Mike Pelfrey’s Sinkerby Jonathan Hale
June 10, 2009
Of types of pitchers I find a certain fascination in the sinkerballer, who can somehow throw a large part of pitching strategy and deception out the window and just throw the pitch one after the other. Not an overpowering pitch like a fastball, devastating like a big curve or slider, or deceptive like a changeup, the sinker is just an annoying pitch that leads to grounders and the ball staying in the park. With former sinkerballer kings Brandon Webb in the NL only just now coming back from shoulder surgery and Chien-Ming Wang's collapse the subject of another article, Mike Pelfrey has been the new poster boy for the sinker, putting together a nice little season (or at least he was when I started writing this, having allowed three runs or fewer in three starts before being shelled by Pittsburgh his last time out) despite having struck out just 26 batters in 59 1/3 innings. Let's put him under the microscope and see how he uses what is essentially his only pitch and what makes it effective:
First, here's a summary of all the pitches Pelfrey has thrown over his career, although they have come and gone over that time. Coming in to 2008, Pelfrey ditched his curveball in favor of the slider that has become his only real secondary pitch. His curve has made the occasional reappearance, but he is not throwing it at the moment:
Pitch % Thrown Movement (Horiz/Vert) MPH Fastball 80.1 (-7.84, 7.31) 92.6 Slider 13.2 (0.62, 3.72) 84.7 Curve 1.4 (3.21, -3.00) 75.2 Changeup 4.7 (-8.82,4.87) 83.8
The movement values are calibrated in inches, with (0,0) as a (mythical) pitch not affected by spinn—the easiest way to think about it is that (-5, 10) is about average for a four-seam fastball coming from a right handed pitcher. For a look at the average values for each pitch, check out John Walsh's article summary of different pitch types.
When he has been able to control them the movement on Pelfrey's secondary pitches has been reasonable, but let's focus on his sinker. With an average sink of just three inches and tail of just over two compared to a normal two-seam fastball, it hardly sounds impressive. However, I think one of the keys to the pitch is how much difference there is between one sinker and another. Here's the same data in graph form:
Note how large the group of sinkers is, spreading from normal four-seam fastball movement to about a foot away, much more than you would see for any other pitch. But it doesn't look like Pelfrey is throwing two different kinds of fastballs, because there aren't multiple groups of pitch movement, just a gradual and even variation in the movement on them up to a foot both horizontally and vertically. It's kind of hard to see with all the little dots, so here's a heat map to show how far his fastballs range in terms of movement:
And while there is some relation between the speed that he throws his slider at and the movement on it, there's still a huge spread at every speed. Pelfrey's sinker has ranged from below 88 mph to above 95 (although on a year to year basis is has been consistent), and there is some negative correlation to how fast he throws his sinker and the amount it drops (i.e. the harder the throws it, the less chance it gets to drop on the way to the plate).
However, while one of Pelfrey's mid 90's sinkers is going to have more drop in the long run, the amount on any individual pitch could range between 10-12 inches. That kind of unpredictable and large variation has to be difficult to deal with, and maybe makes up for not changing speeds or pitches as often as other pitchers—Pelfrey's sinker acts like several pitches in one.
Moving on from what it does, here's where Pelfrey's sinker goes. Because it tails, or moves in the opposite direction as a breaking ball, Pelfrey uses his sinker differently to right and left-handed batters. Against Lefties, he throws it to the very outside corner away from them, even usually off the plate. This graph is looking from the catcher's perspective and shows Pelfrey's sinkers to LHB (the yellow box is the league's average strike zone):
Against right-handed batters, he spreads his sinker around much more and is willing to throw them to both sides of the plate. Note the dead zone on the outside corner to righties though—his sinker has so much movement he very rarely is able to back door them by hitting that corner.
You would expect sinkers to more effective and usually thrown lower in the zone, but the above graphs show how many of Pelfrey's end up at the belt or above. In a previous article I showed that sinkers just lose their home run preventing abilities when thrown at the top of the zone, they aren't actually any more likely to be hit hard.
To see how this plays out for all results, we can use linear weights to see how effective his sinker has been at different heights, with more negative values indicating a better pitch. Dividing the vertical location of all the sinkers that Pelfrey has thrown into half-foot increments we can see that there was no real difference in performance against them within the strike zone, which ranges from an average of about 1.5 to 3.5 feet above the plate—and actually his low sinkers were slightly less effective as they resulted in more balls:
I'm sure that there are plenty of different sinkers out there, and Pelfrey is only known for his because he doesn't have anything else. But because of that we can see how one-trick ponies like him can get away with what they do—not necessarily with overpowering drop to their fastballs, but with a lot of variety both in terms of velocity and movement—and not as many worries about leaving a pitch up in the zone as a normal pitcher.
Jonathan Hale be found mixing cold hard statistics with reactionary conjecture at The Mockingbird . He welcomes questions and criticism via e-mail.