Pitching Outside the Box, Literally

So long as you didn’t bring the party too hard over the Christmas holiday, there’s a chance you remember the article I submitted for the readership’s consideration last week. In said article, I roundly praised research conducted by Lookout Landing’s Jeff Sullivan this past August — research in which he explored the relationship between pitcher contact rates and strikeouts. Moreover, I posted a Top 10 Leaderboard of the starting pitchers (50 IP and up) with the best Contact%.

Well, in the comments section of said article, user Toffer Peak brought to our collective attention a study done by user matthan over at DRaysBay. Matthan is the user name of Matt Hanna, and his work is an exciting complement to Sullivan’s as it gives us some idea of the importance of Out-of-Zone Swinging Strikes (OZSwStr%) relative to In-Zone Swinging Strikes (InZSwStr%).

The relevant article provides all the answers your little heart would desire — complete with a Google spreadsheet of every pitcher from last year — but the relevant content for our purposes is this formula that Hanna concludes is the best fit for calculating Expected Strikeout Percentage (eK%). Said formula goes:

eK%=(ClStr%*.9)+(Foul%*.5)+(InPly%*-.9)+(InZSwStr%*1.1)+(OZSwStr%*1.5)

The Adjusted R-Squared is: 91.4%

The surprising result here is the degree to which OZSwStr% is weighted over and above InZSwStr%. Nor does that even account for the fact that the average for OZSwStr% (4.89%) is already about twice as a high as InZSwStr% (2.73%).

Once we adjust for that difference as well, OZSwStr% comes out to roughly 2.5 times more important than InZSwStr%. If I’m being honest, I’ll say right now that that runs counter to what I would’ve guessed. My impression has always been, if a pitcher can throw a pitch past a swinging batter but still place said pitch within the strike zone, then he (i.e. the pitcher) would be truly unhittable. What Hanna’s research suggests is quite the opposite, in fact: A pitcher who is able to induce swings (and misses) at pitches out of the zone is, in fact, most likely to tally big strikeout numbers.

This research is quite relevant to the present interweb site, as FanGraphs carries both O-Contact% and Z-Contact% on every player page and in the leaderboards section.

And though, much like Forrest Gump, I’m not a smart man, I thought it might make sense to create a leaderboard in which O-Contact% (or OZSwStr%) was given its due. To that end, what follows is a Top 10 list of the starting pitchers with 50+ IP who led the league in what I’ll call Adjusted K. In fact, what I did was to find how many standard deviations all such pitchers were from the mean in both O-Contact% and Z-Contact%. I then multiplied the O-Contact% standard deviation by 2.44 and averaged it with the Z-Contact% standard deviation. Here are the results (SDO = Standard Deviations from O-Contact% mean / SDZ = Standard Deviations from Z-Contact% mean):

This list greatly resembles the one we looked at last week — with one exception, that is: Freddy Bloody Garcia. Granted, he only pitched 56 IP through nine starts last year, but it appears to be a skill he’s carried throughout his career, as his 46.5% lifetime mark suggests.

So that’s one thing. Now here’s another question of some interest, I think: Which pitchers posted the biggest O-Contact%/Z-Contact% splits in 2009? In other words, which pitchers are best at getting swinging strikes outside the strike zone despite allowing somewhat frequent contact within it. Truly, this would be a list of pitchers who use their talents most efficiently, getting swings and misses outside of the zone, where they are more valuable. Here’s a list of such pitchers (SD O-Z = Standard Deviation of O-Contact% minus the standard deviation of Z-Contact%):

Some of those guys are what you might describe as a pretty big deal. Carpenter and Wainwright, certainly, were at least part of the Cy Young convo in the National League — and both accomplished the feat while conceding a below-average contact rate on balls in the strike zone.

There are certainly other questions to ask about this work. I’m in Paris right now, though, so I’m probably not gonna ask them for at least a couple days.