Getting defensive

Defensive Efficiency Rating (DER) measures the percentage of balls in play that are converted into outs by a team’s defense. The better the defense, the higher the team’s DER, and the lower their batting average on balls in play (BABIP) against. DER + BABIP = 1.

In 2006, the Indians were 25^th in baseball, posting a DER of .686---meaning their BABIP was a very high .314 (remember, league average is generally around .300). This would explain a lot of their pitching problems in 2006, especially considering that they were 29^th in pitching strikeouts (only the Royals had fewer). So the Tribe had a ton of balls in play and an inordinate amount of them became hits. That’s not a good recipe for success.

However, in 2005, the Indians ranked 3^rd in DER, at .720 (meaning their BABIP was .280). Their 2005 pitching staff was pretty similar to their 2006 staff; more importantly, their 2005 and 2006 defenses were very similar.

Grady Sizemore (CF), Casey Blake (RF), Jhonny Peralta (SS), Ronnie Belliard (2B), Aaron Boone (3B), and Victor Martinez (C) played most of both seasons at their respective positions. The main change came in left field, where Jason Michaels replaced Coco Crisp; however, Michaels was a centerfielder by trade and had a reputation of being a solid defender. So what happened to raise the Indians’ BABIP from .280 to .314 in one season with basically the same defense?

Well, there are a couple of possibilities: 1) the pitching staff got tremendously worse; or 2) the defenders simply did not perform as well.

Let’s examine the first idea first. Surely, the Indians’ pitching was not nearly as good in 2006 as it was in 2005. In 2006, their staff ranked 6^th in the AL with a 4.41 ERA. The starters were 3^rd in the league with a 4.31 ERA, but the relievers (as you likely know) were awful, finishing 11^th in the AL with a 4.66 ERA.

In 2005, the Tribe was tied for first in the AL with a 3.61 ERA. Their starters were 5^th, with a 3.96 ERA, and their bullpen was 1^st, with a 2.80 ERA. Despite Kevin Millwood’s departure and Jason Johnson’s awful stint in 2006, the Indians’ starters remained very good, thanks to CC Sabathia’s improvement, Jake Westbrook’s regression to the mean (in a good way), and Jeremy Sowers’s strong performance.

The bullpen was definitely a lot worse. However, “worse” does not necessarily equal giving up more hits on balls in play; as we know, generally this is a luck-based statistic. Therefore, while the bullpen was definitely worse, this does not account for the change in BABIP from 2005 to 2006.

More likely, it would seem that the Indians’s defense in 2006 was a lot worse than their defense in 2005. If this is true, we’d expect to see a significant difference in the BABIP’s of pitchers who were on both the 2005 and 2006 staffs. Let’s see:

(2005 BABIP is listed first, then 2006 BABIP)

CC Sabathia: 284-291

Jake Westbrook: 287-322

Cliff Lee: 277-296

Bob Wickman: 269-292

Rafael Betancourt: 294-273

Jason Davis: 325-345

Fernando Cabrera: 277-293

As you can see, of the pitchers who logged a significant amount of innings in front of both defenses, only one (Betancourt) had a better BABIP in 2006. It looks as if there was a significant difference in defense.

How did the defense get that much worse? Two reasons: 1) many of the players performed a lot worse in 2006 than they did in 2005; and 2) Coco Crisp performed much better in 2005 than Jason Michaels did in 2006. Let’s evaluate.

Baseball Prospectus offers a defensive stat called Fielding Runs Above Replacement (FRAR). This is defined as “A fielding statistic, where a replacement player is meant to be approximately equal to the lowest-ranking player at that position, fielding wise, in the majors. Average players at different positions have different FRAR values, which depend on the defensive value of the position; an average shortstop has more FRAR than an average left fielder.” Thus, it is important not to compare the FRAR of two different positions, but it can be important to compare two different players at the same position.

We can look at FRAR and hopefully see what changed in the Tribe’s defense from 2005 to 2006.

I will list the player, their position, their 2005 FRAR, and then their 2006 FRAR

Grady Sizemore (CF): 17-19

Casey Blake (RF): 18-10

Aaron Boone (3B): 15-2

Jhonny Peralta (SS): 34-46

Ronnie Belliard (2B): 33-15

Ben Broussard (1B): 7-2

Victor Martinez (C): 22-7

Coco Crisp (LF) 2005 only: 22

Jason Michaels (LF) 2005 only: 1

Some other notable numbers:

Andy Marte (3B) 2006 only: 7

Ryan Garko (1B) 2006 only: 1

Josh Barfield (2B) 2005 only: 32

It looks as if Casey Blake, Aaron Boone, Ronnie Belliard, and Victor Martinez suffered considerable declines in their defense from 2005 to 2006 (interestingly, Blake, Boone, and Belliard are all on the wrong side of 30, and Martinez, while only 27, has been catching for awhile). Furthermore, it would appear that the downgrade from Coco Crisp to Jason Michaels was considerable.

So what does this mean going forward? Well, first of all, it is reasonable to assume that, like last year, the Indians won’t walk many batters in 2007 (they finished with the second fewest walks of any team in 2006). Paul Byrd and Jeremy Sowers rely on not walking anyone; CC Sabathia, Cliff Lee, and Jake Westbrook all have very modest walk-rates. No one in the bullpen has huge problems with walking too many batters. On the flip side, only CC has an above-average strikeout rate, while Westbrook, Byrd, and Sowers all strike out very few hitters. Plus, the bullpen has many pitchers who rely on batters putting the ball in play. All in all, the Indians pitching staff can expect a lot of balls to be put in play again next year.

Thus, their defense is going to be extraordinarily important. If the Tribe can curb their apparent bad luck and get back into the top half of DER, their pitching staff should improve tremendously, as they will give up far fewer hits.

Are they likely to improve in 2007? For starters, it’d be tough to get much worse. Even if their defense was legitimately bad, it’s tough to maintain a 14%-below-average rate over two seasons; some regression to the mean is likely due. Furthermore, several of the culprits of poor defense last season---Belliard and Boone---are going to replaced by younger players who are likely to be better defenders (Barfield and Marte). Although Jason Michaels was a poor defender in 2006 for the Tribe in left field, he was quite solid in 2005 for Philadelphia in center field, so there might still be hope for him. Grady Sizemore and Jhonny Peralta actually improved from 2005 to 2006, and the Indians have challenged Peralta to improve his defense even further. Casey Blake is likely on the decline, but his potential replacements, Shin-Soo Choo and Franklin Gutierrez, both have excellent defensive reputations.

The Indians are unlikely to be as bad defensively in 2007 as they were in 2006. If they regress to the mean even a little, they will improve a lot; there remains a chance that not only will the defense improve, but it will be well-above average. Either way, the Tribe’s defense is unlikely to be as big of a liability next year.