Where Balls in Play are Allowed, and What it Doesn’t Mean

As much as we’ve all grown accustomed to citing FIP, and trusting FIP, the theory is always kind of staggering at first, no matter who you are or what your background. The principle is that pitchers have little control over the results of their balls in play allowed. The evidence is convincing and exhausting. And it’s so challenging to come to terms with, because it flies in the face of what people are taught playing baseball growing up, and because different pitchers have different abilities to put the pitches where they want. How could location possibly be that unimportant, at least in that one particular regard? Don’t some spots lead to worse contact than others? Can’t some guys throw more pitches to those very spots?

Some people are still working on the investigation, and of course we know that BABIP isn’t completely random. For example, there’s the meaningful difference between groundball pitchers and fly-ball pitchers. But the general conclusion’s still valid. It never stops being a little weird when you stop and think about it, and what’s presented below contributes to the weirdness.

We have lots of information on which guys induce swings at pitches out of the zone, and which guys don’t. It’s good to get swings at pitches out of the zone, because those swings are less likely to lead to contact. They’re also less likely to lead to damaging hits and home runs, because the strike zone covers most hitters’ sweet spots. As a pitcher, you want to be able to trick the hitter into chasing something in or out or high or low. Do that often enough and you’re cut out for the major leagues, probably.

We also have information on out-of-zone contact rates. Those numbers get cited somewhat frequently. What I don’t read much about, however, is the location of pitches hit and hit fair. It can be as simple as a percent of balls in play that were hit on pitches in the zone. I thought I’d dive in for a quick examination, just to see what might turn up. My first thought was that pitchers who allowed more fair balls on pitches in the zone would allow higher BABIPs, because those pitches are more hittable. Allow more fair balls on pitches out of the zone, and those probably get weaker contact, and those probably lead to fewer hits. It sounded good enough to me.

In generating data, I turned to Baseball Savant, and looked at all pitchers who threw at least 1,500 pitches last season. This yielded a sample of 151. I figured out how many fair balls each allowed on pitches out of the zone, and then on pitches in the zone, and then I generated this long and really easy table. I figured showing the whole list would be better than just showing ten players at either end. From Samuel Deduno to Jered Weaver, here are the 2013 results:

