Bryan Caplan  

Moneyball: "Your gut makes mistakes and makes them all the time."

PRINT
Basel III's Deadly Cocktail... Billy Beane Arbitrage...
I'm no sports fan, but as a betting man, Moneyball intrigued me.  I thought the ending dragged, but the peaks were high.  My favorite (from the shooting script, so the final scene may be a little different):
BILLY:

Grady, you don't have special powers.  You don't have the ability to look at a guy and "just know" because you're a scout with special powers.  Your gut makes mistakes and makes them all the time.  I've watched you sit at kitchen tables for years and tell the parents of a 17 year old kid, "Trust me, when I know, I know, and when it comes to your son, I know" and you don't.  It's mythology.
My biggest question after the movie: How good is the statistical evidence that teams that embraced sabermetrics won more games as a result?


Comments and Sharing





COMMENTS (18 to date)
David R. Henderson writes:

Bryan
The evidence is quite good. Oakland did well with it. Then Theo Epstein, who believed in the same approach, moved to Boston Redsox and had both that method and a high payroll to work with. The results in Boston in the last decade speak for themselves.

FiftySeven writes:

Almost all teams use sabermetrics on some level now, and when they don't, it shows. Though, the teams that are most successful blend scouting and statistics.

Ryan writes:

During the time Moneyball was taking place, saber groups were consistently beating the odds in Vegas for predicting end of the year records. That isn't true anymore since it's not public knowledge.

However, people began realizing around the middle of the last decade that there is significant value added in scouts. It's a tacit knowledge thing.

The Tampa Bay Rays now are rated as having the best front office by people into sabermetrics. They consistently win while keeping a very low payroll despite playing in the most competitive division in baseball. See The Extra 2% by Jonah Keri for their story.

Andrew writes:

David;

I think that's a little misleading. Clearly Oakland did very well in the late 90's and early 2000's will a very small payroll. On the flip side, in the famous 'Moneyball' draft, the A's selections did no better than average.

In terms of Boston, they have the second-highest (tied with Philly, below the Yankees) payroll in the league, you can't say they're doing anything special.

What really bothers me is that there are likely general equilibrium effects from the hunt for value. If, in 2004, everyone decides to stock up on OBP, and watches for that stat in their prospect ranks, how do developing pitchers react to facing casts of batters that instead of hacking now want to take lots of pitches and draw walks? It's not clear the precise lessons of Moneyball will survive.

8 writes:

Andrew, they did something different after failing to win a championship for 80 straight years. Maybe the effect was marginal, but it paid big dividends.

Rob writes:

Bryan,

I think the most compelling evidence that other teams have adopted money ball strategies is the lack luster record of the A's over the last several years. Most people cite the A's subsequent sub par performance as evidence that money ball doesn't work. On the contrary, the A's found an arbitrage opportunity in the late 90s and exploited it through the early aughts and over achieved relative to their pay roll. After the success was well known to baseball aficionados the money ball approach was adopted by teams with bigger payrolls. Consequently, the A's are back to where they started. So in summary, I see the A's current lack of success as evidence of the money ball approach gaining acceptance among other ball clubs.

DRDR writes:

The evidence is strong that the 2003-2004 Red Sox benefited from applying a Moneyball philosophy. These two teams' offenses ranked among the most prolific of all time. Theo replaced about half the linkup in 2003, traded a truly average "All-Star" named Shea Hillenbrand. You look at that lineup, they vastly improved at 4 positions. (see http://www.baseball-reference.com/teams/BOS/2003.shtml ) Contrast with 2002 in OPS+ (>100 means better than league average).

I'm not sure how you'll get better evidence than a few case studies of individual teams like the A's and Red Sox earlier this decade. Clearly the value of the strategies pursued by these teams eroded as other teams discovered them. Paul DePodesta, the Asst. GM who the Brand character was loosely modeled after, was a failure as Dodgers GM 2004-06. The Red Sox haven't been nearly as successful lately. No doubt everyone applies some Moneyball lessons today. That's why the Sox & Yankees now play 4-hour games regularly and always work deep in counts.

It's an overly simple lesson to conclude from Moneyball that innovating in Sabermetrics is going to be the path to a sustainable competitive advantage. It may be true that the next competitive advantage is going to be in aspects of the game that aren't easily quantifiable, now that teams are devoting so many resources to quantification. So many free agent signings are busts today, guys who just stop working once they get the big deal: maybe teams should hire more psychologists?

