One of the most interesting things that have come up as a part of the early Marlins season is the changes in the profile of Clay Hensley. When he first broke into the league with San Diego, he was a soft-tossing right-hander with a mediocre control but ability to get ground balls. At some point between 2006 and 2008, he went from mediocre to awful control and lost any semblance of effectiveness, even out of the bullpen. After spending some time in the minors and pitching decently (and in the same fashion as before, might I add), he finally got another opportunity with the Marlins after a solid spring campaign. Initially pegged for the fifth starter spot, Hensley was moved to the bullpen after the team acquired Nate Robertson.
In his bullpen role, Hensley has succeeded in a vastly different way than he had before. After being a groundout guy with very little else of interest, he has all of a sudden turned into a strikeout machine. Hensley has struck out 27 of 77 batters faced, an astonishing 35.1% rate. How is he doing it? What are the changes he’s employing to get these results. Maybe Pitch f/x has some of the answers for us.
Here’s what I heard
Dave McGrath and I talked about this earlier today, and he had mentioned that during the TV broadcasts, Rich Waltz and Tommy Hutton had made mention of the drastic change by Hensley, attributing it in part to a renewed changeup similar to that thrown by Tim Lincecum. Has this new changeup been the difference?
Pitch f/x only reliably (or as reliable as it gets) goes back to 2008 (it wasn’t installed everywhere in 2007 and there were many bugs to fix), but luckily for us Hensley pitched 39 1/3 innings that year out of the bullpen. He wasn’t very good, putting up a 4.47 FIP (5.09 xFIP) and a 5.31 ERA. I took a look at his pitches that season and compared them to his pitches so far this year. First, let’s start with the breakdown of pitches thrown.
|Pitch Type||2008 Count||2008 %||2010 Count*||2008 %*|
* Four-seam fastball and sinker classifications were even for 2010, but were lumped together as fastballs
There are some pretty drastic differences in percentage use there. Based on these changes, you might think that the drop in slider usage and increase in the use of other secondary pitches may have a big thing to do with his success. But how were those secondary pitches? Here’s what they looked like charted in a horizontal vs. vertical movement chart.
Obviously there are a lot fewer pitches in 2010, but you can see that the classifications are generally correct in the change in usage. There is a lot more blue and a lot fewer pitches closer to the neutral region (0 vertical and horizontal break) that are typically considered sliders. The change is real. In addition, you can see that there is so far a slight change in the movement of the changeup. FanGraphs collects Pitch f/x data as well and has the change’s average vertical break at 2.5 inches, meaning the pitch has more sink than it use to have in 2008 (4.4 inch average vertical break).
I don’t know how much difference this break is having, but I can tell you that the results of his changeup are a lot better.
|–||2008 CH||2010 CH|
Does that explain it? Maybe some of it. For some reason, Hensley is having more success with his changeup than in years past. He’s putting more of them in the zone, hitters are taking more of them in the zone, swinging at a lot more of them out of the zone, and whiffing more often. But what about the pitch he dropped? Take a look at the slider from 2008.
One thing I see from this is that Hensley had a difficult time controlling his slider. Sliders should stay away from the zone, so the zone% was not surprising. However, check out the chase rate (pitches swung at out of the zone out of total pitches out of the zone). You will note that it was not all that much better than the changeup rate, while getting far fewer called strikes than the change. Maybe Hensley realized that he was having too much difficulty controlling the slider and that it was not getting proper results in terms of chases and swings.
What about the loss of ground balls? How can we explain that precipitous drop? Well, without looking at the stats, I can give a pretty good guess. Grounders are primarily induced by fastballs, typically sinker-type pitches. None of Hensley’s breaking balls produced more than 45% GB% on balls in play. With the drop in fastball usage, naturally we would expect Hensley’s ground ball rates to drop as well.
Surprisingly, the change in usage by Hensley actually does explain much of the difference in his play. However, it doesn’t mean that this difference will continue. His 2010 changeup is unlikely to stay this good, and all of his pitches will likely regress a bit in performance. Still, the move towards more breaking pitches, I feel, has to have something to do with the newfound strikeout stuff. I’ll investigate more when I get the time (and more sample size for Hensley).