The inside change-up: courage or folly?

Last time I wrote about Keith Hernandez’s excellent book Pure Baseball, which
I truly enjoyed. I especially liked the stuff on pitching and the focus of that piece was Keith’s take on locating three kinds of pitches:
the fastball, the slider and the change-up. Let’s start with the
change-up—here is what Keith wrote about locating the pitch:

The change-up is never thrown purposefully inside. Never. If the
change does what it’s designed to do and gets the hitter off stride,
about all he can do with the pitch over the outside part of the plate
is to hit it weakly toward the end of the bat. But even if he is off
stride he can still get the head of the bat on the inside change-up
and pull it with power, sometimes with one arm. The pitcher who throws
an inside change-up runs a major risk that he will soon be, in the
immortal words of George Hendrick, “rubbing up a new one”.

This explanation didn’t exactly convince me, mostly because I didn’t
understand it. I believe I understand now, however, thanks to the
comments left on Ballhype
by sabermetrician Mitchel Lichtman (aka
MGL). Basically, Lichtman’s point is that the batter needs to swing early on an
inside pitch and the swing on a change-up tends to be early anyway,
hence the likelihood of “rubbing up a new one.”*

Ted Lilly delivers a change-up (note the grip). The pitch is likely headed, despite conventional wisdom,to the inner half of the strike zone. (Icon/SMI)

* That is a good phrase by Joggin’ George Hendrick, isn’t
it? Anybody out there remember Hendrick? Back in his St. Louis days,
he had a reputation as a lazy bum (hence the “Joggin'”), but I always
liked him. Sure, he did sit down on the outfield grass during
Cardinal pitching changes, but I can’t recall him dogging it going
down to first on a ground ball, stuff like that. He also would talk to
us bleacherites during those pitching changes, which we all thought
was very cool.

Ok, so inside change-ups are dangerous pitches and nobody ever throws
them purposefully. “Never,” as Keith said. “Never” is a big word, so I thought I might have
a look at the PITCHf/x data to see if I could learn something about inside change-ups.
Keith did say
“purposefully,” and even the PITCHf/x data, marvelous as it is, cannot
tell us where the pitcher was trying to throw the ball.* Still
we can look at change-up pitch location for some pitchers and see if we can learn something.

* One thing I’ve been thinking about is trying to get an idea of
how well pitchers can actually locate their pitches. That will vary
of course, depending on pitcher and type of pitch, but let’s say the
best control pitcher out there, throwing his best-controlled pitch
(almost for sure the fastball). Can he put it within six inches of
where he wants it? Eight inches? A foot? That’d be an interesting
number to know, I think.

Anyway, back to inside change-ups. To try to get some feeling
for the inside change, the first thing we can do is pick a few
pitchers with good or famous change-ups and see if they avoid the
inside part of the plate. Why don’t we start out
with Mark Buehrle and Jamie Moyer? Buehrle’s change-up was ranked
very high in a previous study I did, while Moyer’s change-up is very
famous*. Below you see location diagrams for change-ups by these two
lefties.

*If you are wondering who had the best change-up ever, well have I got a book for you. The incomparable Neyer/James Guide to Pitchers lists the 10 best practitioners of each kind of pitch. Here is their list of the top 10 change-up artists in baseball history: 1) Stu Miller, 2) Jean Dubuc, 3) Ed Lopat, 4) Jamie Moyer,
5) Pedro Martinez, 6) Trevor Hoffman, 7) Doug Jones, 8) Ellis Kinder, 9) Bill Sherdel, 10) Andy Messersmith. Go on, head over to Baseball Reference to look up Dubuc and Sherdel. I’ll wait …

Right, so here are the location diagrams for Buehrle and Moyer (change-ups only). By the way, I’m only looking at change-ups to opposite-hand hitters, which are
the vast majority of change-ups thrown. The charts:

Well, looks like Keith is on to something here. Only a very small
number of these pitches were thrown on the inside third of the
plate. If you look at a lot of location plots like these (and I
have looked at quite a few), you see many that look like these two; most pitchers really do keep the change-ups away.

