We spent a lot of time last season trying to determine what sort of effect Joe Kerrigan had on the Pirates' pitching staff in his first season on the job. A lot of that focused on Ross Ohlendorf, who Kerrigan worked with a lot in spring training, and how the velocity on his fastball changed during the season.
This year his big project is likely going to be Kevin Hart, who's already been tabbed by the coaching staff as a early favorite for the fifth starter role, despite his rough finish with the Pirates in 2009. One reason for that is that Kerrigan's already been working with Hart extensively; he told a Bradenton paper today that Hart's mechanics were a mess when he came to the Pirates from the Cubs. According to Kerrigan his release point was all over the place, varying between the stretch and the windup, and while they worked to fix his mechanics his performance suffered.
You have probably guessed by now that my initial reaction to reading this story was to try and load up Excel with Kevin Hart PitchFX data to determine if we could really detect a difference in his release point. This wouldn't be easy (we'd have to use all data from one park, say Wrigley Field since Kerrigan supposedly worked on fixing this problem, and then try and decide exactly what "different" means in this situation), but apparently MLBAM, in all of their infinite wisdom, changed something about the way that PitchFX data is served to the public and it screwed with Brooks Baseball, my go-to PitchFX site. It looks like Dan's working on sorting through the problem, but the expanded tables that I use to make my own charts are unavailable for now and since that's what I need to really be able to tackle this (there are other PitchFX sites, but none that make all of the data accessible in a form I can use and none that let me do the sort of sorting I'd want to do to approach this question) we'll just have to look at something else.
The next best way to look at this, as far as I can tell, is just to look at Hart's splits. With runners on base in 2009, Hart held hitters to a .238/.332/.375 line. That's not spectacular, but it's certainly effective. It's a little harder to parse strikeouts and walks (BREF only gives me plate appearances instead of innings, making it harder to figure the rates out) and his 27 strikeouts and 21 walks with runners on base (some back of the envelope math says Hart threw the equivalent of around 40-45 innings with runners on base) clearly aren't great, but that pitcher is certainly better than the one we saw in Pittsburgh down the stretch.
What's alarming is his line with the bases empty in 2009. Hitters hit .373/.453/.595 while Hart was throwing from the windup. His strikeouts and walks aren't very different (25 and 23 in 180 plate appearances, vs. 27/21 in 193 from the stretch), but hitters just teed off on him when no one was on base. In the story, Hart says that it's the windup that he was struggling with and in this case, the numbers certainly bear that out. He attributes it to switching between starting and pitching out of the bullpen, which kind of makes sense. Entering games in the middle of innings with guys on base means more pitching out of the stretch. More pitching out of the stretch could throw his windup mechanics all out of whack and lead to the disaster that we saw down the stretch.
This is just my own speculation, but I wonder if Hart was tipping his pitches from the windup. Less than 200 plate appearances in either direction from one season doesn't provide us with a huge sample size, but the biggest difference between his lines with runners on and without runners on isn't walk rate, it's that hitters pounded the crap out of him when the bases were empty. That, to me, makes it seem more like hitters just weren't fooled by what he was throwing. Whatever the case, there's certainly something for Kerrigan to work with here and it's something worth watching throughout camp.
Bradenton Herald link via BBTF