Place Player %IZ %OOZ
1 Samuel Deduno 81% 19%
2 Nathan Eovaldi 77% 23%
3 Chad Gaudin 76% 24%
4 Felix Doubront 75% 25%
5 Derek Holland 74% 26%
6 Cliff Lee 74% 26%
7 A.J. Burnett 74% 26%
8 Madison Bumgarner 74% 26%
9 Tyson Ross 74% 26%
10 Shelby Miller 73% 27%
11 Tony Cingrani 73% 27%
12 Jon Niese 73% 27%
13 Jose Fernandez 73% 27%
14 Matt Cain 73% 27%
15 Garrett Richards 72% 28%
16 Justin Grimm 72% 28%
17 Corey Kluber 72% 28%
18 Jordan Zimmermann 72% 28%
19 Tyler Chatwood 72% 28%
20 Stephen Strasburg 72% 28%
21 Phil Hughes 72% 28%
22 Ervin Santana 72% 28%
23 Ubaldo Jimenez 72% 28%
24 Tom Koehler 72% 28%
25 Bud Norris 71% 29%
26 Patrick Corbin 71% 29%
27 Jhoulys Chacin 71% 29%
28 Edinson Volquez 71% 29%
29 Justin Masterson 71% 29%
30 Scott Diamond 71% 29%
31 CC Sabathia 71% 29%
32 Trevor Cahill 71% 29%
33 Tim Hudson 71% 29%
34 Nick Tepesch 71% 29%
35 Clayton Kershaw 71% 29%
36 R.A. Dickey 71% 29%
37 Jorge De La Rosa 70% 30%
38 Matt Moore 70% 30%
39 C.J. Wilson 70% 30%
40 Bartolo Colon 70% 30%
41 Roberto Hernandez 70% 30%
42 Martin Perez 70% 30%
43 Charlie Morton 70% 30%
44 Mike Leake 70% 30%
45 Juan Nicasio 70% 30%
46 Paul Maholm 70% 30%
47 Josh Collmenter 70% 30%
48 Tim Lincecum 70% 30%
49 Yovani Gallardo 70% 30%
50 Jason Hammel 70% 30%
51 Jose Quintana 70% 30%
52 Erik Bedard 70% 30%
53 Felix Hernandez 70% 30%
54 Jacob Turner 69% 31%
55 Aaron Harang 69% 31%
56 Wade Miley 69% 31%
57 Matt Garza 69% 31%
58 Kris Medlen 69% 31%
59 Edwin Jackson 69% 31%
60 Matt Harvey 69% 31%
61 Anibal Sanchez 69% 31%
62 David Price 69% 31%
63 Alexi Ogando 69% 31%
64 Jeremy Guthrie 69% 31%
65 James Shields 69% 31%
66 Dan Haren 69% 31%
67 John Lackey 69% 31%
68 Dylan Axelrod 69% 31%
69 Yu Darvish 69% 31%
70 Marco Estrada 69% 31%
71 Scott Feldman 69% 31%
72 Jon Lester 69% 31%
73 Jeff Samardzija 69% 31%
74 Carlos Villanueva 69% 31%
75 Chris Sale 68% 32%
76 Jake Peavy 68% 32%
77 Brandon McCarthy 68% 32%
78 Jake Westbrook 68% 32%
79 Bronson Arroyo 68% 32%
80 Dan Straily 68% 32%
81 Julio Teheran 68% 32%
82 Travis Wood 68% 32%
83 Lance Lynn 68% 32%
84 Zack Wheeler 68% 32%
85 Mike Minor 68% 32%
86 Chris Capuano 68% 32%
87 Hisashi Iwakuma 68% 32%
88 Scott Kazmir 68% 32%
89 Kevin Correia 68% 32%
90 Andrew Cashner 68% 32%
91 Zach McAllister 68% 32%
92 Wade Davis 68% 33%
93 Dallas Keuchel 67% 33%
94 Hector Santiago 67% 33%
95 Mat Latos 67% 33%
96 Jonathan Pettibone 67% 33%
97 Michael Wacha 67% 33%
98 Jeremy Hefner 67% 33%
99 Andy Pettitte 67% 33%
100 Jarrod Parker 67% 33%
101 Ricky Nolasco 67% 33%
102 Wily Peralta 67% 33%
103 Jeff Locke 67% 33%
104 Adam Wainwright 66% 34%
105 Homer Bailey 66% 34%
106 Joe Saunders 66% 34%
107 Dillon Gee 66% 34%
108 Chris Tillman 66% 34%
109 Rick Porcello 66% 34%
110 Kyle Lohse 66% 34%
111 Gerrit Cole 66% 34%
112 Tommy Milone 66% 34%
113 John Danks 66% 34%
114 Chris Archer 66% 34%
115 J.A. Happ 66% 34%
116 Wei-Yin Chen 66% 34%
117 Francisco Liriano 66% 34%
118 Cole Hamels 66% 34%
119 Ian Kennedy 65% 35%
120 Luis Mendoza 65% 35%
121 A.J. Griffin 65% 35%
122 Joe Kelly 65% 35%
123 Bruce Chen 65% 35%
124 Esmil Rogers 65% 35%
125 Jerome Williams 65% 35%
126 Joe Blanton 65% 35%
127 Jason Marquis 65% 35%
128 Mike Pelfrey 65% 35%
129 Max Scherzer 64% 36%
130 Zack Greinke 64% 36%
131 Hyun-Jin Ryu 64% 36%
132 Gio Gonzalez 64% 36%
133 Alex Cobb 64% 36%
134 Barry Zito 64% 36%
135 Miguel Gonzalez 64% 36%
136 Clay Buchholz 63% 37%
137 Ryan Vogelsong 63% 37%
138 Kyle Kendrick 63% 37%
139 Jeremy Hellickson 63% 37%
140 Justin Verlander 62% 38%
141 Ryan Dempster 62% 38%
142 Jordan Lyles 62% 38%
143 Randall Delgado 62% 38%
144 Jason Vargas 62% 38%
145 Doug Fister 60% 40%
146 Hiroki Kuroda 60% 40%
147 Mark Buehrle 59% 41%
148 Ivan Nova 58% 42%
149 Lucas Harrell 57% 43%
150 Eric Stults 57% 43%
151 Jered Weaver 57% 43%

