Zack Greinke Is Pulling a Felix
Pitch names aren’t very imaginative. What is a fastball but a ball that’s thrown fast? The ball, in fact, thrown the fastest, relative to the curveball — which curves — or to the knuckleball — thrown as if off the knuckles. The changeup is also entirely explained by the name, although this one requires you know something about another pitch. The changeup is supposed to change things up, when a hitter is looking for a fastball. It’s supposed to represent a change of speed. Absent a fastball, a changeup is nothing but a slower fastball. The changeup needs to change something up to survive.
The game has seen a lot of pitchers. For the overwhelming majority of them, we’re sitting on pretty limited information. Surely, there have been some outliers over the years, pitchers who have done unusual things with their pitches. According to conventional wisdom, a good changeup needs to be about 8 mph to 10 mph slower than the fastball. Anything less than that, it’s thought, and there’s not enough of a change of speed. The best-known exception to this idea is Felix Hernandez, who’s been known to throw changeups in the low-90s. Felix’s changeup is one of the best in the game, so he’s served as evidence that there’s more than one way to throw a hitter off with a change. Henderson Alvarez specifically cited Felix as the reason he’s willing to throw his own changeup harder. It can be rewarding to push the limits.
Zack Greinke, too, is pushing the limits. Obviously, he’s pushing the limits of un-scored-upon-ness. But he’s also become a lot more like Felix than you might have realized.
Greinke is one of those constant tinkerers. That’s neither a good thing nor a bad thing. That’s a necessary thing in order to remain successful. The league is always adjusting, and good players always need to be able to adjust right back. Every player is a work in progress, but some of Greinke’s adjustments have been more widely publicized. He’s referred to himself as maybe the most changed player in baseball in the past 10 years. Regardless of whether that’s true, Greinke will openly admit he never feels settled.
He’s always working on something. He’s forever been working on his changeup. It didn’t have the most promising origins. Pulling from an Eno article, a year and a half ago:
Take the change-up; he’s thrown it his whole career, but it hasn’t been the same pitch. “It was an awful pitch, they just wanted me to work on it,” the pitcher said of the Royals’ coaching staff. “It’s been bad for a long time.” But the hard work is paying off (“Every year it gets a little better”) […] — “It’s probably my best pitch to lefties now,” Greinke agrees.
To get a sense of the progress Greinke has made with his change, look at his usage of the pitch, which reflects his level of trust in it. From Brooks Baseball, going back to 2008, when Greinke became a starter for good:
The trend, generally, is upward. Greinke now is throwing more changeups to lefties; roughly one for every three pitches. He’s also throwing more changeups to righties; roughly one per 10 pitches. Greinke used to be thought of as fastball-slider pitcher. Now he’s more fastball-changeup guy. The slider’s still good, though, and the above plot leaves something critical out. It’s not just that Greinke’s throwing more changeups. It’s that he’s throwing more different changeups.
Covering the same years as above, here are Greinke’s fastball-changeup velocity differentials. Put more simply, here is the difference between Greinke’s average fastball speed and his average changeup speed. The trend is unmistakable.
At the beginning, Greinke featured an exaggerated velocity gap. He subsequently pulled that closer to average, but then he just never stopped shrinking the difference. From greater than 10 mph, to roughly seven, to less than five, to barely three. The gap has gotten smaller every year. This year, Greinke’s changeup is faster; his fastball is slower.
We have pitch data on this site going back to 2002, so I made some use of it. I looked at individual starting-pitcher seasons, going back to 2002, and set an innings minimum of 50. Within the pool, here are the five smallest fastball-changeup speed separations:
- 2013 Felix Hernandez, 3 mph
- 2014 Felix Hernandez, 3.1 mph
- 2012 Felix Hernandez, 3.4 mph
- 2015 Zack Greinke, 3.4 mph
- 2015 Felix Hernandez, 3.6 mph
Felix, Felix, Felix, Greinke, Felix. Greinke is tied with the third Felix. This is most certainly unusual. For Greinke, this also is most certainly effective. It’s hard to evaluate a pitch without considering its role as part of the bigger picture, but from Brooks, here’s a plot of the slugging percentages Greinke’s allowed on changeups:
Usage up, speed differential down, slugging way down. More directly, you could refer to the pitch values we have on our leaderboards. There, you’ll find that Greinke, in 2015, has thrown the most valuable changeup in baseball. Chris Sale is right behind him, then Felix. More indirectly, Greinke’s changeup is mostly a weapon against left-handed hitters, and after years of allowing a roughly 30% hard-hit rate to lefties, he’s trimmed that to about 21% the last season and a half. It’s tremendously difficult to know how good an individual pitch is. A pitch only works in the context of all the pitches. But all the Greinke changeup indicators are good, and if you’d prefer another approach, you could realize that Greinke is having a historic season. He’s also throwing more changeups than ever before. The two probably aren’t unrelated.
