Analyzing Jeremy Sowers, Part III

It is quite likely that you are at least skeptical of Voros McCracken’s DIPS theory (or perhaps you simply don’t believe it). Perhaps Sowers is simply an exception, assuming the theory is even true on the first place.

One theory put forth as a possible explanation of “exceptions” to the DIPS theory is that pitchers who strike a lot of batters out might also be able to get lots of weakly hit balls that are easily converted into outs. Whether or not this is true, it surely does not relate to Sowers. Sowers’s stuff is barely even average. His fastball sits around 88, and he lacks any plus-plus pitches. Furthermore, his strikeout rates reflect this---in an at-bat, Sowers is not overwhelming a hitter with his stuff.

Sowers struck out only 3.57 batters per nine innings, a figure that ranks him third-to-last in all of baseball. The only pitchers who struck out fewer batters were Carlos Silva (who had a 5.94 ERA, a walk rate much lower than Sowers’s, and generated 30% more ground balls than fly balls) and Chien-Ming Wang (who survived on his ability to generate 306% more ground balls than fly balls).

Okay, so Jeremy Sowers is not a power pitcher. If you’ve ever seen him pitch, you don’t need me to tell you this. Perhaps we can compare him to other “crafty” or “finesse” pitchers who seem to thrive on getting weakly-hit balls. Greg Maddux, Tom Glavine, and Jamie Moyer come to mind as pitchers that fit this mold. In fact, I’ve even heard Sowers compared to Glavine. Here are their BABIPs for the last three seasons:

Maddux: .261, .299, .290

Glavine: .284, .288, .318

Moyer: .268, .296, .327

Fairly random, and once again, all higher than Sowers’s BABIP. Furthermore, despite his reputation of being “crafty”, Tom Glavine has struck out 5.38 batters per nine innings. Glavine’s career BABIP-against is .282. Although he does seem to be good at controlling balls that are put in to play, his BABIP is still significantly higher than Sowers’s. Maddux’s career BABIP is .283. Moyer’s is .284. All are very good. All are much higher than Jeremy Sowers’s .256 with the Indians, or .263 with Buffalo.

So what do we make of a pitcher who doesn’t strike anyone out and gives up far fewer hits than we expect him to? In short, he was lucky. The sample size was small, and over a longer period of time regression to the mean would occur, and Sowers’s ERA would be a lot higher. This might indeed be the case.

However, Sowers’s track record is still short. ERA aside, his 2005 and 2006 were very different years: he was legitimately excellent in 2005, striking out many, walking few, and giving up few homers. In 2006, he also gave up few homers, but he didn’t strike out as many and walked more.

Sowers will certainly survive on his command and control (yes, these are different). He’s likely an exception to the general rule that power pitchers are the best kind of pitchers. However, his apparent ability to prevent balls in play from becoming hits is at least somewhat of a fluke. Thus, in order to maintain success going forward, he will need to preserve his ability to prevent homers, and increase his ability to strike batters out. If he is able to do this, he will likely be a very good #3 starter in the majors, a guy who you can expect to pitch 200 innings and maintain an ERA of around 4.00-4.50 (perhaps a little bit better if he can really improve his K-rate).

However, because much of his 2006 was attributable to luck, Sowers is a prime candidate for the infamous Sophomore Slump, which, in actuality, is regression to the mean. His long-term prognosis is excellent, but I fear that expectations are going to be unjustly high for 2007.