“Home plate is 17 inches wide, but I ignore the middle 12 inches. I pitch to the two-and-a-half inches on each side.”
I love to dig into the detailed data we have today to test common beliefs or quotes like the one I just reported. Most times this kind of research ends up confirming what decades of baseball knowledge already tell us—and I strongly suspect I’m working on one of such cases. Nevertheless I like to perform analyses like this because they give us the opportunity to quantify more precisely something we already know.
Nobody would challenge the assumption that painting the corners leads to pitching success. I was willing to take the 2.5 inches offered by Spahn and use them for graphical representations in future posts. But I grew curious about how accurate Spahn’s estimate of the “pitcher’s zone” was. Here’s what happened after I put my “testing hat” on.
First, I calculated (Linear Weights) Run Values for pitches in the “Spahn’s zone” and comparing them to pitches thrown at different locations. I took the liberty to add another 2.5 inches on both sides of the plate to the Spahn’s zone (obviously the 2.5 inches just in/out-side of the plate). Here are the values for all the pitches recorded in 2008 by the PitchF/x system.
As we would have expected, pitchers are more successful working on the corners (negative numbers mean runs prevented by the pitchers).
Following are the run values split by batter handedness.
Again, nothing surprising. What we see is that the inside part of the Spahn’s zone is not as valuable to the pitcher as the outside part. We know that good power hitters love to crush inside pitches and deposit them into the stands.
The charts that follow show how the importance of hitting the corners changes for different pitches. Since in 2008 the MLBAM classifying algorithm wasn’t very good at discerning four-seamers from two-seamers, I grouped them together as fastballs, though I acknowledge that painting the corners with sinkers might have an importance not equal to doing it with four-seam fastballs.
Looking at these data it appears that sliders need to be placed on the corners more than curves—we were expecting this, too, since curves rely more on the vertical component of their location. We also see that change-ups are to be placed only on the outside corner.
Until now we have taken Spahn’s words (and numbers) for granted and split the horizontal coordinates of the pitches accordingly. Next I’d like to see if he nailed the right dimension when he picked 2.5 inches as the border where pitchers make their living.
The following chart depicts the run values for the fastball in a continuum, instead of in five separate zones.
Here the lower the line, the better for the pitcher. It’s hard to think that Warren Spahn could have been more accurate when picking his numbers: maybe, for the right-handed batters you have more room than the 2.5 inches on the outside part of the plate and something less on the inside part, but we can say that he nailed it.
Let’s see the other main pitches.
Even better for the slider. Here the difference between inside corner and outside corner is not very high; both the slider in and the backdoor are effective against RHBs as well as against LHBs.
The plot for the curve appears a lot flatter. As we anticipated, for this pitch the vertical component (totally ignored in this article) is the key.
And finally the change-up. Here we see that only the outside of the Spahn’s zone is good for the pitcher; actually the change-up achieves the greatest success way off the plate.
According to the Neyer/James Guide to pitchers, Spahn broke into the big leagues with his arsenal consisting of fastball (plus sinker), curve and change. Later in his career, he added a screwball (that may actually have been a circle change), a slider and a knuckleball. (A palm ball also appears in his listed repertoire).
I don’t know when he proffered the sentence that opens this post; chances are after he added the slider.
References & Resources
John Walsh wrote a tremendous article here at THT on how you figure the run value of a single pitch. You can read it here.