According to the numbers, Deduno allowed 344 fair hit balls, and 279 of those hit pitches within the strike zone. At the other end, Weaver allowed 471 fair hit balls, and 268 of those hit pitches within the strike zone. Intuitively, it’s not surprising to see a command guy like Weaver in that position, nor is it surprising to think that hitters might’ve forced a guy like Deduno to throw more over the plate. In terms of in-zone rate, Weaver finished 2.9 standard deviations below the mean. Deduno, meanwhile, finished 3.4 standard deviations above it. Here’s a comparative chart, showing the locations of pitches hit fair for both guys:

dedunoweaver

This, like usual, is from the catcher’s perspective,and for Weaver you see a lot of balls hit fair on pitches off the plate to the left. Deduno is mostly collected in the middle. This is just a graphical representation of the numbers that looked so different above, and if this is all you knew, you’d think, all right, Weaver was a lot better, then. He probably allowed fewer hits, and fewer damaging hits.

And it’s true that Weaver allowed a .268 BABIP, while Deduno came in at .291. But then, hitters slugged .382 against Weaver, and .375 against Deduno. The four guys right above Weaver at the bottom of the table allowed BABIPs over .300. The top ten pitchers in the table allowed an average .288 BABIP. The bottom ten pitchers in the table allowed an average .300 BABIP. There’s barely any correlation between IZ% and BABIP, and what correlation there is is slightly negative. In two ways, that’s the opposite of what one might’ve expected.

You’d think a pitcher would be better off getting more balls in play on pitches off the plate, relatively speaking. That doesn’t actually seem to hold true in the majors. Maybe it does hold true at other levels, but I suspect the majors are selective for pitchers who are successful however they are, and if you’re a guy who allows balls in play on pitches in the zone, then to cut it in the bigs you have to demonstrate some ability to keep those hit balls from hurting too bad. Not all pitches in the zone hit fair are created alike. Nor are all pitches out of the zone hit fair created alike. It turns out that this, like everything, is complicated, yielding results that demand further and closer attention. Follow-up studies. The whole enchilada.

It could be that guys toward the bottom throw more hittable pitches out of the zone than guys toward the top. The bottom ten averaged an O-Contact% of 68%. The top ten averaged an O-Contact% of 60%, and the correlation between O-Contact% and IZ% comes out to about 0.3. Then there are going to be other factors as well, because there always are, in greater number than one could ever imagine. I’m sure you’ve thought of a few. I’m still thinking of more, but I don’t want this post to meander.

Up there: results. They’re not what I expected, and for me, that makes them interesting. Four of five balls hit fair against Samuel Deduno were on pitches in the zone. Four of seven balls hit fair against Jered Weaver were on pitches in the zone. Batters posted a higher slugging percentage against Weaver than against Deduno. There’s no meaningful correlation between IZ% and BABIP. Nobody needed more evidence that there are several different ways to be successful in the majors, but you can throw this on the pile anyway. Don’t stand too close to the pile, for your own safety.




Print This Post



Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.


