Zac Gallen, Throwback

Nobody throws curveballs anymore. They’re old hat, as Michael Baumann just got done telling you. They don’t fit modern pitch design. Sliders do all the things that curveballs do, and mostly better. Look at the league changing right in front of our eyes:

There’s nothing particularly odd about this change. Sliders, of both the sweeping and tight variety, get better results. Even as they’ve exploded in popularity, even as more and more pitchers have added mediocre sweepers to broaden their arsenals, the numbers speak for themselves. Sliders have been more valuable than the average pitch this year. Curveballs have been less valuable than the average pitch. Sliders seem easier to teach, too, at least anecdotally. You don’t hear about a lot of pitching factories turning guys into Charlie Morton, but seemingly every child in America learns a sweeper these days.

I don’t think that trend is likely to change anytime soon. But today I’d like to show some appreciation for pitchers still going strong with their curveballs. More specifically, I’d like to talk about Zac Gallen, who keeps chugging along, throwing curves and succeeding where the rest of the league isn’t.

Gallen’s approach is delightfully direct. The things you learned about pitching as a kid? He still lives by them. He throws his fastball high in the zone with classic four-seam shape. He drops a curveball off of it, or a changeup if he’s facing a left-handed batter. It’s not rocket science. But one of the cool things about baseball today is that we can use rocket science concepts to bolster our intuition, so let’s do that for Gallen’s anachronistic pitches, at least a little.

Gallen isn’t just throwing a curve, he’s throwing a 12-6 curve, perhaps the most endangered subspecies of the pitches we call “curveball.” Vertical breaking balls have fallen out of favor in recent years. Breakers with horizontal movement are the new hot pitching trend, offering a breaking ball designed for pitching with a platoon advantage that pairs well with a changeup. 12-6 curves perform similarly against both lefties and righties, so they don’t quite fit the mold of “platoon pitch.”

Another issue: It’s hard to tell a credible story when you’re tunneling a good looping curveball off of a fastball. What do I mean by that? Think of it this way: Gallen’s curveball falls roughly 40 inches more than his fastball on its path home. The strike zone is around 20 inches tall. If he threw his curveball at the same initial trajectory as a top-of-the-zone fastball, it would bounce off of home plate. That means that a batter who recognizes curveball will have an easier time dealing with it – when the gaps in movement are that large, the hitter can simply pick up on spin and not swing. That approach doesn’t work as well when two pitches have a smaller gap in movement. For example, Gallen’s slider falls about 16 inches more than his fastball; hitters who see spin and decide not to swing might receive a called strike for their troubles.

Luckily, there’s a countervailing benefit to throwing curves that helps make up for their inherent weaknesses. At Saberseminar last weekend, White Sox director of pitching Brian Bannister found a descriptive way of explaining something I’d previously grasped only implicitly: Pitchers are trying to avoid the middle approach angles. You want your fastballs coming in at as shallow an angle as possible and your breaking balls diving. Hitters swing on a slight upward plane, and the harder a pitcher can make it to meet that plane, the more success they’re likely to have. Gallen’s fastball/curveball pairing does exactly that.

His four-seamer doesn’t exactly stick to this maxim; it has a roughly average approach angle. In fact, it’s a fairly average pitch overall, with velocity right around league average, standard four-seamer shape and movement, and so on. Our pitch grading models agree with me on that, but you don’t even need them to know this. If you’ve seen Gallen throw a few fastballs, you’ll understand it right away. Fancy math and the eye test concur – this is what a very normal fastball looks like.

Here’s where things get back to Bannister, though: Gallen takes advantage of that fastball shape by throwing a rainbow curve with a massively different shape. His curveball’s vertical approach angle is around -10 degrees, as compared to his fastball, which is around -4.8 degrees. That’s a huge gap; no one who meets the admittedly arbitrary cutoff of 500 curveballs and 500 fastballs thrown this year has a larger discrepancy.