Thomas DeMeo writes:

If the basic theme of Moneyball were extended to economics, some Billy Beane like character would be telling you that you asked the wrong question.

Steve Sailer writes:

The movie is very good and the screenplay is one of the smartest I can remember. I am in awe of how much of the book they got into the movie in a comprehensible fashion.

Of course, what's not mentioned is the interaction between moneyball and steroids. Beane's strategy appears to have been to find players with a natural ability to not swing at bad pitches, tacitly encourage them to juice up, which discourages pitchers from throwing the ball over the middle of the plate because they'd crush it for a homer, which leads to even more walks. The Giambi brothers are the most obvious examples but there are a lot more. Bill James's World Series winning Red Sox were similar with juicers Manny Ramirez and David Ortiz.

Steve Sailer writes:

As much as I admire the sabremetric revolution intellectually, it helped make baseball worse by exacerbating trends toward steroids and protracted games. Baseball before Bill James figured out how to game the Game was more elegant and more fun. Sure, stealing bases, for example, was pretty stupid, but it was exciting. Watching some lumbering steroid-user walk on 11 pitches, then stand on first base until the next guy with a giant head from HGH use walks is pretty torpid.

Steve Sailer writes:

Oakland was the center of the baseball juicing universe under the leadership of Billy Beane's predecessor as General Manager Sandy Alderson, the first GM to appreciate Bill James. In 1993, a baseball agent told me that Jose Canseco was the Typhoid Mary of steroids, now infecting Texas after infecting Oakland (e.g., Mark McGwire). A decade later Canseco wrote a book detailing exactly that scenario: he had been the first to openly proselytize for PED use in lockerrooms. (Before that it appears to have been furtive and solitary.)

Steve Sailer writes:

Bill James, by the way, consistently ignored or, if pressed, pooh-poohed the impact of steroids and HGH on statistics. Even in the middle of the last decade he was still writing ridiculous stuff about how Barry Bonds' special maple wood bat might help explain his insane statistics. James had the statistical tools to expose the steroid scandal, but if he had, he wouldn't have gotten that swell job in Boston.

frankcross writes:

I don't think you can blame moneyball for steroids. And I wonder what the evidence is for the claim that Beane encouraged them to juice up. I suspect that the word "tacitly" means there is none.

drobviousso writes:

I find baseball fans' obsession with steroids and other drugs both fascinating and exasperating.

John T. Kennedy writes:

@frankcross I don't think Sailer was blaming Moneyball for steriods, just observing that Beane took advantage of them.

Surely it was obvious to Beane (and countless other baseball execs) that some of his players were breaking the rules - rewarding them for the results of that is tacit encouragement.

Steve Sailer writes:

The book explicitly makes a nature-nurture argument: strike zone judgment is nature, but power is nurture. "Power can be acquired," says Beane in making this argument. The book doesn't follow up on how exactly power, more power than a player's genes would seemingly allow, can be acquired, but it's obvious, isn't it?

Jim Glass writes:

Paul DePodesta, the Asst. GM who the Brand character was loosely modeled after, was a failure as Dodgers GM 2004-06.

Because at 31 he had none of the skills needed to be a GM. It was no failure of sabremetrics.

Knowing a lot about statistical analysis gives you none of the business and baseball skills needed need to be a GM. Otherwise they'd be recruiting GMs out of math departments. If you have the skills to be a GM, then adding the statistical knowledge makes you a better one.

There's one scene in Moneyball were Beane gets on the phone with the Mets Steve Phillips and robs Phillips blind in a trade. Personally outmaneuvers him in negotiations and gamesmanship. That illustrated the fundamental GM skills that made Beane who he is and Phillips a TV commentator (until he lost that job too). They don't teach those in econometrics class.

The Moneyball philosophy is moving into the NFL too, where the top practitioners are Bill Belichick ('nuff said) who has a degree in economics, and Lions coach Jim Schwartz, who's doing a very impressive job lifting a perpetual cellar-dwelling team up into the light. So on the evidence it helps an NFL coach to be a stat head. But being a stat head does not in any way qualify one to be an NFL coach, obviously.

Jim Glass writes:

Grady, you don't have special powers. You don't have the ability to look at a guy and "just know" because you're a scout with special powers. Your gut makes mistakes and makes them all the time.

Beane was just applying Bill James's philosophy:

"There is what we know and what we can demonstrate. What we know but can't demonstrate is [prohibited by EconLog commenting guidelines]".

Comments for this entry have been closed
Return to top