Fielding Symposium Part I: Baseball’s Next Statistical Revolution?

30813235

In Moneyball, Michael Lewis chronicles Oakland A’s general manger Billy Beane’s use of unconventional statistics (sabermetrics, using baseball-speak) to field a competitive team despite Oakland’s small budget.  The A’s found that statistics like on-base percentage (OBP) and slugging percentage were undervalued on the open-market, allowing Beane to sign and trade for better players at a lower cost. Due to the popularity of Moneyball and the increased popularity of sabermetrics more generally, these statistics are not nearly as undervalued as they used to be. So, what’s a general manager like Beane to do now? What methods of evaluating players are currently undervalued?


One answer has to lie in fielding. One of the more interesting insights of Moneyball was how much teams erroneously relied on traditional scouts’ conventional wisdom that was contradicted by insights from statistical research. Evaluations of fielding have depended on scouting because of the lack of meaningful fielding statistics. Until recently, the main statistic used to evaluate fielders was the atrocious fielding percentage (FP). FP measures the percentage of times a hit ball or a thrown ball is properly handled by a fielder. Statistically speaking, it is the sum of total putouts and assists divided by total chances, which constitute putouts, assists, and errors. The obvious problem with this statistic is it fails to account for a player’s range. A slow centerfielder may commit few errors during the season but may let many more balls drop for hits than a speedy centerfielder. This is not accounted for at all by fielding percentage.

Luckily, a better fielding statistic—zone rating (ZR)—has gained popularity in recent years. Like fielding percentage, zone rating divides total chances by total opportunities. But, a predefined zone on the field for each position—rather than a combination of plays made plus errors—defines total opportunities. Total chances simply consist of the plays made within the zone. There have been some new subtleties added to ZR, such as a better method of accounting for plays made outside of the zone. Yet, what’s relevant is that ZR accounts for a player’s range much better than FP. The zones were determined fairly reasonably, based on areas in the field in which outs are converted more than fifty percent of time. ZR has revealed that certain players whose fielding was rated highly through conventional baseball fielding wisdom were actually quite overrated.

There are, admittedly, drawbacks to ZR. The most obvious one is that it doesn’t account for defensive positioning. If a manger likes to play his rightfielder deep, he is less likely to reach bloop singles than an outfield playing shallow. This has little to do with the rightfielder’s fielding ability and more to due with the manager’s strategic choices.* Moreover, especially in the outfield, different parks play differently. ZR does not account for this shift in difficulty and may punish outfielders that play in tougher parks.

*This actually highlights another interesting point: Fielding is easier to measure statistically for teams than for individual players. You don’t need to worry about factors like positioning since that is an inherent part of team fielding. If a manger positions the players poorly, then they are going to convert fewer outs and this will be reflected in the fielding statistics. Interestingly, according to Baseball Prospectus’s team fielding statistic (defensive efficiency), the World Series champions since 2001 have placed 4,1, 16, 14, 2, 7, 2, 10 in team fielding.

However, The New York Times reports an exciting technological development for evaluating fielding:

“A new camera and software system in its final testing phases will record the exact speed and location of the ball and every player on the field, allowing the most digitized of sports to be overrun anew by hundreds of innovative statistics that will rate players more accurately, almost certainly affect their compensation and perhaps alter how the game itself is played… In San Francisco, four high-resolution cameras sit on light towers 162 feet up, capturing everything that happens on the field in three dimensions and wiring it to a control room below. Software tools determine which movements are the ball, which are fielders and runners, and which are passing seagulls. More than two million meaningful location points are recorded per game.”

This is a big deal. This technology could account for defensive positioning and range in a much more precise manner. Moreover, because it can measure speed, it can be used to tell the distinction between a hard-hit ground ball and a weakly hit one, even if they are hit to the same zone; the same holds for different types of line drives and fly balls. Interestingly, they are hoping to make the data accessible so the statistically minded can play around with the data and develop better statistical measures. This could revolutionize fielding statistics; it should be particularly interesting to see if one dominant statistic evolves to replace ZR or if different teams use different statistics. A small market team that pioneers a particular fielding statistic may have a chance to compete, just like the A’s could with OBP and slugging percentage. The fact that winning teams have generally been towards the top in terms of team fielding statistics gives credence to the belief that fielding matters. At the very least, though, fielding—by necessity—has been undervalued due to the lack of refined and accurate ways to gauge it statistically. This is about to change.

Stay tuned for John S’s reply…

5 responses to this post.

  1. Posted by Doc on July 16, 2009 at 11:19 AM

    Keith Hernandez is the best fielding first baseman ever and should be in the Hall of Fame. I would love to see how he would be rated with the new stats and technology. Compare his hitting stats to third baseman Brooks Robinson who made the HOF solely for his fielding. Each was an MVP, but Keith hit over .300 seven times while Brooks did it once. Both played power-hitting positions and were not strong in that department, but Robinson was better thought of by the press. Keith was one of the first to admit cocaine use and unfortunately that affected the voting. The guy was also the best clutch hitter the Mets ever had.

    Reply

  2. […] Lewis’s seminal sports book, Moneyball,  played a pretty big role in previous posts by Josh and John […]

    Reply

  3. […] Moneyball has had on baseball. OPS shifted from undervalued to properly or even overvalued (and, you know what’s next) and teams continued to hire Art Howe (well, that wasn’t a good thing). More than simply […]

    Reply

  4. […] technology is certainly something that I look forward to seeing develop in the next decade. …Baseball’s Next Statistical Revolution. …How Twitter, Facebook, and Google Develop: Twitter, Facebook, and Google are all huge. […]

    Reply

  5. […] and statistical analysis are two NPI (or Josh) favorites. Michael Lander combines them, performing a statistical analysis of […]

    Reply

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: