Another look at the sinker

In one of my former articles (it’s written in Chinese so I can’t share it with you until I rewrite it in English), I found out that groundball pitchers are generally more economical on their pitches: Namely, they throw fewer pitches per inning by 1) throwing fewer pitches per at-bat and 2) inducing more double plays. I speculated that groundball pitchers throw fewer pitches because their major weapon is a sinker (two-seamer, sinking fastball, or whatever you want to call it) instead of a four-seam fastball, and a sinker results in fewer swings-and-misses, leading to more balls in play than a four-seam fastball (given that they’re at the same velocity, of course). Back then, it was just speculation, but now it’s a different story. With the help of the new MLB Enhanced Gameday data this year, we can take a deeper look at this topic.

The ideal way to do this is to find someone who throws a lot of fastballs and sinkers, and pitches for one of the nine teams with the Enhanced Gameday system set up in their home stadiums. Off the top of my head, there are two pitchers who meet this criteria: Brad Penny and A.J. Burnett. After I looking through the data, unfortunately, I found it difficult to distinguish Penny’s fastball from his sinker, leaving me with only Burnett to work with. And let’s look at this graph first:

image

The plot above shows, for all 614 pitches thrown by Burnett that have been captured by Enhanced Gameday thus far (intentional balls and pitchouts excluded), the starting speed of the pitch on the horizontal axis and the horizontal movement on the vertical axis. You can see that there are three groups of pitches. The dots in the blue area on the top left are Burnett’s curveball, and those in the grey areanare his change up, and the ones in the red area on the right are his fastballs, both four-seam and two-seam. I suppose this is quite straight forward, so without furthur ado, let’s take a look at another plot:

image

This one shows the horizontal movement and vertical movement of the pitches on, respectively, horizontal and vertical axis and it only include pitches faster than 92.5 mph, which, as you can see from the first plot, should leave us with just the fastballs and sinkers. From this plot, we can see that there are two group of pitches, as shown by the blue and red colored areas. I subjectively picked a place to draw a line separating fastballs from sinkers (I can’t think of a better way to do this. I’ve tried to sort by break length and break angle, but it didn’t do any better). Now with the help of this arbitrary line (the red line shown on the plot), I’ve separated Burnett’s fastballs and sinkers and looked at their data. Here’s what I found:

Pitch          Number    Speed_0(mph)   Speed_1(mph)   Break(in)  Miss%   BIP%
Fastball        238           96.5           86.5       4.15      19.1     30
Sinker          155           96.1           86.2       6.41         8   50.7

Speed_0:starting speed
Speed_1:end speed
Break: break length
Miss%: times of swing and miss divided by the times when the batter takes a swing on the pitch
BIP%: times of ball put in play divided by the times when the batter takes a swing on the pitch

From this table, we can see that Burnett’s sinker has larger break than his fastball and is just a bit slower. As you might think, if one pitch is at the same speed as the other one but has more break, it’s harder for the hitters to make contact. However, that’s actually not the case here. The percentage of swings and misses induced by the sinker is, surprisingly, much lower than that of the fastball, and it’s also much easier for the hitters to put a sinker in play.

Of course, Burnett is only one pitcher. Therefore, I tried to approach the question in another way. I chose some sinkerballers and some non-sinkerballers who play in parks with the Enhanced Gameday system to see how hitters do against their fastballs/sinkers. And here are the results:

Sinker:
Pitcher        Number    Speed_0 (mph)  Speed_1 (mph)  Break(in)  Miss%     BIP%
M.Batista        303           93.3           84.1       6.04        8      49.3
A.Burnett        155           96.1           86.2       6.41        8      50.7
F.Hernandez      212           96.5           86.2       5.59     15.3      47.1
R.Halladay       305           91.6           82.9        8.1     10.4      61.5
D.Lowe           595           90.1           81.4       8.08     16.5        54
C.Wang           117           94.1             84       6.79      8.6      48.3

Fastball:
Pitcher        Number    Speed_0 (mph)  Speed_1 (mph)  Break(in)  Miss%     BIP%
A.Burnett        238           96.5           86.5       4.15     19.1      30.0
K.Davies         212           91.8           83.9       3.80     26.5      48.2
J.Lackey         333           91.6           83.2       4.20     15.2      43.0
Je.Weaver        349           89.8           81.1       3.37     13.9      43.1
C.Young          609           91.9           80.7       2.38     22.1      34.2

The sinkerballers, on average, throw harder than the non-sinkerballers, though their ability to induce swings and misses is lower, with half of them not even top the double-digits while the lowest number for non-sinkerballers is 13.9%. As for the percentage of balls put in play, the sinkerballers are also apparently higher than the non-sinkerballers. With the small sample size, we can’t draw any strong conclusions about it. Nonetheless, I’m willing to hypothesize that there really is something special (no idea on what it is, though) about the sinker that makes it easier for hitters to make contact. When MLB puts Enhanced Gameday System in more parks, we might be able to get a better handle on it.

Oh, by the way, there’s is one more thing worth noticing. Among the sinkerballers, the one with the highest swing and miss rate, Derek Lowe, is also the one who has the lowest velocity. I think that it is easier to make contact with a sinker, but that effect would be more pronounced with pitchers who throw sinkers with less total break, such as Wang or Jake Westbrook. I would expect guys like Lowe and Brandon Webb, who get more movement, to get more swings and misses. I don’t know if this is true or not; it’s just my gut feeling from watching the game. Maybe with more data, we can figure this out someday.


Comments are closed.