Pitching to zones, part two

Last time I checked in, I was discussing the outer and inner edges of the plate and the frequency with which pitchers throw to those regions. Though I was originally intending to keep going with the corners, today I’ve decided to open it up a lot more—instead of focusing on the edges, I wanted to see how the league average pitcher throws to all different regions of the strike zone.

For this study I set up five “vertical” zones and five “horizontal” zones; horizontal: inside out of the zone, inner third of the zone, middle third, outer third, outside of the zone; vertical: low out of the zone, lower third of the zone, middle third, upper third, high out of the zone*. For the sake of the most straightforward presentation, I’ll only be looking at horizontal or vertical zones and not permutations of the two (so for example, nothing for “low and inside” today).

The numbers, which come from the 2010 season, I’ll be presenting today will be split up by pitch type and opposing batter handedness, two things that are very important when it comes to zone analyses.

*For this study, the horizontal zone is one foot from the center of the plate on both sides. Vertically, I’m using 1.5 feet from the ground to 3.5 feet from the ground, normalized to the batter’s zone using the sz_top and sz_bottom in the PITCHf/x data.

I wanted to have some sort of batter normalization as opposed to having a fixed zone; ideally, I would have batter zones aggregated over all games in order to strip out the fluctuation in the PITCHf/x operators’ zone markings, but I haven’t done that yet. For more about this difficult subject, read Mike Fast’s article on batter height and strike zone.

Anyway, let’s start by looking at four-seam fastballs.

Four-seam fastballs

121,861 (38.1 percent of total pitches) against same-handed batters; 142,086 (37.2 percent) against opposite-handed batters


























Handedness Low OZ Low IZ Middle High IZ High OZ
Same 9.6% 21.1% 31.2% 24.3% 13.9%
Opposite 8.4% 19.9% 31.0% 25.5% 15.2%

























Handedness Inside OZ Inside IZ Middle Outside IZ Outside OZ
Same 8.7% 18.3% 30.8% 27.9% 14.2%
Opposite 9.0% 18.4% 27.7% 27.2% 17.7%

As Dave Allen showed us in 2009, four-seam fastballs do not have a big platoon split, and fastball location seems not to be affected—both in terms of height and width—by the handedness of the batter. I would have expected there to be fewer pitches thrown to opposite-handed batters in the middle of the plate than to same-handed batters, and while it does look like there’s a slight difference, it’s not incredibly significant. For the most part, batter handedness does not have an effect on where four-seam fastballs are thrown.

Two-seam fastballs

56,302 (17.6 percent) against same-handed batters; 73,905 (19.4 percent) against opposite-handed batters


























Handedness Low OZ Low IZ Middle High IZ High OZ
Same 13.5% 27.0% 32.6% 19.2% 7.7%
Opposite 13.2% 26.2% 32.6% 19.9% 8.1%

























Handedness Inside OZ Inside IZ Middle Outside IZ Outside OZ
Same 17.4% 24.6% 28.4% 20.6% 8.9%
Opposite 4.9% 12.4% 25.5% 32.2% 25.0%

There are a number of things I find interesting here. First of all, you can see how batter handedness plays a much bigger role for two-seamer location than it does for four-seamer location. With rare exceptions, all fastballs tail in on a same-handed batter, but two-seamers do more so than four-seamers; because of this, it is natural for pitchers to throw the tailing fastballs at the area of the plate where its movement naturally takes it.

To same-handed batters, there is a higher preponderance of inside two-seamers than outside two-seamers. This makes the pitchers who can perfect the backdoor/frontdoor “Greg Maddux pitch” all the more special.

Another thing I noticed about the two-seam data was that, as I had expected, there are a lot more pitches below and at the low part of the zone than there were for four-seamers. This makes sense, considering that two-seamers are designed to be thrown low in the zone in order to get groundballs.

Cutters

19,893 (6.2 percent) against same-handed batters; 19,890 (5.2 percent) against opposite-handed batters

A Hardball Times Update
Goodbye for now.

























Handedness Low OZ Low IZ Middle High IZ High OZ
Same 16.4% 25.1% 30.7% 19.4% 8.5%
Opposite 12.0% 23.2% 32.1% 21.9% 10.9%

























Handedness Inside OZ Inside IZ Middle Outside IZ Outside OZ
Same 3.3% 10.9% 27.0% 34.6% 24.1%
Opposite 19.8% 28.5% 25.4% 16.5% 9.7%

The cutter is an interesting pitch because it includes those used by pitchers as fastball variants (such as Mariano Rivera) and those used by pitchers as breaking pitches (such as Phil Hughes and Kyle Davies). Here, they’re lumped together into the same “cutter” group.

Cutters are pretty much reverse-sinkers and have some glove-side movement inside on opposite-handed batters. Given this, it makes sense that cutters have the reversed location trends of sinkers. They’re typically going to be inside to opposite-handed batters and outside to same-handed batters.

