In the months since our introductory piece on the Seam-Shifted Wake effect, many different wings of the industry have taken a stab at proving its (SSW) existence and relevance from player to player across varying pitch types (Baseball Prospectus, Alan Nathan, Glenn Healey and Lequan Wang, Trip Somers). With strong evidence that SSW not only exists, but can also strongly influence performance, many solutions have been proposed to try and quantify its value.
While some analyses (including our own) have used Spin Direction (or Axis) Deviation to tackle performance-related Seam-Shifted Wake questions, this method does not explain the full relevance of SSW and other non-Magnus effects. For example, consider that:
- Although the league theoretically benefits from seam-effects on sinkers, not every player benefits from such an effect.
- Spin Direction measures the ratio of Horizontal Break and Vertical Break, rather than the direction and magnitude of movement itself. Thus, Spin Direction’s relation to the quality of a pitch is not direct.
- While one player may benefit from Spin Direction Deviation in one direction, another may benefit from deviation in the opposite direction.
Fortunately, in early 2021, Baseball Savant revealed their new Spin Direction Leaderboard, which allowed us access to directly measured Spin Efficiency (or Active Spin), Spin Rate, and Observed Spin Direction on a per player-season basis.
With the above parameters, one can reverse engineer how much a pitch should move based on spin and trajectory out of hand (similar to that of Rapsodo) and juxtapose it to how much it actually moved. By comparing the two, we can determine on a player pitch type basis who benefits from SSW (and other non-Magnus effects) using our proprietary pitcher “stuff” metric, Stuff+. Hopefully, this will be a step forward in defining the value behind non-Magnus effects, and determining which pitchers and pitch types stand to gain the most from Seam-Shifted Wakes.
NON-MAGNUS CONSIDERATIONS BEYOND SSW
Although Seam-Shifted Wakes are often credited as the sole source of non-Magnus effects on the baseball, there are other important factors that can influence pitch flight. Solely using deviation between Observed and Inferred Spin Direction can be misleading in its raw form, as there are often other wrinkles at play that can cause a discrepancy between the two.
Sometimes, a ball’s flight can be obstructed by forces out of a pitcher’s control, namely air conditions at ball release. Air density, wind, and other environmental factors at pitch release can cause irregular movement. While there are methods to make adjustments for these factors, it is a tall task to control for all of these at the exact moment a pitch is released, as we know that weather is oftentimes unpredictable. Coors Field in particular can suppress a pitch’s movement by ~25%. If we don’t attempt to control for environment, we can’t begin to attribute movement to other sorts of external forces, like SSW.
GYRO SPIN TO TRANSVERSE SPIN
This may be an effect that many are unfamiliar with, and it is certainly a bit more perplexing if you don’t come from a traditional physics background. Many are taught that Gyro Spin doesn’t contribute to the movement of a pitch, but that isn’t entirely correct. Gyro Spin (or Rifle Spin) can indeed contribute to movement depending on the pitch’s trajectory, the amount of Gyro Spin it has, and which direction the Gyro Degree is pointed to.*
*In this piece, we classified non-Magnus as forces unrelated to the expected Magnus force due to spin out of hand. While Gyro to Transverse falls into this category, technically, it is a Magnus effect as the axis is shifting during ball flight.
To gain a better perspective on the magnitude of this effect on the baseball, we compared Gyro Spin out of hand to the difference between Trajectory Horizontal Break and Spin-Based Horizontal Break values from in-facility Rapsodo sessions. In this case, a change of -1 inch would be 1 inch of additional sweep that occurs during ball flight.
We can see that at most, a RHP can look to gain ~2 inches of gloveside movement in the extreme case of 2500 RPM of Gyro Spin. While an additional 2 inches is nothing to sniff at, this effect generally provides somewhat trivial gains, with the median SL/CT in our dataset adding .75 inches of gloveside movement.
We owe an awful lot to Barton Smith and his continued research on Seam-Shifted Wakes, as his work has spawned many insightful and impactful pieces of public research. The SSW Effect can be summarized by the following: when altering the seam orientation of the ball, you can break the symmetry of the ball’s wake, causing turbulence on one side of the baseball that creates force, and thus movement, that wouldn’t be expected when looking solely at Magnus.
MEASURING NON-MAGNUS EFFECTS
Up until 2020, inferring non-Magnus effects would’ve been an impossible task at the Major League Level, as there were no in-game pitch-tracking technologies that directly measured spin characteristics at release, which are needed to surmise expected break. With the introduction of Hawk-Eye last season, an optical (meaning camera-based) tracking system that does measure these values, the wheels were set into motion.
With this newfound, directly-measured data in hand, we were able to measure expected movement, or Spin-Based Breaks (both horizontal and vertical), out of hand, by deriving average velocity and spin vectors for a player pitch type. From here, we removed drag and the effect of environmental conditions for both break sets so that we could isolate dragless movement from ball release.
When comparing this data with Trajectory Breaks, which describe the total movement of the pitch, we can attribute any difference between the two to either non-Magnus effects or Hawk-Eye error.
