Cole Hamels' changeup.
A good changeup is a good thing for a pitcher to have, but you know what's even better than a good changeup?
An elite changeup. That's what. A pitcher can build a career on one of those, and there are a few who can tell you all about it.
But what exactly sets an elite changeup apart from, you know, a regular old changeup? What makes elite changeups so darn special?
If you watch enough baseball, you probably already have a good idea. But I like to have fun with numbers and pictures, and this is a dandy of an opportunity to have fun with numbers and pictures.
If we're going to talk about what makes an elite changeup elite, the first thing we need is a non-elite changeup. We need a sort of lab rat, one to pick apart and, to an extent, to pick on.
The numbers point toward New York Mets right-hander Jeremy Hefner. His changeup has just the right amount of "meh" for this task.
According to Brooks Baseball, opponents are hitting Hefner's changeup at a .333 clip with a .750 slugging percentage. Per FanGraphs, no other pitcher in baseball has surrendered more runs below average with his changeup than Hefner.
But Hefner's changeup was good. Hitters hit .242 against it with a .364 slugging percentage in 2012, and he saved 2.4 runs above average with it. It was actually his best pitch.
What's happened to it this year? From the looks of things, the scouting report has gotten around and hitters simply aren't being deceived by it anymore. It's not hard to pick apart why.
Deceptiveness: On Release Points
The deceptive act of a changeup starts before the ball even leaves a pitcher's hand. The idea is to throw a pitch that looks like it's going to be a fastball, and the deception act can easily fall apart if a pitcher's arm broadcasts that something not very fastball-y is coming.
That's been an issue for Hefner in 2013, as you can kinda-sorta tell by looking at his release points over at TexasLeaguers.com. It's more clear if we focus on a single start, however, and his start against the New York Yankees on May 29 is as good a case study as any.
The following GIF is going to show you three images. The first is the release points for the four-seam fastballs Hefner threw in that start. Then his changeup release points are going to be overlayed on top of that. The last image will highlight where the overlap between the two release points was strongest.
There was quite a difference between Hefner's release point for his four-seamer and his changeup. He was basically throwing his four-seamer from one slot, and his changeup from another. That's not recommended.
Now consider Johan Santana, who's forever going to be known for having one of the sickest changeups the world has ever known. Much of its effectiveness is owed to Santana's ability to throw it from roughly the same release point that he throws his fastball.
The following GIF shows the same thing the above GIF does, but with the release point plots from Santana's no-hitter last June:
Look at that. Then look at it again. Then swoon. What you have just witnessed is an illustration of a perfect changeup release point. What it looks like, in essence, is a perfect fastball release point.
Any variation between release points can be a problem. These are major league hitters we're talking about, after all. If they get the slightest hint that a changeup is coming, they will be capable of hitting said changeup.
In Hefner's case, one guy who could vouch is Brennan Boesch.
In an at-bat against Hefner in that May 29 game, Boesch saw a fastball from Hefner on the first pitch and a changeup from him on the third pitch, according to Brooks Baseball. I can't re-post it here, but there's a graph that shows the two pitches, naturally, came from notably different release points.
Boesch did this to the changeup:
If it wasn't the release point that tipped Boesch, there are a couple of other things that could have been responsible for that pitch's fate.
Command and Speed: On Location and Velocity Differentials
Changeups need to be well-located as much as any other pitch, and Hefner left the one that Boesch hit right out over the heart of the plate. That's where pitches go to die, and they have a much smaller chance of cheating death if they're not fast.
This changeup may have had a chance if Boesch was tricked enough by it to have his balance thrown off. But he wasn't. You can see how well he was staying back on the ball in this shot:
Part of the reason it was easy for Boesch to stay back on the ball was its velocity, which is another primary weakness of Hefner's changeup.
Brooks Baseball has the average velocity of Hefner's four-seamer at 91.48 miles per hour and the average velocity of his sinker at 90.85. His changeup comes in at an average of 85.00 miles per hour, meaning there's only about a five- or six-MPH difference between his changeup and his hard stuff.
This shortcoming was on display in the Boesch at-bat. Brooks Baseball says the first-pitch fastball Hefner threw him was about 92 miles per hour. The changeup he threw him was 86 miles per hour.
What a pitcher wants to have is something closer to a 10-MPH difference between his fastball and his changeup. An ideal changeup has more than that much separation.
This is another case where Santana would be a perfect guy to turn to, as Baseball Info Solutions (via FanGraphs) says he had a roughly 14-MPH difference between his fastball and changeup back in 2004 when he won his first Cy Young award.
But alas, there's no video footage from back then. So we're going to have to make do with Clay Buchholz instead.
