The Fielding Revolution

added 3/17/2006 by Scott Barzilla

There has been no greater chasm between statheads and traditionalists than the question of fielding. A little more than five years ago, most of the stat luminaries were telling us that fielding accounted for two percent of the variance of runs allowed. Of course, the traditionalists scoffed at that notion and rightfully so. Ironically, the work of Voros McCracken spurned a revolution in fielding metrics that has culminated in a book called “The Fielding Bible”.

Of course, the Fielding Bible reveals that many traditionalists are wrong about who the great fielders are (at some positions), but they were right about how important fielding was in the grand scheme of things. However, the statistical approach to fielding could bring more improvement in that area on the big league level than traditional scouting did. Teams already subscribe to the Fielding Bible and now it is available to general public. Let’s consider the performance of the hometown nine this past season.

               Runs Allowed       Defensive Eff.
Astros 2004        698                0.687
Astros 2005        609                0.706

This seems minor, but what this means is that Astros fielders got to more than 70% of the balls in play in 2005 versus nearly 69% in 2004. That one percent of difference played a big role in the number of runs they allowed. When you think about it, the staffs were pretty much the same from one year to the next. Sure, a healthy Pettitte and the success of Weaver and Qualls played a role in that, but the defense did as well.

In the race between traditionalists, stat geeks, and pundits, the pundits are usually the last one to get anything. Whenever you hear the good folks at ESPN or Fox talk about a transaction they usually gloss over the affects the move will have on fielding. However, we can’t ignore the biggest question in camp and the affect it will have on fielding. Will Bagwell be able to go and what affect will that have on the club’s defense. We can look at fielding runs above average from Baseball Prospectus to give us a clue.

                    2003    2004   2005   Total   AVG
Brad Ausmus         14        1      8     23     7.7 
Jeff Bagwell        -5      -13     -1    -19    -6.3
Craig Biggio       -15      -16     -4    -35   -11.7
Morgan Ensberg       9      -12      8      5     1.7
Adam Everett         2        5      6     13     4.3
Lance Berkman        3       -7     -2     -6    -2.0
Preston Wilson     -13       -2    -11    -26    -8.7
Jason Lane           0        1     -4     -3    -1.0

With Bagwell         -       -      -     -48   -16.0 

This is definitely not the way we want to go on defense. Ironically, you can see how many of the same fielders performed worse in 2004 than 2005. Biggio’s numbers correspond to his last two seasons as a second basemen, so those aren’t quite fair, but give us a ballpark of where he could end up in 2006. The interesting folks are Morgan Ensberg and Lance Berkman. I guess we know that fielding slumps are just as prevalent as hitting slumps. Let’s see what happens when we take Bagwell out and put Taveras in.

                    2003    2004    2005    Total    AVG
Brad Ausmus          14       1       8      23      7.7 
Lance Berkman         0       0      -4      -4     -1.3
Craig Biggio        -15     -16      -4     -35    -11.7
Morgan Ensberg        9     -12       8       5      1.7
Adam Everett          2       5       6      13      4.3
Preston Wilson        0       1       0       1      0.3
Wily Taveras          0       0      15      15     15.0
Jason Lane            0       1      -4      -3     -1.0  	  

Without Bagwell       -       -       -      17      5.7

In other words, the change creates a pretty considerable run differential (21.7 runs). Unfortunately, this isn’t the end of the story. If things were this easy Bagwell would be packing his bags already. Offense is still a considerable part of the equation. I’m still dumbfounded by the number of people who suggest we shouldn’t worry about the offensive production we get from certain positions.

At the end of camp, the Astros brass will have to decide whether the offensive production they will get from Bagwell (since the lineup would effectively only have that one change) is great enough to offset the loss in fielding. Since Bagwell is in decline this will be a daunting question to answer. It’s also daunting because Taveras’s offensive production can go either way.

Overall Observations

Those of us in the stat world have been accused of a lot of things. In many ways I find the hostility to be odd. At the end of the day we are all looking for truth and we all make mistakes. I can fake amnesia and pretend that I didn’t say that fielding was unimportant, but my detractors already have the quote lined up. If I admit that I was wrong I can move in this pursuit.

The release of the Fielding Bible doesn’t remove the debate, it just crystallizes it. The folks that released that are far smarter than most of us and they deserve a round of applause. What each of us is looking for is an integration between offense and defense that we can use to accurately rate players overall. I guarantee that the Astros don’t gloss over Adam Everett and Brad Ausmus’s offense anymore than they gloss over Lance Berkman or Craig Biggio’s defense.

Simply put, they have a different rating system than we do. Some teams are open to the new methodologies while others are not. What gets lost here is that it took the sabermetrical community nearly twenty years to develop offensive metrics that became generally accepted inside baseball. The question for baseball teams is whether they want to follow the progress now or wait until the results are done. Some are already subscribing to the Fielding Bible and their clubs are using some of the information inside. Sometimes the sabermetrical community isn’t ahead of the curve, but like the technology industry, the stat community moves pretty fast.