26 Responses to “Where Balls in Play are Allowed, and What it Doesn’t Mean”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Plato says:

    Only the best of the luckiest of hitters can consistently swing at pitches out of the zone and consistently get hits

    Vote -1 Vote +1

  2. pinch says:

    If I worked in the analytics department of a team, I would read all of Jeff Sullivan’s articles. I feel like you frequently come up with really interesting ideas that could reveal really interesting/counterintuitive ways to predict success and end them with “more research! That I sadly do not have the time/resources to do” and I would love to be able to leap in and do said research. Good article Jeff!

    Vote -1 Vote +1

  3. Oh, Beepy says:

    I feel that maybe the problem is that the measurement you’re taking allows for too many variables.

    Some of the guys near the top of the list are there because they are control artists who don’t throw outside the zone often enough for an appreciable percentage of their OOZ pitches to be hits, creating noise in the rankings. Perhaps to compensate for this there would need to be some sort of adjustment into a counting stat of sorts as opposed to a straight-rate stat, although I haven’t thought that out and it seems flawed.

    I feel that the fact that the best starting pitchers in baseball are scattered throughout is a function of what makes them the best in the first place. They usually throw four or more pitches, and they usually throw their fastballs for strikes.

    This is just spitballing but I think you’re onto something and you just still have too much noise.

    Vote -1 Vote +1

    • Oh, Beepy says:

      I forgot to add ‘they allow fewer hits and therefore a smaller sample’ before the four pitches sentence.

      Vote -1 Vote +1

    • tz says:

      In other words:

      Cliff Lee is at #6 because he pounds the strike zone, so by default the majority of his pitches hit or unhit are in the zone.

      A.J. Burnett is at #7 because his breaking stuff is too difficult to put in play when it breaks out of the strike zone.

      Vote -1 Vote +1

  4. Bob says:

    Samuel Deduno is an extreme ground ball pitcher and Jered Weaver is an extreme flyball pitcher. Deduno can pound the strike zone with sinkers and not get hit that hard while Weaver nibbles because a fastball down the middle is more likely to be hit in the air harder.

    Deduno’s BABIP will be higher and his SLG should be lower than Weaver due to the batted ball type (ground balls are more likely to be hits but give up fewer bases than fly balls).

    Also, the HR/FB % is typically higher for ground ball pitchers than it is for fly ball pitchers.

    Vote -1 Vote +1

    • GilaMonster says:

      There isn’t really a link, Nova and Fister are two of the more extreme GB% pitchers in baseball/

      Vote -1 Vote +1

    • Tim says:

      Deduno can pound the strike zone with sinkers

      In what universe? Dude has a career Zone% of 45.1% and a walk rate north of four and a half. I’m amazed he manages to top the list just because he hardly ever throws strikes.

      Vote -1 Vote +1

  5. Krell says:

    It would be interesting to see this data broken down a little further, distinguishing if pitches in certain out-of-zone locations have a higher BABIP allowed than others.

    My inclination would be to think that pitches that are low would lead to a significantly higher number of ground balls, and ground balls result in a higher BABIP than fly balls. And because of that, low pitches might be skewing the sample.

    I would expect that the results would more closely align with expectations if low pitches were excluded from the sample. Also, it would be interesting to see if there’s as strong or stronger of a correlation based on pitch height, regardless of horizontal positioning.

    Vote -1 Vote +1

  6. Matthew Cornwell says:

    Jeff Zimmerman had an article this year which showed BABIP on all balls in play down the middle from 2008-2012 had a BABIP of .310. Pitches on the edges of the plate produced a BABIP of .289 and contact off the plate produced a BABIP of .273.

    How does this jive with this article’s findings?

    Vote -1 Vote +1

  7. Jon L. says:

    It looks like the best control pitchers get high contact percentages in the zone, simply because so many of their pitches are in the zone. Pitchers with great stuff, on the other hand, can induce more contact outside the zone. Therefore, skill can predict higher or lower contact percentages, leading to a lot of noise in the data.

