Towards an Objective Measure of Hanging Pitches

While working on something Erasmo Ramirez said — that his slider was always in the zone anyway, so he should probably use it to steal strikes rather than for swinging strikes — it became obvious that breaking pitches are much less effective in the zone than out when it comes to swinging strikes. Curves, in particular, are much better outside the zone. You get about one third of the whiffs on a curve in the zone as you do outside of the zone.

Separately, I’m working on a piece for The Hardball Times Annual about command. In it, a few pitchers talk about the difficulty of commanding breaking pitches. “Nobody throws anything that’s truly straight,” is how Trevor Bauer put it.

While sorting the in and out of zone whiff rates, and thinking about command, it came to me that the two are related. Maybe that’s a duh, but a big part of quantifying command is the problem of breaking balls and changeups and their movement. A breaking ball in the zone may often be a hung breaking ball, which contributes to the lower whiff rates.

Let’s take a look at the pitchers that have the most disparate results on their non fastballs inside and outside the zone first, and then try to find a way to spot these pitchers by movement.

Below, I’ve taken all pitchers’ pitches thrown at least 200 times this year and found their in-zone and overall whiffs per swing rates. Then I’ve indexed the difference between those two numbers to pitch type — sliders aren’t as bad as curves in the zone, after all. Then I sorted the list for the biggest difference in baseball.

These pitches are much worse inside the zone then they are outside the zone.

The Pitches That Do the Worst in the Zone

Name	Pitch	All Whiff	Zone Whiff	Diff+	zWhiff+
A.J. Ramos	CH	57%	24%	297	134
Chasen Shreve	CH	44%	16%	252	87
Pedro Strop	SL	54%	14%	240	90
Miguel Gonzalez	CH	37%	12%	226	64
Keone Kela	CU	51%	13%	223	100
Joaquin Benoit	CH	43%	19%	212	107
Chad Bettis	CH	29%	6%	211	31
Luis Avilan	CH	40%	17%	204	93
Andrew Heaney	SL	42%	9%	201	56
Jeurys Familia	SL	50%	16%	197	107
Brett Cecil	CU	48%	15%	197	113
Trevor Bauer	CU	44%	12%	188	91
Francisco Rodriguez	CH	40%	19%	184	107
Taylor Jungmann	CU	46%	15%	183	113
Zack Greinke	CH	33%	13%	179	71

All Whiff = whiffs/swing
Zone Whiff = whiff/swing on pitches inside the strike zone
Diff+ = Difference between whiff% on all pitches and in-zone pitches, divided by league difference, *100
zWhiff+ = zone whiff rate divided by league whiff rate on that pitch type, * 100
Minimum 200 thrown

Of course, these aren’t all created the same. The change thrown by A.J. Ramos, and the curves thrown by Taylor Jungmann and Brett Cecil, these pitches are still better than average on getting whiffs on pitches inside the zone. They’re just much better outside the zone.

But take a look at Pedro Strop‘s slider, or Andrew Heaney’s slider. They go from excellent swing and miss pitches to below average inside the zone. Are those hanging breaking balls?

Results-based analysis isn’t very helpful for looking forward, so I thought I’d look at the difference in movement on non-fastballs inside and outside the zone.

Here are the pitchers who experience the biggest difference between the movement on their in-zone non-fastballs and their overall movement.

The Pitches That Move The Least Inside the Zone

Name	Pitch	count all	Zone PFx_z	All PFx_z	Diff Z
David Price	KC	256	0.2	-0.7	0.89
Felix Hernandez	SL	227	-2.9	-3.6	0.67
Wade Miley	SL	438	1.9	1.3	0.63
Will Smith	SL	389	0.5	-0.1	0.57
Neal Cotts	SL	499	4.0	3.5	0.55
Danny Salazar	CH	460	3.9	3.3	0.53
Michael Lorenzen	CU	207	-5.3	-5.8	0.50
Dan Haren	FS	211	6.6	6.1	0.49
James Shields	CH	639	5.3	4.8	0.49
Sergio Romo	SL	436	2.7	2.2	0.49
Jerome Williams	CH	316	3.9	3.5	0.43
Mike Pelfrey	FS	349	0.0	-0.4	0.42
Alex Wood	CH	434	1.5	1.1	0.41
Tommy Milone	CH	444	5.7	5.3	0.41
Noah Syndergaard	CU	490	-0.2	-0.6	0.40
C.J. Wilson	SL	271	0.2	-0.1	0.39

Zone PFx_z = vertical movement on pitches thrown into strike zone
All PFx_z = vertical movement on all pitches
Minimum 200 thrown

Our lists have none of the same names in common! That’s surprising.

Even more surprising is that these pitches, that have such different movement in and out of the strike zone, don’t have very different results in and out of the strike zone. Sure, Danny Salazar‘s changeup gets whiffs on 44% of swings outside the zone and only 25% inside the zone, and James Shields‘ changeup and Will Smith‘s slider have the same sort of splits. But David Price’s curve still gets 16% whiffs on swings inside the zone, and that’s decent for that pitch type. As a group, the guys above have neutral results in and out of the zone — a 5 percentage-point (higher) difference between whiff rates in the zone and overall than league average.

So we’re not quite there. And it’s probably because some guys vary the movement on their breaking balls on purpose — we’ve heard it from Adam Ottavino and Luke Gregerson here before, at least.

Still, there might be something to this idea that big changes in movement could give us clues about hanging breaking balls. Let’s try one more time.

Here are the league’s curve balls, ranked by the size of the standard deviation in their vertical movement. These guys are most likely to throw one curve which features considerable depth and then a second curve without considerably less downward break.

The Curveballs With the Biggest Range in Movement

Name	Count	Avg PFx_z	Standard Deviation PFx_z
Steven Wright	167	0.4	4.72
Jesse Chavez	170	-5.0	3.88
Aaron Sanchez	190	-7.6	3.71
Chris Heston	617	-3.0	3.49
Noah Syndergaard	490	-0.6	3.37
J.A. Happ	275	-4.7	3.36
Erasmo Ramirez	174	0.8	3.31
Roenis Elias	311	-4.1	3.20
Mike Bolsinger	674	-7.8	3.03
Yoervis Medina	129	-4.8	2.87
Michael Lorenzen	207	-5.8	2.81
Vincent Velasquez	165	-4.3	2.78
Tim Lincecum	206	-7.6	2.74
Shane Greene	147	0.0	2.69
John Danks	267	-0.9	2.68
Tanner Scheppers	197	-1.7	2.67
Felix Hernandez	553	-8.7	2.52
Danny Duffy	496	-4.0	2.48
Miguel Gonzalez	251	-5.2	2.48
Carlos Torres	138	-7.6	2.43

Avg PFx_z = average vertical movement
Standard Deviation PFx_z = standard deviation in vertical movement
Minimum 100 curveballs thrown. 2015

In other words, Jesse Chavez’s curveball has the second-widest variation in vertical movement among all major leaguers. And he gets 43% whiffs per swing on the pitch overall, but only 15% on pitches inside the zone.

Given that we know that more vertical movement is good for curveballs, and that curves have the biggest difference between in-zone and overall whiff rates, this analysis seems to fit this pitch-type best. There are some good curveballs on this list, but there are more than a few hangers.

Unfortunately, there’s no relationship between this standard deviation measure and any outcome measures, so we’ll have to keep searching. Perhaps the next step would be to control for count — the curve thrown for a stolen strike can look very different than the one thrown for a swinging strike. A cluster analysis that judged the “tightness” of the movement clusters, while correcting for count, might give us a better measure of command when it comes to pitches other than fastballs.