VALIDITY OF HAWK-EYE SPIN DIRECTION
When looking at different systems, there are a multitude of concerns: where the sensor (in this case a camera) is positioned within a park, the calibration of the system, weather, misreads due to a model error, as well as potential tagging issues. These issues are commonplace, with both PITCHf/x and Trackman both struggling during initial rollout.
Public baseball analyst Lau Sze Yui (@903124S on Twitter) outlined some concerns regarding potential errors with Hawk-Eye in a thread (our very own Alex Caravan also dove in), detailing the differences in Observed and Inferred Spin Directions for different parks and pitch types. In this case, extreme outliers are somewhat worrisome, as they suggest broad differences in Observed and Inferred SDs, which signal a larger non-Magnus effect. If these differences are inflated, that throws a wrench into measuring the magnitude and direction of non-Magnus.
In our own attempts to correct Hawk-Eye’s Spin Directions using the popular Standard Candle Method, we didn’t find much to be concerned about. Especially considering the magnitude of Spin Direction values (0-360°), small discrepancies do little to make us question the validity of Hawk-Eye’s output.
At most, we found the largest amount of park error to be less than 2°. With 21 out of 30 parks coming in below 1° of error, this does not appear to be a league-wide issue.
NON-MAGNUS AT THE MAJOR-LEAGUE LEVEL
With a better understanding of the factors above, and our Spin-Based and Trajectory Breaks in hand, we can start to get an idea of how prevalent non-Magnus effects like SSW are at the major-league level, as well as how they impact the quality of a pitcher’s arsenal.
LEAGUE-WIDE EFFECTS & MAGNITUDES
On a macro-scale, each pitch type is impacted by non-Magnus effects in different ways, with Gyro Spin-heavy pitches like sliders and cutters likelier to be influenced by Gyro to Transverse, and pitches that move to the armside (sinkers and changeups) being more likely to present seam-effects. However, non-Magnus effects are not mutually exclusive, meaning that many can be present at once.
Given the league-wide trends above, some expected outcomes appear. Sinkers, which have been at the center of SSW discussion, are heavily impacted both vertically and horizontally. The average sinker gained more than 3 inches of run and nearly 4 inches of depth due to non-Magnus effects — that is a significant amount of movement. Changeups (and splitters) don’t see quite the same gain with regards to run, but still benefit from a considerable amount of additional drop (which is generally a positive trait for offspeed pitches).
On the flip side, cutters also showed substantial changes, generating around 3 extra inches of gloveside movement and 2 additional inches of drop. Cutters typically have the heaviest gyro components, so they have the potential to be influenced by the Gyro to Transverse effect. Fourseamers, especially those with lower Spin Efficiencies (and thus more Gyro Spin), also generate a fair amount of “cut” but don’t seem to be as prone to losing ride as other pitch types.
Somewhat surprisingly, sliders and curveballs don’t seem to have any noticeable, large, league-wide effects. While at the extremes, there are certainly pitchers that have significant amounts of movement added/subtracted to their breaking balls due to non-Magnus effects, these pitches appear to be less susceptible to non-Magnus effects on average.
Perhaps these effects are clearer when examining a specific pitcher’s arsenal. Below, we can see that recent Astros’ signee Jake Odorizzi has an arsenal that is heavily impacted by non-Magnus effects, with the circle points representing Spin Breaks (expected) and the diamonds representing Trajectory Breaks (actual movement).
Odorizzi has a deep and varied arsenal, and each pitch was affected by SSW (and company) in a different way. For example, his fourseamer lost around 2 inches of rise, which significantly lowered the quality of the offering according to our Stuff+ metric (more on this below), while his changeup (which is more accurately classified as a splitter) benefitted from considerably more drop than what was expected.
In contrast, Max Scherzer’s arsenal appears to be predominantly Magnus-heavy, with most of his pitches impacted only marginally by seam-effects.
Scherzer’s highest non-Magnus gain in Stuff+ comes from his changeup, with the pitch dropping an additional 4 inches relative to expectation. Scherzer also sees a non-Magnus bump in Stuff+ on his cutter, which follows a similar trend to the league, adding a fair amount of gloveside movement.
IMPACT ON STUFF
As noted when mentioning the pitfalls of using Spin Direction as a proxy for movement, an inch of Horizontal Break is not always equivalent in value to an inch of Vertical Break, and vice versa. The quality of a pitch is multi-faceted, and the presence of non-Magnus break doesn’t always impact a pitcher’s “stuff” as we think it would. Internally, we use Stuff+ to assess pitch quality, taking into account the physical characteristics of a pitch, controlling for count, batter talent, location, and other external factors. Stuff+, like other “+” metrics, has a league average of 100. So, for a FB with a Stuff+ of 105, we can say that it is 5% better than league average.
Typically, Stuff+ is based upon Trajectory Breaks, as those are the break values that we have access to through MLB’s public data streams. These also represent the actual movement of the pitch. However, when comparing these traditional Stuff+ values to Stuff+ based on Spin Breaks (how we expect the pitch to move), we are able to determine how much value a player gains from harnessing (inadvertently or not) forces other than the Magnus effect.
