The Seam-Shifted Revolution Is Headed for the Mainstream

Hey there! I want to give you a heads up about this article, because it doesn’t fit into a normal genre I write. Today, I won’t be telling you some new insight about a player you like, or creating some new nonsense statistic that tries to pull meaning from noise. This is a story about how baseball analysis is changing right before our eyes. A group of scientists and baseball thinkers are redefining the way we think about pitch movement, and I think it’s worth highlighting even if I don’t have anything to add to the conversation yet, because this new avenue of research is going to be front and center in Statcast-based analysis over the next few years.

“Seam-shifted wake,” as Andrew Smith, a student of Dr. Barton Smith (no relation) coined it, is a source of pitch movement that the first attempts at understanding the physics of a pitched baseball overlooked. It has already changed the way that coaches and pitchers approach pitch design, and due to recent data advances, it’s about to be everywhere. So let’s go over how we got here, to this newly observable way that pitchers deceive hitters, by starting at the beginning and working forward.

At its core, baseball is a game about one person trying to throw a ball past another person. There are other trappings — bases and baserunners, umpires, a strike zone, the mythology of Babe Ruth, and a million other sundry things. At the end of the day, though, everything starts with the pitcher trying to throw a ball past the batter.

Accordingly, baseball analysis over the years has focused on describing the flight of that ball. For a time, that simply meant describing the shape of pitches — they don’t call them curveballs for nothing. The next step was velocity — radar guns let us appreciate fastballs numerically rather than merely aesthetically.

In the past 15 years, the amount and scope of pitch-level analytical data has exploded. First, PITCHf/x quantified pitch location and movement. When we report a pitcher’s chase rate or how often a batter swings at pitches in the strike zone, it’s because the location where each pitch crosses the plate is recorded and logged. When we say a pitcher has eight inches of horizontal break on their slider, it’s because new technology allows us to measure it.

When Statcast debuted in 2015, it added another wrinkle: radar tracked the spin rate of each pitch in flight, putting a numerical value on something that had previously been only qualitative; a pitcher’s ability to generate movement through spin. Doctor Alan Nathan has written several authoritative studies discussing the value of this spin data.

One of Dr. Nathan’s key insights — one shared by many R&D departments across baseball but most eloquently explained here — was that raw spin rate doesn’t all lead directly to movement. If you’ve heard this part before, please feel free to skip ahead — you can Control+F to the words “frustratingly complex.” If you haven’t, though, let’s detour in this article full of detours to a quick discussion of spin.

Picture a car driving straight down a road. The tires on the car spin as the car moves, and if the car isn’t turning, the axis they spin around is pointed directly sideways (think of the car’s axels), perpendicular to the car’s forward movement. That’s called transverse spin. A baseball thrown with this spin — a perfectly backspinning fastball, say — creates movement due to the Magnus effect.

When people talk about fastballs rising, what they mean is that their transverse spin creates lift via the Magnus effect, making them fall less than a ball thrown with no spin would. When they talk about curveballs plummeting, it’s the same force applied in a different direction. Transverse spin — tire spin — creates movement via the Magnus effect.

Next, picture a football thrown with a perfect spiral. This spin, gyroscopic spin, doesn’t create movement via the Magnus effect. If you threw a baseball this way, the ball would still spin, but not in a way that creates Magnus-based movement; the relative angle between axis and flight path determines whether Magnus forces are created. In practice, most pitches are thrown with a mixture of gyroscopic and transverse spin. The exact composition of spin depends on grip, release, arm angle, and myriad other things the pitcher controls.

Here was a new insight: how much total spin a pitcher can generate is useful, but it’s largely useful for talking about what that pitcher can do, not what they’re currently doing. The higher the percentage of spin that is transverse (or “active”), the more Magnus-based movement a pitcher can generate.

I’ve attempted to be careful in mentioning that the movement we’re talking about here is specifically break created by the Magnus effect. You can model a ball’s total break as that produced by this interaction, but that’s just a model. Reality remains frustratingly complex, however, and as it turns out, baseballs move in ways that don’t fit into a tidy equation that can be solved using only spin, initial direction, and gravity.

A group of researchers led by Dr. Smith have spent years exploring, quantifying, and explaining a separate force, which Smith calls “seam-shifted wake.” In a series of posts, Smith was able to produce variable movement despite exactly identical spin orientation and magnitude by changing the location of the baseball’s seams. From there, he delved into the intricacies that led to this heretofore undiscovered source of movement. Researchers at Driveline Baseball appear to have been doing the same work in parallel, and published their own research on it late last year.

