts provide many benefits which cannot be achieved using a computer.  

My View
My view on “Sabermetrics” can be summarized by a quote from analyst Maury Brown, who said “Moneyball works in giving you another set of tools.” In other words, I believe Moneyball and traditional scouting can serve as effective complements to one another. 

Boston is the perfect example of a team which perfectly implements the Moneyball theory along with traditional scouting.  Boston’s general manager, Theo Epstein, famously hired statistical analyst Bill James to work as a consultant for the team.  But Boston also focuses on some of the same principles discussed in scout’s honor, most importantly focusing upon player development and finding players in areas less scouted by other teams.  Larry Dierker, one of my favorite baseball people, wrote a great article about how he implemented Moneyball principles when he was with the Astros. 

Just as Moneyball describes, selecting baseball players is analogous to making investments.   Just like business investors, baseball teams should employ the best statistical analysis possible when making their investment.    But as Warren Buffet said, “beware of geeks bearing formulas.” As many investors learned, the best statistical analysis is not effective unless combined with sound traditional judgment.   Therefore, to me, for a team to be successful, they need to find a method of making “stats and scouts” co-exist effectively.