For fastballs, the effects of SSW can cause dramatic shifts in pitch quality. At the top end, sinkers that generate an extreme amount of, well, sink due to shifted-wakes find themselves in a much better position than they would be if seam-effects weren’t present. It’s not an exaggeration to say that sinkers must have the presence of seam-effects, or else they will likely fall well below the threshold of an effective major-league offering (barring extremes in regards to velocity or command of the pitch).
On the opposite end, we see some aces that get docked pretty heavily, and there’s logic behind this. Out of hand, the sinkers of Darvish, Woodruff, and Burnes have a movement profile more similar to their fourseamer, featuring more ride and less run. During flight, due to seam-effects, these pitches are all brought into a movement profile that we refer to as the “dead zone”, where Vertical and Horizontal Break are almost equal and outcomes are more favorable to a batter.
Because Stuff+ is very high on heaters that have excess Vertical Break, but lower on dead zone sinkers, this suggests that SSW actually pushes these pitches into a profile that isn’t as effective.
As noted before, cutters usually have heavy gyro components, so they have more potential to be influenced by the Gyro to Transverse effect alongside any potential seam-effects. Cutters also have Observed Spin Directions very similar to a fastball (for example, Burnes’ cutter is at 192° & his fourseamer is at 190°), so when they end up generating more gloveside movement than expected, it leads to big gains in Stuff+.
Sliders are an interesting case, as we are aware that most don’t have much to add from non-Magnus forces. But, for some, the pitch appears to be more “slurvy” out of hand, only to end up with a monster, sweep-heavy profile (Yohan Ramirez, Jake Diekman, Mike Clevinger, Justin Topa).
But, because it’s likely that these pitches already rate out well without any unexpected movement, in most cases there aren’t massive gains to be had. There are some sliders, though, that seemingly “back-up” on the way to the plate, and lose HBreak in the process (Seth Romero, Jack Flaherty), which results in a pitch with a worse profile.
Offspeed pitches are often the trickiest to properly grade out, as much of their success is determined by how well they play with a pitcher’s mix of fastballs (Harry Pavlidis covered this idea in a series at BP). Additionally, we group splitters and changeups together, and these pitches have varying components that could be influenced by outside forces in a variety of ways.
Traditional changeups, like sinkers, are more susceptible to seam-effects due to seam orientation at release. However, changeup and splitter grips are a dime a dozen, and with so many varying grips, it’s hard to pinpoint which lead to the most extreme movement changes due to non-Magnus.
Generally, though, additional drop for a changeup is a valuable trait, as it helps increase separation between the offering and a pitcher’s FB. Those who are dinged the most when looking at Stuff+ tend to add lift to their offspeed, limiting the effectiveness of the pitch.
EXTENDING ANALYSIS BEYOND THE BIG LEAGUES
If you’re curious about Spin and Trajectory Breaks and have a Rapsodo unit with CSV access, you’re in luck. Given that Rapsodo is optical-based, like Hawk-Eye, it has the ability to pick up key parameters out of hand that help calculate both Spin-Based and Trajectory-Based Breaks. When you compare the two metrics against one another, you can start to measure the magnitude of non-Magnus effects both vertically and laterally.
If subscribed to Rapsodo’s Advanced or Advisor Cloud tiers, you have access to CSV files that include both break sets. However, you’ll need to be mindful of measurement error, placement of the unit, calibration, and other potential issues before diving in. As such, it’s important to view this solution as just a rough estimate of non-Magnus movement, rather than ground truth.
While it is clear that Seam-Shifted Wakes are not the only non-Magnus force to impact ball flight, they appear to be accountable for the lion’s share of non-Magnus movement. At the extremes, SSW can contribute ~9 inches of lateral and/or vertical movement to a pitch, changing the profile and subsequent outcomes achieved by a pitch type altogether.
Most pitch types at the MLB level get a sizable boost in quality due to these effects, but like most hard and fast rules, there are some exceptions. For example, around 42% of pitches in our 2020 MLB sample had a lower Stuff+ relative to their Spin-Based estimate, indicating that the value in obtaining a SSW effect should be considered in relation to a pitch type’s initial spin-profile rather than without context.
Not mentioned in this piece is the component to SSW that is perhaps the most interesting — late break. Unfortunately, because a pitch’s trajectory is still reported via Statcast using the 9 parameter fit, we still have work left to do to uncover when non-Magnus forces take place during ball flight.
Theoretically, being able to parse out “late break” would be a boon for understanding pitch quality, but until that data is available to the public at the pitch-level, we will have to continue to engineer workarounds at the per-season level like the ones provided in this piece.
Given that we’re still scratching the surface with SSW, we’ve provided a free spreadsheet (normalized so that all values are from the RHP perspective) that contains both Hawk-Eye’s Trajectory Breaks and derived Spin-Based Breaks. With our understanding on non-Magnus effects changing week-by-week, we hope that this can be a resource to the community as we continue to dissect and evaluate the true movement characteristics of a pitch.