Tracking Kuroda’s splitter usage

After a somewhat slow start to 2012 that was perhaps overly scrutinized, and could easily be attributed to changing leagues, Hiroki Kuroda has been very successful of late.  In the month of June, Kuroda has only once failed to complete 7 innings, with that one failure a 6-inning outing against the Braves.  In June has pitched to a 2.38 ERA, with 32 strikeouts against 9 walks in 34 innings, and opposing hitters have managed just a .641 OPS.  This is the pitcher that the Yankees hoped they were getting when they signed Kuroda, and so far he hasn’t disappointed.

His most recent outing against the Indians was particularly impressive, as he gave up just 1 run in 7 innings of work, and struck out 7.  Of interest to me was that 6 of the 7 strikeouts came in the splitter.  Going into the season, I knew that Kuroda’s bread and butter pitch was his sinker, but that the splitter was his go-to strikeout pitch.  In the early going, however, Kuroda didn’t seem to be using the splitter as often, and instead, favored the slider.  I’m not sure if it was a matter of not getting ahead in counts (where the splitter may be a more effective offering) or if Kuroda was just not as comfortable with the pitch, but the difference in usage was striking.

I decided to turn to Brooks Baseball to see if the data supported my recollection.  Overall this season, Kuroda has thrown splitters 12 percent of the time.  However, the usage has fluctuated somewhat game to game (as seen if the chart below.  Don’t mind the dates on the x-axis, Excel was being stubborn and I didn’t have time to fix them).

The data do suggest that Kurdoa’s splitter usage has gradually increased over the course of the season, and that it was particularly low at the beginning.  There are 2 outings (on the high end) that are worthy of note, May 5 and last night.  On May 5, Kuroda threw a whopping 27 splitters on a day when he threw only 91 pitches (nearly 30% usage).  While Kuroda used the splitter frequently in that game, it was not very effective, as it induced only 3 whiffs (11%) and a strike rate of about 50%.  The pitch was particularly effective last night, as it had a whiff rate of 38 percent, and was thrown for a strike more than 75 percent of the time.

The differential success could be in part a factor of the competition, but perhaps there was something else at work.  On May 5, the splitter had 4.21 inches of vertical break and -5.01 inches of horizontal break.  By comparison, last night’s splitter had 1.43 inches of vertical break and -7.01 inches of horizontal break, a pretty substantial difference.  It is possible that the increased success of the pitch last night was due to the increased horizontal break, as the additional movement in that direction might have fooled hitters.

Additionally, it is possible that vertical break had an impact: May 5 had less than a 3.5 inch difference in vertical break between the fastball and the splitter, while the difference last night was nearly 9 inches.  The increased difference in break between fastball and slider could also make the splitter a more effective pitch, preventing hitters who are sitting fastball from making contact with the pitch.

From this analysis, it is clear that Kuroda wasn’t just using the splitter last night than he was earlier in the season (with the exception of May 5), he was also using it more effectively.  This could be a factor of warmer weather, comfort, or some other factor, and I will definitely be watching to see if his usage fluctuates again over the course of the season.  Kuroda is definitely at his best when the splitter is working as a strikeout pitch, though he has shown the ability to be effective without his best splitter.