The backdoor cutter is rare, and the frontdoor cutter is even more rare. My guess would be that pitchers don’t want to get burned on an inside mistake. Also like sinkers, cutters are typically located below or at the bottom of the strike zone.

Switching gears now to offspeed pitches and breaking balls…

Sliders

63,066 (19.7 percent) against same-handed batters; 35,535 (9.3 percent) against opposite-handed batters


























Handedness Low OZ Low IZ Middle High IZ High IZ
Same 26.8% 28.8% 26.3% 13.3% 4.8%
Opposite 25.7% 27.0% 26.2% 14.9% 6.3%

























Handedness Inside OZ Inside IZ Middle Outside IZ Outside OZ
Same 2.4% 9.4% 25.9% 34.0% 28.3%
Opposite 23.4% 28.5% 26.6% 15.0% 6.5%

Like cutters, but with more movement, sliders move down and away from same-handed batters. Because of this, the typical spot from a right-handed slider seems to be down and away to a righty batter or, less frequently, down and in to a lefty batter. Given the similar movement of cutters and sliders, the plate locations are also quite similar, though you’ll see a lot more sliders below the strike zone than cutters below the strike zone.

Changeups

20,058 (6.3 percent) against same-handed batters; 59,232 (15.5 percent) against opposite-handed batters


























Handedness Low OZ Low IZ Middle High IZ High IZ
Same 30.2% 27.5% 24.1% 13.0% 5.3%
Opposite 27.6% 27.7% 25.3% 13.5% 6.0%

























Handedness Inside OZ Inside IZ Middle Outside IZ Outside OZ
Same 9.5% 21.2% 31.7% 25.0% 12.6%
Opposite 2.6% 10.4% 27.0% 33.8% 26.2%

Changeups tail armside, so they typically will be used against opposite-handed batters as a pitch that moves away. Logically, changeups against opposite-handed batters are typically located outside.

Here’s what I find strange, though—when changeups are thrown to same-handed batters (which is not very often), a lot of them are thrown outside, which would mean that the pitch is moving back towards the center of the plate. I would have expected a greater percentage of pitches to be on the inside third or inside off the plate, given that that’s the natural trajectory of the pitch.

Next, let’s look at split-fingered pitches, which are typically used as change variants.

Splitters

3,134 (1.0 percent) against same-handed batters; 6,293 (1.6 percent) against opposite-handed batters


























Handedness Low OZ Low IZ Middle High IZ High OZ
Same 37.7% 26.5% 21.3% 10.5% 4.1%
Opposite 34.7% 26.7% 21.5% 11.6% 5.4%

























Handedness Inside OZ Inside IZ Middle Outside IZ Outside OZ
Same 9.1% 22.0% 31.6% 23.7% 13.7%
Opposite 4.3% 15.1% 29.8% 30.2% 20.5%

Splitters have similar movement to changeups, though they usually have less horizontal tail and more vertical drop. Splitters are also thrown for a ball more than any other pitch type; this can be explained by the fact that 35 percent of them are thrown low and out of the zone.

I won’t dwell on splitters too much, because while they are often considered to be a distinct pitch type, they are more accurately described as a subset of changeups. Also, there isn’t a very large sample size to work with, and most pitchers don’t throw a splitter.

Finally, a look at curveball data.

Curveballs

27,327 (8.6%) against same-handed batters; 35,153 (9.2%) against opposite-handed batters


























Handedness Low OZ Low IZ Middle High IZ High OZ
Same 31.2% 25.6% 23.4% 12.9% 6.9%
Opposite 29.3% 24.5% 24.1% 14.3% 7.8%

























Handedness Inside OZ Inside IZ Middle Outside IZ Outside OZ
Same 5.2% 14.4% 28.7% 29.4% 22.3%
Opposite 8.8% 19.6% 30.0% 25.8% 15.8%

To me, curveballs are the most interesting in this kind of analysis. Like cutters, curves are often used in two different ways. One is to get batters to chase low and out of the zone (A.J. Burnett’s is a good example); another is to pick up called strikes by freezing batters on pitches in the zone (Barry Zito’s hook is designed to do this).

Since curveballs, on average, move gloveside and have the most movement of any pitch, a popular spot for the “looking” curve is the outer part of the plate. The curveball’s movement allows it to look like a ball outside out of the pitcher’s hand before coming back over the plate.

Based on these numbers, that appears to be the preferred approach against opposite-handed batters. It is different for same-handed batters, as there are a lot more back-door curveballs than front-door curveballs. Again, my suspicion is that pitchers don’t like to throw pitches there due to fear of making a mistake inside.

***

Hopefully, these data—which just scratch the surface of a better understanding of the “how” and “why” of pitch location—are informative and interesting. As I said in my last post, there are a whole lot of things that must be done with these numbers in order to know what we really need to know, which means that I expect to be continuing with these posts for quite some time.


Comments are closed.