Ok, but these are just two pitchers*; what about all the others? Does
anybody work the change-up on the inside part of the plate? Well, as
much as I like the plots with the colored dots and all, I’m not really
enthused about looking at the 130 location plots for pitchers who have
at least 100 PITCHf/x-captured change-ups thrown in 2007.

*Two left-handed pitchers. You’ve probably noticed that many of the great change-up artists
are left-handed: Santana, Hamels, Moyer, Glavine, etc. There’s a
good reason for that: change-ups are used primarily against
opposite-handed batters and lefties face far more opposite-handed
batters than righties. Southpaws need to develop a pitch that will keep
those platoon-advantaged hitters off-stride. There have been many
great change-ups thrown by right-handed pitchers of course—the names Pedro Martinez and Trevor Hoffman spring immediately to
mind.

So, let’s try a different tack. Let’s go through the PITCHf/x data
and for each pitcher show how often he throws the change-up inside
compared to how often he keeps it away (see the Resources section for
more details). Let’s call the ratio of inside change-ups to outside
change-ups the CourageQuotient (CQ for short), because you need
to be brave (or stupid) to intentionally come inside with the change.
Mark Buehrle’s CQ is 0.1, meaning for each inside
change-up, he threw 10 outside change-ups. Got that? OK.

The following list shows the 10 pitchers who most kept their
change-ups outside. Steve Trachsel only threw two inside change-ups to
compare with 68 thrown outside.

+-------------------+-------+-----+---------+--------+-------+
| pitcher           | speed |  NP | outside | inside | CQ    |
+-------------------+-------+-----+---------+--------+-------+
| Trachsel_Steve    |  80.8 | 106 |      68 |      2 | 0.029 | 
| Durbin_Chad       |  79.4 | 113 |      66 |      2 | 0.030 | 
| Moseley_Dustin    |    82 | 119 |      87 |      3 | 0.034 | 
| Zito_Barry        |  72.7 | 269 |     171 |      9 | 0.052 | 
| Mitre_Sergio      |  84.1 | 104 |      61 |      4 | 0.065 | 
| Vargas_Claudio    |  81.9 | 122 |      73 |      5 | 0.068 | 
| Davies_Kyle       |  81.4 | 183 |     111 |      8 | 0.072 | 
| Marcum_Shaun      |  82.1 | 278 |     182 |     15 | 0.082 | 
| Gabbard_Kason     |  80.1 | 186 |     105 |      9 | 0.085 | 
| Smoltz_John       |  85.5 | 245 |     148 |     13 | 0.087 | 
+-------------------+-------+-----+---------+--------+-------+

OK, so these are the extreme never-throw-the-change-inside guys. The
average CQ is around .25, or one change thrown inside for every four
thrown outside. So, even on average, pitchers are really keeping their
change-ups outside. Not always, but usually. Of course, we don’t know
if the relatively few inside change-ups are the result of pitchers
missing inside (this would be Hernandez’ view) or if they are trying
to throw it inside.

Now let’s look at the leaders in CQ.
These are the guys who come inside most often with the change-up:

+-------------------+-------+-----+---------+--------+-------+
| pitcher           | speed | nCU | outside | inside | CQ    |
+-------------------+-------+-----+---------+--------+-------+
| Lilly_Ted         |  79.2 | 169 |      42 |     66 | 1.571 | 
| De La Rosa_Jorge  |  81.8 | 124 |      49 |     39 | 0.795 | 
| Hamels_Cole       |  82.3 | 226 |      80 |     52 | 0.650 | 
| Hill_Rich         |  82.4 | 107 |      44 |     28 | 0.636 | 
| McGowan_Dustin    |  87.9 | 196 |      77 |     47 | 0.610 | 
| Matsuzaka_Daisuke |  82.3 | 115 |      49 |     29 | 0.591 | 
| Beckett_Josh      |    91 | 152 |      64 |     34 | 0.531 | 
| Reyes_Jo-Jo       |  84.1 | 147 |      64 |     34 | 0.531 | 
| Pettitte_Andy     |  79.7 | 105 |      52 |     26 | 0.500 | 
| Tomko_Brett       |  84.7 | 135 |      59 |     29 | 0.491 | 
+-------------------+-------+-----+---------+--------+-------+