Buchholz isn't relying on his changeup in 2013 as much as he used to, but it's still one of his best offerings. Per Brooks Baseball, he's holding hitters to a .222 average and a .333 slugging percentage with his changeup, and one of the best things about it is the velocity separation that he has.
An average Buchholz fastball comes across the plate between 92 and 93 miles per hour. His average changeup comes in at about 81 miles per hour, meaning there's usually an 11- or 12-MPH difference between it and his hard stuff.
Consider a matchup against Tampa Bay Rays catcher Jose Molina back in April. You can see that the release points of Buchholz's pitches in the at-bat were very similar, and what we're after are the final two pitches of the at-bat.
Strike two was a changeup on the outside part of the plate that Molina was badly fooled on. I don't have video, but here's a GIF of images pulled from the MLB.TV archives:
Here's a slightly better look at how much Molina was out on his front foot as this pitch was nearing the plate:
This changeup from Buchholz did everything it was supposed to do. It was located on the outside part of the plate, it was slow enough to get Molina off-balance and in the end all he could do was swing over the top of it.
You can see the pitch that finished the at-bat at around the 1:45 mark in this highlight reel:
That's a two-seamer at 92 in virtually the same place that Molina had just seen a changeup at 81. After having his bat slowed down on the previous pitch, he had no chance.
Now, there's one other thing Buchholz's changeup had that Hefner's didn't, and that's our cue to transition to the next point.
Even More Deceptiveness: On Movement
A good changeup must have movement, but the best changeups have all of the movement.
Go back and re-watch the video of the Boesch home run and keep an eye on the movement of Hefner's changeup. It didn't stay completely still the whole way to the plate, but it started to break pretty early and didn't break that much before Boesch sent it into orbit.
If one were to illustrate the movement of the pitch in an image, it would look like this:
Not exactly the easiest thing to show in a still image, to be sure, but the relative flatness of it shows through well enough.
Compare that with the movement of Buchholz's changeup:
Late break and sharp break. Just what you want in a changeup.
The movement on Buchholz's changeup is hardly legendary. It looks pretty good next to the movement of Hefner's changeup, but quite pedestrian next to a couple other changeups that I can think of.
The best right-handed changeup I've ever seen, for example, belonged to Pedro Martinez. He broke it out to strike out Barry Larkin in the 1999 All-Star Game, and its filthiness was on full display.
It's the first pitch you're going to see here:
Feel free to swoon once again. Or you can re-watch the video again and say, "Weeeeeee!"
Either would be appropriate, for the movement of that changeup went a little something like this:
That's a pitch that looked like it was going to be a foot outside, but then it darted back toward the plate and down at Larkin's knees. The word "unhittable" applies.
In modern times, of course, there's the nastiest changeup of them all. Otherwise known as the one that belongs to Cole Hamels.
This has been a rough season for Hamels, but there's nothing wrong with his changeup. Per Brooks Baseball, hitters are only hitting .139 against it with a .188 slugging percentage. According to FanGraphs, Hamels has saved more runs above average with his changeup than anyone else, just as he did in 2011 and in 2008.
Lest you've forgotten what Hamels' changeup looks like, check out the 0:30 mark here (you might be forced to go to MLB.com):
Granted, that was opposing pitcher Jacob Turner at the plate, but even the love child of Miguel Cabrera and Chris Davis would have a hard time hitting that pitch. The movement on it went a little something like this:
Great movement was one thing that changeup had going for it, and it also had terrific location down below Turner's knees. Just like we talked about.
We also talked about velocity differential, and this changeup had that too. Per Brooks Baseball, Hamels threw Turner six fastballs in that at-bat, topping out at 94.97 miles per hour. That changeup was clocked at 85.6 miles per hour.
Those numbers seem a bit generous to me, but that sort of velocity differential is fairly typical of Hamels. Brooks Baseball says he's generally throwing his fastball at 92 and his changeup at 84 in 2013, which is close enough to the 10-MPH difference pitchers want to have.
Another thing Hamels was doing in that start against the Marlins was masking the delivery of his changeup extremely well, as he usually does. With another assist from TexasLeaguers.com, here's some more animated trickery:
Hamels got himself a $144 million extension last year, and his changeup is a big reason why. It's an elite changeup that indeed has all the key trappings of an elite changeup. Hamels can place it, deceive hitters, slow it down and make it move.
A lot of pitchers throw changeups, but only a few can throw them as they're meant to be thrown. And when they bust them out, good things happen.
Thus concludes our chat on changeups. I'd like to thank Mr. Hefner for volunteering, and I'm hereby putting an offer on the table to buy him a cold one the next time he visits the Bay Area.
Or maybe two would be better.
If you want to talk baseball, hit me up on Twitter.