Finding Baseball’s Least-Effective Pitch

We have a pretty good idea of baseball’s best pitches. You’ve got the Aroldis Chapman fastball. You’ve got the Kenley Jansen cutter. The Adam Wainwright curveball. The Stephen Strasburg changeup. The Cole Hamels changeup. The Felix Hernandez changeup. The Corey Kluber whatever it is. The Clayton Kershaw curveball. The Kershaw slider. The Kershaw hypothetical splitter that, in my imagination, he doesn’t throw because he doesn’t need to because of his curveball and his slider. There’s no clear winner, but there are plenty of candidates, and all of them are amazing.

We don’t have as good an idea of baseball’s worst pitches. The truth is baseball’s worst pitches don’t get thrown often outside the bullpen. They’re projects in which pitchers don’t have confidence, so you don’t see them in games. But we can skip over to something related, something that might stand as a decent proxy: We have the data to identify baseball’s least-effective pitches. At least among pitches that are thrown more than once or twice a month. This is one of the uses of the FanGraphs pitch-value data, and if you set a 50-pitch minimum, the second-least effective pitch this year has been Wei-Chung Wang’s changeup. And the first-least effective pitch this year? That honor belongs to Drew Smyly.

Smyly, being a pitcher, throws a fastball. We’re not going to look at his fastball. He’s also got a curveball. But we’re not going to look at his curveball. Not infrequently, he’ll come in with a cutter. We’re not going to look at his cutter. Smyly has a changeup he doesn’t throw very often. We’re going to look at his changeup.

You should know all about run values, and if you understand run values, you understand run value per 100 pitches. Wang’s changeup, mentioned above, has been worth -8.4 runs per 100 pitches. That’s terrible. Smyly’s changeup has been worth -8.9 runs per 100 pitches. That’s even worse. That’s the worst mark in baseball in 2014, and while the numbers will change over the remaining few weeks, at this particular moment the numbers are set and the numbers are damning.

Most of Smyly’s changeups have gone for balls. That’s bad for the run value. Many others have gone for hits — some of those for extra bases. That’s worse for the run value. Smyly, like a lot of lefties, has a pretty pronounced platoon split, and it’s something he’ll need to improve on if he wants to become a more reliable starting pitcher down the road. To get better, Smyly doesn’t need to get a better changeup, but that would be the most direct path, which is why he’s still working on the pitch. Perhaps the Rays figure they can help him out, either with the change or by turning his change into a splitter. The good news for Smyly is his changeup isn’t a finished product. The bad news is right now all he has is a scattered assortment of parts without instructions.

Here, basically, is where Smyly’s changeups have gone. The here caveat is that I might’ve misclassified a few deliveries:

That might not scream “terrible pitch” right away, but you can see some signs. Lots of misses too low. More misses up and also within the strike zone. Location inconsistency. The map tells you Smyly’s changeup has been inconsistent, and the rest of the numbers tell you the balls have been watched and the strikes have been punished.

Here is Smyly missing a little with a changeup and giving up a home run:

Here is Smyly missing a lot with a changeup and giving up a home run:

Here is Smyly missing a lot with a changeup and giving up a home run:

Here is Smyly missing a little with a changeup and giving up a double:

Here is Smyly missing a lot with a changeup and throwing a ball:

Here is Smyly missing a lot with a changeup and throwing a ball:

Smyly’s no idiot, of course. He doesn’t throw the changeup very much, and he’s aware it’s developing. But it would’ve been easy this past spring to figure Smyly was in for a big improvement. From January:

[…]he says he likes where he’s at with his changeup, a pitch he never really threw much in college or the year he spent in the minors. Or last year, for that matter.

“I kind of banked it early in the season,” Smyly said. “Because when you’re coming in for three or four batters, I just wanted to go straight to my go-to pitches. I didn’t really want to mess around. But I’ve been working on my changeup a lot, just trying to get it back.”

From February:

Smyly was sharp early but got dinged up a bit in the second inning, during which he threw a good number of changeups, a pitch he is focusing on this spring as he returns to the rotation from the bullpen.

“The more spring goes, I hope to get a little better feel for it,” Smyly said. “Even pitchers with a really good changeup, I think they might say that this early in spring. It’s a feel pitch, so you’ve got to keep throwing it in game situations, and eventually it’ll just be there and hopefully you can rely on it.”

From March:

“It’s a huge pitch for me and righties,” Smyly said. “I’m feeling more confident with my changeup, and today it was my best off-speed pitch. And I never thought I’d say that. It’s a great pitch to throw when you are behind in the count.”

He worked with Tigers pitching coach Jeff Jones on the pitch, and also credits his success with it on repetition and gaining a feel.

“I’m really just trusting throwing it,” Smyly said. “I was timid to throw it.”

In spring training, Drew Smyly was really trusting throwing his changeup. In April, he threw just one every 40 pitches or so. That frequency has hardly budged over time. He has reason to be timid again, as his changeup has been woefully ineffective. That doesn’t mean the run value perfectly captures the quality of the pitch, but it’s probably a good indicator. Besides, even if you throw a decent changeup, if it gets hit hard, you’re going to remember that. Smyly’s tried to fold a changeup into his repertoire and it just hasn’t been a pitch worth respecting yet.

As long as we’re here, at least for curiosity, we can look at years other than 2014. We’ve got pitch-type information going back to 2002. Let’s set a new minimum of 100 pitches thrown. The second-least-effective pitch since 2002 has been Adam Wilk’s changeup. The least-effective pitch since 2002 has been Drew Smyly’s changeup.

Turns out the pitch also was terrible in 2012. And while it basically disappeared in 2013, the few that were thrown didn’t go great. Sometimes it can be a little difficult to tell Smyly’s pitches apart, and because of the need to classify all these numbers come with their own error bars, but it’s not like Smyly right now is trying to re-discover a changeup that worked for him. It hasn’t worked for him in the past. He’s trying to work on a new thing. His career changeup run value is -8.1 per 100 pitches. Wilk’s changeup shows up at -6.6. Smyly’s in last by a bunch.

To leave you with a little bit of hope, here’s a Smyly changeup from just the other week:

This demonstrates an important idea: Smyly isn’t incapable of throwing a good changeup. He’s just incapable of consistently throwing a good changeup. The difference between good and bad is just the frequency of good versus the frequency of bad, and no one in baseball has 100% of one and 0% of the other. Some percent of Clayton Kershaw’s pitches are good pitches. Some percent of Kevin Correia’s pitches are good pitches. Correia throws a lower rate of good pitches, but some of his pitches are terrific. Kershaw-level, even.

Smyly has thrown good changeups. What he wants is to be able to more reliably throw good changeups. That’s why he’ll continue to work on the pitch, and that’s why the pitch still has future upside. Sometimes the most meaningless word in sports is “consistency,” but that’s a lot of what good pitching is. If Smyly becomes consistent with his changeup delivery, he’ll have a good changeup and he could be a good pitcher. The changeup delivery just hasn’t been consistent yet. So the results have been consistently dreadful.