This leads us to a major breakthrough in the way that we can talk about spin. Before 2020, league-wide data on spin axis was inferred based on movement. In other words, analysts, myself included (though I’m more of an overzealous layperson and less of a scientist), measured the movement of the ball relative to its initial path. From there, they calculated the angle that would produce Magnus-based movement agreeing with the way the ball actually moved.

On an individual basis, pitchers already knew this wasn’t a perfect description of reality. By fixing high-speed cameras on their own deliveries, they could directly observe spin axis as the ball left their hand, and it didn’t always match up with what the movement-based angle calculations inferred. At a league level, however, those cameras didn’t exist. If you wanted a spreadsheet with every pitch thrown in 2019, your spin data was inferred, not observed.

In 2020, Statcast began collecting data with Hawkeye cameras. Those cameras could simply observe the axis of the ball at release, rather than inferring it based on movement. The two numbers don’t always agree. Why not? Seam-shifted wake! Smith, Nathan, and Harry Pavlidis published a pitch-level investigation that tracked the disagreement between observed angle (what the camera sees) and inferred angle (the movement-based calculation).

MLB and Baseball Savant now release graphics that show this divergence, as explained by this excellent article. Between this new data and Smith’s ongoing investigations, there’s plenty to unpack when thinking about pitch movement, but one thing is for sure: working out movement based on Magnus acceleration isn’t a complete picture of what’s happening between the mound and the plate.

Analysis of this data is still in its infancy, but the ramifications are potentially huge. Take just one example: spin mirroring, the idea that batters have trouble distinguishing between spin that moves in exactly opposite directions, relies on measuring the angle of spin. If you use pre-2020 data, however, you’re using the inferred angle, which is in some cases notably different from the way the ball actually spins.

The purported reason for this mirroring is that spin in opposite directions is visually indistinguishable. The problem is that the inferred spin axis data we were using to look for mirrored pitches didn’t always correctly match reality. If a ball leaves a pitcher’s hand with a 12:00 spin orientation only for seam-shifted wake to create horizontal movement, it might end up with an inferred spin orientation of, say, 10:00. Visually, however, the 12:00 spin is what a batter would see.

Well, maybe it’s what the batter would see. There’s even more confusion, though, because hands, baseballs, and all of reality exist in three dimensions, but we currently only record two. A pitch with 12:00 spin orientation but plenty of gyroscopic spin won’t resemble a pitch with 12:00 orientation and no gyroscopic spin from the batter’s perspective, depending on seam orientation. Heck, seam orientation changes what spin looks like as well. There’s far more to spin mirroring than what we can currently divine from publicly available data.

Over the next few years, I expect to see a raft of new research into seam-shifted wake, seam orientation, and how the two interact. There’s the pure physics research, working out how exactly turbulence created by seam movement can impact the ball’s flight. There’s data on how the wake works in practice; do pitches with more deviation between implied and inferred spin axes perform better? Do some pitchers succeed largely due to this effect? Eno Sarris is already looking into some of these questions, just to give one example.

I hope to do some of this research myself, and I’m sure that others at FanGraphs do as well. I’ve been spending a fair amount of my free time playing around with this data, and there’s still far more to learn. If there’s one lesson I’d take away from all of this though, it’s that we shouldn’t expect this next revolution in public-side analysis to be the last word.

Every revolution in data collection around pitching leads to more questions. Anyone who thinks their knowledge is final, that their plan for measuring pitch movement is perfect, eventually falls victim to one of the key limitations of every model: models only approximate reality, they don’t replace it.

If your model of spin axis involves only the Magnus effect, seam-shifted wake won’t exist for you. If your model of a hitter’s skill involves only his swinging strike rate and production on contact, you’ll miss any adjustments he makes in two-strike counts, because count-aware hitting doesn’t exist for you. These aren’t the same things — one’s a physical model and one’s an abstraction — but they both have one identical feature: they can’t account for what they don’t include.

With that caveat aside, the new era of movement analysis is going to be exciting. We’ve long wondered how some sinker- and changeup-heavy pitchers keep batters off balance despite middling velocity and movement. Smith’s research suggests that those two pitches, particularly when thrown with gyroscopic spin, can create meaningful non-Magnus movement. Why can’t batters square up Kyle Hendricks? The ball moves strangely!

In a year, we might think very differently about all kinds of different pitches, in the same way that measuring spin gave us new information about four-seam fastballs. We’ll certainly think differently about changeups, because we’re woefully short of ways to analyze their strange movement now. No matter how you slice it, we’ll be looking at the same kinds of pitches and thinking new things about them. It’s an exciting time to be learning new things about baseball, and we have physicists, cameras, and baseball laboratories to thank for it.

Update: this article has been updated to reflect the fact that Andrew Smith, not Dr. Barton Smith, coined the term “seam-shifted wake.”