You want some visuals? Of course you do. A 1-and-1 Greinke changeup to a lefty:
A 1-and-2 Greinke changeup to a righty:
Greinke throws his changeup with Felix’s speed differential. He also throws his changeup at roughly the same speed as Felix’s changeup, which means he throws his fastball at roughly the same speed as Felix’s fastball. The two aren’t identical pitchers, nor are they strikingly similar pitchers. Still, they have some strikingly similar traits, in the little picture and in the big one. Greinke used to have what he called an awful changeup. Now his changeup might be his best weapon, and he’s using it in a way no one would recommend. No one except one of the best pitchers in the world, and those who’ve seen him and realized the changeup can be remarkably versatile.
At some point, Zack Greinke will be finished developing. By then, I imagine, he’ll be retired.
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.
Isn’t a knuckleball thrown off the fingernails? Of course knuckleball sounds a heck of a lot better than “fingernailball.”
I always thought the knuckleball was called that because of its knuckling flight path not because of the grip.
A knuckleball is said to have a knuckling flight only because it takes the description of it’s flight from it’s delivery, not because there is such a thing as knuckling flight. There isn’t. Before Abner Doubleday’s illegitimate son invented the knuckleball the rare ball seen to travel without spin was called teh gyroscopically challenged ball. In my yut I threw a gyroscopically challenged true forkball which nobody claimed to have forking flight. I forking dare any knucklehead to challenge the truth of this true statement.
Da Fuq did I just read
The pitch gets its name because when it’s thrown, the batter’s/catcher’s point of view sees the pitcher’s knuckles above the baseball when it’s thrown.
I did not know that. Of course, when I hit, I don’t see much of anything, so I can’t relate to that perspective. This is why I read and comment on baseball articles rather than playing professionally.
When I pitch slow-pitch softball, though, I throw a knuckleball with my actual knuckles. The ball is simply too big for the fingernail grip. It’s obviously not moving fast enough to dance like a knuckler, but it’s worth throwing just for the moment’s pause it generates in the hitter as it approaches. 🙂
“…thrown AS IF if off the knuckles.”
The movement he gets on his changeup, even at the elevated velocity, man, no wonder it’s so tough to hit. I wonder if there’s a tendency for hitters to expect the norm, to expect that 10 mph velocity difference, and that’s what makes Grienke and Felix so successful with such hard changeups. They’re still a little out in front if they think it’s a fastball, but maybe they’re behind if they’re expecting a change. I have a hard time judging timing without contact as a reference.
Maybe Greinke’s doing better because his teammates started washing their hands after they poop. Ever think of that?
He’s throwing his change-up faster to get that dirty baseball out of his hand quicker.
Why is it that a fast changeup is not considered a sinker / 2-seamer? Is it the movement? How do we draw the line in between one and the other?
Yeah, I was wondering this, too. Especially considering that PitchFX is looking at pitch velocity, H_movement, V_movement, and maybe spin, I don’t understand how it’s able to make that distinction, especially when the differential in velocity is so low.
Put another way, if Greinke threw a 2 seamer with a fair amount of sink and arm-side movement, wouldn’t PitchFX think that it was a changeup?