    Vote -1 Vote +1

  8. dcs says:

    Why use BABIP and then SLG for comparison? Why not use SLGBIP? And I think it’s better not to take out the homers, By focusing on BABIP, you are taking out the pitches that result in the hardest contact (HR) and pitches that result in the weakest contact (K). So the ‘weird’ results are, to a significant degree, just a byproduct of the choices of what pitches to exclude. plus luck, of course.

    Vote -1 Vote +1

  9. Im quite certain that balls in play out of the zone result in a lower BABIP than balls in play in the zone. this study does not really refute that..because the individual events are not taken into account, only the total average.

    perhaps weavers in the zone balls in play had a much higher safe rate than deduno’s…and therefor even though weaver induced more in play contact on balls out of the zone..his overall averages turned out to be higher….

    Vote -1 Vote +1

  10. Matt says:

    “There’s barely any correlation between IZ% and BABIP, and what correlation there is is slightly negative. In two ways, that’s the opposite of what one might’ve expected.”

    Instead of looking at the correlation of a pitcher’s IZ% with his overall BABIP, wouldn’t it make more sense to look at a pitcher’s BABIP on IZ pitches with the same pitcher’s BABIP on OOZ pitches? Might not be so easy to extract the data, but think it would be more informative.

    Just looking at the baseball savant page for Weaver, and I think I get: IZ: 77 hits, 14 of them HR. So that’s a BABIP of 63/254 = 248, AVG of 77/268 = 287, SLG of 140/248 = 565.

    OOZ: 61 hits, 3 HR. BABIP = 58/200 = 290, AVG = 61/203 = 300, SLG = 82/203 = 404.

    Would be interested to see how IZ SLG vs OOZ SLG correlated, and also the same for AVG/BABIP.

    Vote -1 Vote +1

  11. tz says:

    This is all very interesting stuff.

    However, I’m still floored by one basic fact:

    This is the first I’ve ever heard of Samuel Deduno. SMH

    Vote -1 Vote +1

  12. DavidKB says:

    The vagaries of BABIP are probably the most fascinating thing uncovered by the whole sabr movement.

    Regarding this article, a question: Why do you break things up by pitcher? Why not break it up by pitch type, velocity, and/or location? Should the BABIP of a 94 mph fastball middle away change because of who threw it? Of course the answer is partly yes.. maybe the better pitchers set up their pitches in sequence. Maybe you could look at first pitches only.

    Okay, I get it. There are a million studies you could do. Chop chop Sullivan! The people are waiting!

    Vote -1 Vote +1

  13. Spit Ball says:

    I’m not at all surprised to see the difference between Deduno and Weaver. Deduno is more over the top with his delivery then Weaver. Weaver’s fastball must be crazy to hit as a right handed batter or even a lefty. He is tall with long arms and his delivery is more three quarters. When he’s right his fastball moves away from a right handed player in an extreme manner.

    Vote -1 Vote +1

  14. Dan says:

    “How could location possibly be that unimportant, at least in that one particular regard?”

    I would imagine it has something to do with the composition of the bat and the ball and also the atmosphere, the air. Has anyone ever looked at differences between NCAA aluminum and wood BABIP?

    Vote -1 Vote +1

  15. Dan Rosenheck says:

    I think you’re looking at the wrong variable here. There may not be a relationship between the share of fair batted balls that come on strikes and BABIP; I haven’t studied that specifically. But there certainly is an extremely strong correlation between opposing batters’ contact rate when they swing at a ball in the strike zone and BABIP the following year. See http://www.fangraphs.com/blogs/hitting-em-where-they-are

    Vote -1 Vote +1

  16. randhyllcho says:

    Help me out with what we’re actually looking at here? Some pitchers who throw in the zone get hit more than those that don’t, but some who throw out of the zone get hit more than those that do? Is this a look at deception? Game theory? Ability to throw strikes?

    Vote -1 Vote +1

  17. Mike says:

    Jared Weaver, makes sense the guys nasty. im surprised to see Bheurle up there though.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>