Balls, Sticks, & Stuff
In Defense of Moneyball and Sabermetrics
One of the baseball blogs I read quite often is Swing and a Miss. The author, Tom Goodman, manages to create thoughtful posts on a daily basis. Today is no exception, though I must disagree with his post on the use of sabermetrics and what the book Moneyball means to baseball. I do not mean to pick on Tom, but I think his post is very representative of the traditionalist/anti-sabermetrics/anti-Moneyball school of thought. It isn't so much that I have a problem with traditionalists disagreeing with sabermetrics, the problem I have is when traditionalists disagree with sabermetrics only after they have mis-characterized it.

The Moneyball Philosophy
Traditionalists are mistaken on the nature of the teams sabermetrically inclined organizations try to build. It seems as if traditionalists believe that the A's are only interested in acquiring players with quirky wrinkles in their statistical histories. But that is not the case. If you were to have a panel of baseball traditionalists name the top ten hitters in baseball today and then asked a panel of sabermetricians to name the top ten hitters in baseball today, I am sure the two panels would name nearly the same ten hitters in nearly the same order. Teams like the A's, Red Sox, Dodgers, and Blue Jays use in-depth statistical analysis as an additional tool to evaluate players, to be used in conjunction with traditional scouting techniques. If Billy Beane and Theo Epstein didn't believe in the value of traditional scouting, they wouldn't retain traditional scouts on their payroll. In interview after interview, sabermetrically labeled GM's reiterate that they are not turning their back on traditional scouting, the use of sabermetrics is just an additional tool they emphasize when building a team.

Theo Epstein and the Red Sox are an example. Tom claims on Swing and a Miss that the Red Sox are only successful because Theo Epstein inherited players such as Pedro Martinez and Manny Ramirez from the previous Red Sox regime and because anyone can figure out Curt Schilling is a pitcher to trade for, Young Theo shouldn't deserve any credit there either. But what traditionalists leave out is that because Epstein used all the tools at his disposal, objective quantifiers and subjective qualifiers, he saw the value in players such as David Ortiz and Kevin Millar, both of which have been significant cogs in the Red Sox wheel over the last two years. Sabermetrics will tell you that all five of these players were valuable entering the 2003 season, whereas traditional baseball analysis will tell you that only Pedro and Manny and Curt were valuable.

Using the baseball traditionalists line of thinking, Noah would have had to build his ark with just a hammer and anything more would be egg-headed. Beane and Epstein would have been willing to let Noah use a hammer, wood glue, caulk, and whatever else he thought would help.

Sample Sizes and Historical Context
Sabermetrics relies on statistics, and statistics carry more and more weight as the sample size gets larger and larger. Statistics characterizing a small sample size mean nearly nothing, but statistics describing hundreds or even thousands of occurrences can carry serious weight. A .500 batting average after ten at-bats will garner little attention, but a career average of .285 after 5,000 at-bats means much more. Ironically, the track record of sabermetrically inclined baseball teams is suffering from judgments made on very small sample sizes: just a few playoff series over just four years. More traditional baseball thinkers point to the failures of the A's the last four years in the playoffs and say that is proof that Billy Beane's Moneyball-thinking does not work in Major League Baseball.

Traditionalists are correct in that an approach to building a team should be judged on the number of championships it acquires, but they are wrong in not allowing the sabermetric movement in baseball to develop over time and see where it leads. For instance, in football, the West Coast Offense has been the approach that the majority of Super Bowl winners have used in the past two decades, and because of that, nearly everyone can agree that it has been - and continues to be - a superior style of offense. But, we've made that judgment after two decades.

When Joe Montana connected with Dwight Clark in the back of the endzone in order to advance to the Super Bowl, the West Coast Offense was merely a blip in the history of the NFL. Sabermetrics is at a similar stage in history. It has taken the accumulation of time for us to appreciate the West Coast Offense, and it will take a similar accumulation of success - or failure - for us to finally understand the impact sabermetrics will have on Major League Baseball. To put it another way, at this point in baseball history after essentially four playoff series, to say that sabermetrics doesn't work in the Bigs is like saying in 1996 that Tiger Woods was never going to be a dominant player on the PGA Tour because he didn't win the first tournament he entered as a professional. And so, it is wrong for traditional baseball thinkers to quickly to shut the door on teams such as the A's, Red Sox, Dodgers, and Blue Jays.


Because I believe in the power of large sample sizes and the objectivity statistics can bring to any endeavor, I tend to lean towards the sabermetric side of the argument. However, I am not drinking from the Kool-Aid so much as to think it is the end-all-be-all for baseball. There are certain things in baseball like baserunning and defense which are very difficult to quantify. And anyone who has played a sport knows that chemistry is important, but measuring it in order to learn to create it is an entirely different story. The one thing I am positive of is that I will enjoy every bit of watching it play out over the coming years.

<< Home

  • 07/01/2004 - 08/01/2004
  • 08/01/2004 - 09/01/2004
  • 09/01/2004 - 10/01/2004
  • 10/01/2004 - 11/01/2004