The Good and the Worse of Taijuan Walker’s Changeup

It’s obvious, now, that Taijuan Walker is going to enter the season as a member of the Mariners’ rotation. He survived an offseason of trade rumors and beat out the recently-demoted Roenis Elias, and this is the kind of thing that can happen when you’re a pitcher who allows but a single run in 25 spring innings. There was, though, a point at which it looked like Walker and Elias might share an intense competition. So Walker came out guns a’blazing, immediately throwing a fastball at 95 – 96. Some pitchers use spring training to build up arm strength. Walker began it strong.

Which makes him an interesting guy to analyze. And that, in turn, is facilitated by Walker throwing a lot of spring innings before PITCHf/x cameras. Over those innings, Walker’s allowed nine hits and four walks, to go with 24 strikeouts. It seems like it’s been something of an early breakthrough, with Walker refining his mechanics and adding some depth to a slider. It’s been clear from the day he was drafted that Walker would have a big-league-caliber fastball. Of greater importance has been a changeup, an offspeed weapon for Walker to use against lefties. It just so happens Walker has thrown plenty of changeups this month. It’s been a clear priority, for obvious reasons. What, then, is there to be learned from the evidence?

As is generally the case, I’m going to use some data from both Brooks Baseball and Baseball Savant. And while Walker has made six spring appearances that show up on his MLB.com page, I’m less interested in the first ones. The last three times out, Walker has thrown a combined 17 innings, with 235 pitches. This can work as an approximation of Walker at or around 100%, mostly or fully stretched out. All three games took place in Peoria, meaning all three games were captured by PITCHf/x. That is neat! Onward.

Comparing Walker to himself in 2014, the recent version has thrown a tiny bit harder. There’s slightly different movement on the curveball, and the cutter has indeed become more of a slider, although it’s still registering above 90 miles per hour. The pitch of greatest interest, though, is the change, which Walker throws with something of a splitter grip. Brooks says he’s thrown 54 changeups over the last three games, and those changeups are up a tick and a half from where they were before. The movement is similar, though recently there’s been a bit more tail. Here’s an example of the changeup in action, as brought to you by a pretty lousy Arizona camera angle:

You might notice a couple things. One, that’s a changeup to a righty. That’s not an accident — Walker is trying to learn something or two from Felix Hernandez. And then there’s that “90 MPH” that shows up in the score bug. Specifically because of Felix, we know it’s not unheard of to have a changeup at 90, but for Walker this might be the new norm. From Brooks, over the last three games, Walker’s changeup has had an average velocity of 90.1mph.

There’s not a lot of comparison we can do in spring training, since so little is captured by PITCHf/x, but let’s take Walker’s number and compare it to 2014. Last year, among starting pitchers, the fastest average splitter belonged to Yu Darvish, at 88.9mph. And the fastest average changeup belonged to Felix, at 89.9mph. Four-hundredth of a point behind him: Henderson Alvarez. At this point, then, we’re splitting hairs, and there’s also error introduced by comparing 2014 to March 2015, but if you just take these numbers at face value, the changeup Walker has thrown the last few weeks would’ve been the fastest offspeed pitch for any starting pitcher a year ago. The very fastest. The same arm that gives Walker such an explosive fastball has given him an offspeed pitch that seems like it ought to come with a considerable margin of error.

This is when we get to something a little less promising. It’s one thing for a pitch to have an encouraging quality; it’s quite another for the pitch to be consistent in its delivery. Though it’s not at all the same thing, when I play catch, probably one out of five times I can throw a relatively good curveball. The other four times, I throw a disaster of a curveball, which means the actual curveball is very bad. Walker’s changeup can only be good if he can execute, awesome trait be damned, and this is where changeups have gone lately:

Sure, you see plenty in the right general area. You also see plenty left up, or thrown too far to the opposite side of the plate, and while not every changeup is thrown with the same intent, what can be gleaned from this is that Walker’s command of the pitch is very much still a work in progress. The explanation is always “not enough mechanical repetition”, and now consider this: a changeup is supposed to be thrown with the same arm action as the fastball. Here’s an example of Walker recently releasing a changeup, and Walker recently releasing a fastball:

It doesn’t matter which is which. Those were consecutive pitches, and there’s a subtle difference. The difference is so subtle you might not think it means anything, but this is how mechanical differences manifest. The throwing motions weren’t the same. As a proxy of mechanical consistency, we can compare average fastball release point to average changeup or splitter release point. In theory, we should want them to be similar. Using data from Brooks, here’s a table of 2015 Walker and some 2014 starters with what we think to be good changeups or splitters. You’re seeing the difference, in inches, between the average release points.

Pitcher	Release-point separation (in.)
Stephen Strasburg	0.5
Felix Hernandez	0.5
Hiroki Kuroda	1.2
Alex Cobb	1.5
James Shields	1.7
Hisashi Iwakuma	1.8
Cole Hamels	2.0
Masahiro Tanaka	2.3
Chris Sale	3.0
Taijuan Walker	4.0

Strasburg and Felix were basically right on. Their release points were virtually identical. Walker shows up last, here, and while he’s last by only one inch, that’s also 33% worse than the next-worst number. Now, there’s an obvious thing about this: it’s comparing Walker to guys with really excellent changeups or splitters. Walker’s pitch doesn’t have to be that good. But this is mostly for purposes of illumination. Walker throws a changeup that’s incredibly hard. However, he doesn’t quite mirror his fastball throwing motion, and it stands to reason that will limit the pitch’s effectiveness until or unless it’s ironed out. Because of the velocity it can still be a weapon, but the pitch isn’t yet close to its ceiling.

With Taijuan Walker, there’s so much to like. He’s still just 22 years old — younger than Jose Fernandez and Aaron Sanchez — and he’s among baseball’s most athletic pitchers, with upper-tier velocity. That upper-tier velocity applies to both his fastball and his changeup, and it’s the changeup that will probably most determine how Walker fares against lefties, and thus how he fares as a long-term starter. The changeup Walker’s been throwing the last few weeks would’ve out-paced any other starter’s changeup or splitter a season ago. The upside there is apparent. But there’s still more refining left to do. Which is always the case with incomplete pitchers. Walker came into camp with his arm strength already almost maxed out. The delivery? That’s still coming along.