That’s the math. The plain English of it is that batters shape their swings to hit his fastball, which means a relatively flat stroke. Then they encounter his curveball, which is breaking sharply downward as it reaches their bat. Flat swing, negative-trajectory pitch: The most likely outcome is that the bat hits the top part of the ball and sends it downward. As you might expect, Gallen’s curveball generates plenty of grounders – more than 50% in each of the past three years, as compared to a league average in the low 40s.

Of course, those grounders only happen when batters swing, and as we previously covered, big curveballs suffer most in how easy it is for hitters to identify them. Gallen pitches to minimize that problem, though. If he always threw his fastball to the precise top of the zone, a curveball with the same trajectory would never be a strike. But he doesn’t always throw it there. He misses high frequently – he’s effectively wild with it, in other words:

When Gallen is behind in the count, 6.1% of his fastballs cross the plate at 3.5 feet or higher. When he gets ahead, that number balloons to 34%. If we set a four foot cutoff to handle the really high ones, the number jumps from 3% to 19%. When Gallen’s ahead, he’s trying to miss high. The cloud of possible fastball trajectories expands by quite a bit, which means the cloud of pitches you could mistake for a fastball also expands. That changes the math on fastball/curveball pairings significantly.

Another point of uncertainty is that it’s not like all of Gallen’s curveballs drop by the same amount. He throws them anywhere between 77 and 85 mph, and with greatly varying break as a result. Sure, his average curve falls 50 inches on its way home, but he’ll cook up a 40 or 60 incher from time to time. Given that he’s disproportionately throwing his curve when ahead in the count, he doesn’t need to fool the batter every time; if he can just keep them off balance and force them into a bad decision once in a while, he’ll get the strikeout (or weak contact) and move on with life.

Most of Gallen’s game is built around this variance. He adjusts his fastball location more by count than most pitchers – he’s locating above the zone at half the league average rate when behind, and by comfortably more than average when ahead. He adds and subtracts from his curveball to make things even trickier. He mixes in cutters and sliders that split the difference – not particularly frequently, because they aren’t great pitches, but often enough that hitters can’t ignore them in two-strike counts.

It might sound like I’m saying that Gallen’s innovation is not controlling his pitches well. I’d argue that it’s the exact opposite. He’s harnessing the variance involved in pitching. The average case of his fastball/curveball pairing is nothing special, but the various permutations are a lot better than that because they overlap and fold together. The high fastballs and the strike-stealing curves blend together, and that makes batters worse at both. Think you’re laying off a goofily high fastball? That might be a curveball for a strike. Think you’re going to hammer a lazy hook at the bottom of the zone? You might have misidentified one of those high fastballs you’re working so hard not to swing at, and good luck correcting that error.

Thanks to Baseball Savant, we can see this effect in three dimensions. Here are all of Gallen’s pitches from his most recent start, from a view behind the righty batter’s box:

It’s a messy graphic, but that’s kind of the point. You can see from the strike zone that he lands plenty of his curves (purple) for strikes. You can see from the tracers out by the mound that many of his fastballs (red) mirror the curves thrown for a strike early in their path before ending up well above the zone. He misses low sometimes too, and those pitches pair off of each other; a fastball down the middle and a curve that bounces in the dirt mirror each other fairly well.

That’s how Gallen makes his old school arsenal work. His pitch mix isn’t what you’d call “optimal.” It doesn’t score particularly well in stuff models of any type. Nothing stands out, really, aside from his command. In fact, this whole exercise might seem silly to you. He’s Zac Gallen! Of course he’s good. Why do we need to explain exactly why?

But I happen to think that the way he blends everything together is fascinating. It’s not just about raw stuff, or hitting the corner every time, or even about a new pitch type. It’s about how the whole can be more than the sum of its parts. Every pitcher misses their targets. The great ones incorporate that into their attack. Gallen uses his tools well and understands how to harness his misses. This mastery of uncertainty keeps him chugging along, year after year, with a bunch of pitches that seem just okay and a bunch of results that win him Cy Young votes.