Note that only one of the guys, Ted Lilly, throws the change-up more
often inside than outside. I’ll come back to Lilly in a minute, but
first a few comments are in order. First, note that only a handful of
pitchers throw even half as many inside change-ups than outside
change-ups. So even these fearless change-up artists are throwing the
large majority of their changes outside.

And what about control? Maybe the intention was to throw all these
change-ups outside and they just missed inside. That would be Keith’s
view. Let’s go back to the pitch location graphs and let’s compare
Cole Hamels and Jorge de la Rosa.

De la Rosa’s CQ is higher than Hamels’, but when you look at these
plots, you get the feeling that Hamels was locating his change
in different parts of the strike zone, but most of de la Rosa’s inside
change-ups look to be pitches that he released too late and ended up
low and inside. I can’t prove that Hamels’ inside pitches were
intentional, of course; it’s more of a gut feeling.

Ted Lilly had a CQ of 1.57, which is about twice as high as any other pitcher. I don’t know how many of these
inside change-ups are intentional, but it’s pretty clear that Ted Lilly is doing something that nobody
else in major league baseball is doing. The sheer number of inside change-ups would make you think that some decent
fraction of them must have been intentional. On the other hand, Lilly
does not rely heavily on the change-up (he only throws the it around
11 percent of the time), which makes me wonder why he’d be so “revolutionary”
with his least important pitch, if that makes any sense.

The question that we need to answer is this: What is happening to all those inside change-ups?
Does Lilly have to “rub up a new one” often after throwing one of these changes on the inner part of the plate?
Lilly’s location plot can be seen on the right, but this time I’ve added information on what
happened to each pitch.

Wow, now isn’t this interesting? Look at the change-ups on the inside
third of the plate: almost every single one that was even close to the
strike zone was a “success,” i.e., it resulted in either a strike or an
out. Actually, quite a few pitches that were off the plate inside
were taken for called strikes. See all those green dots on the
inside corner? Maybe these inside change-ups are so
rare, the batter is taken by surprise?

Only one of the hits that Lilly served up on the change-up was thrown
on the inner third, and as for the long ball, the single home run he surrendered was on the
outer third, high in the zone. No, the inside change-up did
not hurt Ted Lilly in 2007.

So, where do we stand with inside change-ups? Well, despite Keith
Hernandez’s assurance that the change-up is never thrown inside
purposefully, we are seeing quite a few inside change-ups. I suppose
all those inside pitches could have been aimed outside but just
missed their target, but that seems unlikely to me.

Furthermore, we have seen that Ted Lilly in 2007, who threw more of
his change-ups inside than any other pitcher, had a great degree of
success with the inside change-up. Lilly, of course, is just one
pitcher, and although a quick look at what happened on inside change-ups
from Hamels and Santana didn’t show that they were getting killed on
those pitches, a more comprehensive study would probably yield some
interesting results.

References & Resources

To calculate the Courage Quotient, I look at all pitchers with 100 recorded changeups against opposite-hand batters in the 2007 PITCHf/x data. CQ is just the ratio of inside pitches to outside pitches. Inside pitches are those that cross over the inner third of the plate or are inside but within about 7 inches of the strike zone horizontally. Outside pitches are defined analogously.

For me, Joe Posnanski* is the best baseball writer on the web right now.

* I stole this asterisk/digression thing from him, and though I cannot hope to approach even a fraction of his talent, I at least have the asterisks. Thanks, Joe.