Like David Aardsma, Shawn Kelley was expected to provide a solid presence in the bullpen during the later innings. Joined with Brandon League and Mark Lowe, the foursome had the makings of an above average core of hard throwing righties to close out games with. No unit on the Mariners has escaped disappointment in 2010 and the bullpen is one of the bigger ones. I have already covered Aardsma, so I turn to Kelley here.
A major reason I had high hopes for Kelley in 2010 stemmed from this post wherein I looked at Kelley's 2009 season before his oblique injury and both his immediate and post-immediate performances after returning from the disabled list. The conclusion that I drew was that Kelley's rough stretch was almost entirely isolated to his time right after he came back from injury; perhaps hinting that he needed more time before being thrust back into Major League work. If anything, Shawn Kelley down the stretch was even better than the Shawn Kelley we saw in April of 2009.
The first thing I decided to do was repeat the final table from that prior article but on 2010 data. Back then I looked at how many of each pitch were called a ball and the percentage of pitches that were either a called or swinging strike (i.e. not a strike due to contact from the hitter). Here is the 2009 data:
Pre: (Through May 5th)
109 FB, 30% / 26%
51 SL, 35% / 32%
Comeback: (July 3rd through August 4th)
126 FB, 35% / 19%
46 SL, 41% / 34%
End: (August 4th onwards)
259 FB, 30% / 27%
110 SL, 28% / 36%
And here is the 2010 data:
294 FB, 35% / 26%
107 SL, 27% / 36%
Kelley is missing more often with his fastball, but is still getting the called and swinging strikes at rates matching his peak. The slider is as good as ever though his usage of it has decreased a small bit as he has actually made use of his change up more often.
Normally, I would say that the rise in balls on his fastball is not that big of a deal because of that stability in his whiff rate, but in this case I think it might have a bigger effect. The single biggest, and only major deviant from 2009 has been in Kelley's walk rate. We all appreciated Kelley so much in 2009 as the one reliever we could count on to throw strikes. He is still throwing a lot of strikes at 66% this season compared to 68% last year, but his walk rate has gone from 6% to 10%. That nearly doubling of walk rate more than drowns out the improved overall whiff rate and corresponding slight bump in strikeouts.
What is pushing the walk rate so much higher when the overall strike rate has only moved a little? First pitch strikes would be my guess. A pitcher's first pitch strike percentage correlates decently (an R^2 around 0.35) with his overall walk rate. Kelley went from a 69% first pitch strike rate in 2009 to a 61% this season. That is a dramatic shift and since pitchers tend to rely on their fastballs to begin most at bats, I believe that Kelley's degrading fastball command might be primarily responsible for causing this large increase in 1-0 counts.