This might help a little, but I’m not sure how PITCHf/x separates the overlaps.
http://www.fangraphs.com/library/pitch-type-abbreviations-classifications/
Yeah, checking out the numbers, the answer is – perhaps unsatisfyingly – it just is. If you go the leaderboards, and look at pitchers who throw both pitches with some kind of regularity, there is almost always a substantial difference (~3 MPH) between a 2-seamer/sinker/splitter and a change up. There are some pitchers who throw both or even three of them at similar speeds, but when you look at their individual player pages you see that the most likely explanation is that PitchFX makes a mistake; the frequency of one or two of those pitches is less than 1%.
So in other words, PitchFX classifies the pitch one way and says ‘Greinke throws a change up’, and just continues to see his pitches that way. Maybe a human programs a pitcher’s repertoire into the machine and Greinke’s pitches just get classified according to that, I’m not sure.
Where this can get you into trouble is by comparing changeups (or sinkers or two-seamers or splitters) belonging to different pitchers without comparing changeups to other pitchers’ changeups AND to other pitchers’ sinkers AND to other pitchers’ two-seamers AND to other pitchers’ splitters. When Jeff writes that Greinke’s changeup is the most valuable in baseball in 2015, followed by Sale and then Felix he isn’t wrong, but it’s putting a lot of faith in PitchFX classifying all those pitches for all pitchers correctly. What we really care to know is ‘who throws a pitch – with arm-side movement that is hard, but not as hard as a fastball, and that dips a little bit – the best in baseball?’. When we sort for PitchFX value for 2015 for all four pitches (two seam, changeup, splitter, and sinker) here are your leaders:
Keuchel 16.4 (two-seam)
Kazmir 14.7 (two-seam)
Greinke 13.9 (Changeup) <–
Burnett 13.2 (sinker)
Gibson 11.6 (two-seam)
Lackey 10.9 (two-seam)
Sale 10.3 (two-seam)
Sale 10.1 (Changeup)
Heston 9.9 (sinker)
Hernandez 9.4 (Changeup)
Volquez 9.3 (Changeup)
Hahn 9.2 (two-seam)
…
Eovaldi 7.1 (splitter) (included as he's the leader for this pitch)
This is more what I was trying to get into.. Thanks for the answer and the research!
I’d assume it’s the grip. If he’s still using a change-up grip, we’ll call it change regardless of velocity.
But PitchFX does not account for grip, right?
Jeff is using Brooks Baseball for his data, who categorize each pitch according to the pitcher that threw it. Without knowing what Felix throws, for example, some of his changeups might end up classified as sinkers or two seamers or something else entirely, but folks at Brooks Baseball know that Felix calls that pitch his changeup, so they classify it as such. As far as I know, PitchF/X doesn’t have an algorithm tailored to specific pitchers, so you’ll see different usage patterns between Brooks Baseball and PitchF/X.
How does Greinke get the sink on his changeup. Is it gravity or is he pulling the string and putting top spin on it?
Maybe he pronates his wrist so much over the top of the ball that it’s almost like throwing a hard screwball, like a reverse slider or something, and that does give him a little bit of a topspin component.
Not having your fingers under the ball – so gripping it more from the sides or deeper in your hand – reduces backspin, and thus reduces rise (or increases drop, whatever way you want to look at it). This is why a splitter drops; less backspin.
filthy
I could watch that changeup at the RH batter’s ankles all day and be perfectly content.
So, essentially, Greinke’s last several starts answer the question “What happens if you give an already elite pitcher Felix’s changeup”?
Jesus.
Pitches are traditionally defined by grip, and there are generally accepted ideas of what kind of pitch is created by that grip.
Ultimately, a pitch should be defined by what happens after it leaves the hand of the pitcher. Furthermore, as addressed above, a pitch is defined by the pitches that are thrown before and after.
Greinke throwing his change at a high velocity creates sharper, later break.
Pitchers that are constantly evolving are truly artists.
Another four years, and Greinke’s change will be faster than his fastball. That’ll really throw hitters.
Well, they do call it a “change up.” No reason that the velocity must be down.
Dickey throws a fastball as a changeup from his knuckleball. I’ll be super impressed if Grienke starts throwing a knuckleball regularly.
We just need someone to throw Dickey’s hard knuckleball and Jamie Moyer’s fastball, which is a few MPH slower.
“Change of pace”, for those of us old enough to recall the original term for the pitch.
So why isn’t it Changeop then?