Seam Shifted Wake — First Looks at NCAA Pitchers
Note from the editor, Chad Raines: This article is written by two of our student analysts. Joey Mylott is a master’s student in biomedical engineering and does biomechanics research in the Wake Forest Pitching Lab. Nihar Maskara is a sophomore mathematical business student. Both students work on our Wake Forest Baseball Analytics team.
With many recent advancements in pitch-tracking technology, pitch movement profiles are getting more and more accurate. The optically enhanced dual radar-based unit TrackMan can produce metrics on the vertical and horizontal break of pitches (+/- 1 inch) along with spin rate (+/- 20 rpm) and spin axis (+/- 3 degrees). While Rapsodo does not release similar accuracy metrics, Driveline Baseball performed a validation study in 2016 of the optical-based Rapsodo 1.0 unit, which was elevated behind home plate. Their results show that Rapsodo outputs accurate metrics and is a viable, low-cost alternative to TrackMan. The more recent Rapsodo 2.0 unit, which sits on the ground 14 feet in front of home plate, has only increased the accuracy of its metrics, but may lose some accuracy as it does not capture the entire ball flight. However, not all similar metrics from different tracking systems can be directly compared because of how differently they are set up and how the calculation changes based on that fact.
Since radar and optical-based tracking both have unique advantages and disadvantages, using a combination of both types of tracking technology can provide powerful insights into pitch movement. Other systems that quantity pitch characteristics exist as well, using high-speed cameras, custom computer vision software, and a combination of other methods. The new HawkEye system used in all MLB stadiums utilizes both radar and optical technology together with an array of cameras and devices to create an all-encompassing in-stadium system. Since TrackMan doesn’t have optical technology to measure the exact spin axis of the ball at release, it infers what the spin axis was based on the observed movement of the pitch.
As far as pitch movement, the most commonly reported metrics are induced vertical break and horizontal break. These values break up the movement of the pitch into two directions. All forces acting on a baseball as it travels towards home plate include the gravity, drag, Magnus effect, and seam-related forces. Now, the only two known physical forces that act on pitch movement are the Magnus effect and seam-related forces. The former of these two phenomena, which comes from the spin of the baseball, is the more common and well-known factor. Essentially, the direction of the ball’s spin creates differences in air pressure around the ball during its flight towards the catcher. These air pressure gradients impart force on the baseball as it travels towards home plate. For a fastball, the pressure difference creates an upward force that opposes gravity and causes the pitch to fall less than it would if gravity was the only force acting on it.
The pressure areas around the baseball change locations based on the ball’s spin axis, and the movement of the pitch will follow the resulting Magnus force. This is why 12–6 curveballs drop off the table and perfect sliders dart sideways across the strike zone.
One of the most recent findings using different combinations of pitch tracking technologies is that second mechanism or force creating pitch movement, which uses the seams. The seam positioning can actually create pitch movement independent of spin. This phenomenon has been named Seam-Shifted Wake (SSW) by Dr. Barton Smith, a professor at Utah State University who is an expert in researching the topic. He has found that placing seams within a certain region of the ball that is near the anterior/posterior centerline of the ball for the majority of the pitch’s rotation creates a deflected wake of air flowing past the baseball.
When the seams remain in this region for the majority of the pitch, the effects of the deflected wake are realized in movement that differs from the expected movement profile from spin-based movement alone. The seams’ interaction with the air is not counteracted on the next ½ or ¼ rotation like with a typical, efficient 4-seam fastball, so the effect of the seams persists for the entirety of the pitch’s flight. A pitch with a SSW will push the ball flight in the direction away from the seams that are able to stay in the important region. Dr. Barton Smith, summarizes this topic extremely well in an article from his blog “Baseball Aerodynamics” that can be read here.
The major metrics of interest for SSW are inferred axis, observed axis, spin efficiency, and gyro angle. A quick summary of these metrics are as follows. The inferred axis is calculated solely from the movement of the pitch, and we used TrackMan to measure this metric. The observed axis is the axis of rotation of the pitch that is observed directly, and we measured this metric using Rapsodo. Spin efficiency is a loaded term that is often confused and misused around the baseball community. For this article, spin efficiency will refer to the metric as measured by Rapsodo, which is the “percentage of raw spin that directly impacts movement” (Aguiar). Gyroscopic spin (sometimes referred to as football spin or bullet spin) is perpendicular to the direction of movement. Gyro angle is measured between a line around which the pitch is spinning and a line between the pitching rubber and home plate.
A pitch with SSW characteristics will have an inferred axis that is different from its observed axis, meaning that the pitch responds to both Magnus force and non-spin induced forces. The difference between these two axes is referred to as axis deviation and is the biggest indicator of a pitch with SSW right now. Given all of this information, the pitches that have the best potential to exhibit SSW effects are fastballs, sinkers, and changeups all with <90% spin efficiency. Sliders and cutters are affected by SSW too, but the spin axis and gyro angle of these pitch types are extremely variable due to a variety of different grips, making them difficult to study.
The reason pitches with <90% spin efficiency have the best potential to exhibit SSW is because the orientation of the baseball due to the gyroscopic spin (football or bullet spin). In order for the seams to remain in the proper region around the “hemisphere plane” and impart SSW effects, gyro spin is extremely important. Barton Smith termed a pitch the “looper” and it is the only real way that we know of to throw a SSW pitch with 100% efficiency (no gyro spin). Since gyro spin does not contribute to spin-based (Magnus) movement, it is inversely related to reported spin efficiency metrics. However, this element of spin allows for the seams to be positioned to where SSW effects are noticeable and will create movement on the pitch that is not caused by spin. Most other SSW pitches employ gyro spin and find the most benefit when the spin efficiency drops below 80–90%, depending on the pitch type.
The need for this gyro spin stems from the pattern of the seams on the baseball, which is essentially arbitrary. Outside of the “looper”, gyro spin is necessary, but not the only important factor. A pitch with the proper amount of gyro spin will not exhibit SSW effects unless it has the proper orientation with the seams in the proper region. However, it’s not like a pitch either has SSW or it doesn’t. The phenomenon is not binary. Every pitch has some degree of SSW, but on most pitches, the force is basically negligible. Some pitchers may already be experiencing SSW effects on their pitches without realizing it. Ideally, pitchers would want to be able to control the amount of SSW they put on a pitch, just like they do velocity, spin axis, and spin rate.
Intrigued by the physics and performance implications of SSW, we decided to look into the phenomenon among the Wake Forest pitching staff. For this project, paired pitch data from TrackMan and Rapsodo 2.0 were used to obtain the necessary data points to identify SSW. Here are the plots of different pitch types for the Wake Forest Pitching staff that we used to identify pitches that already or potentially could take advantage of SSW effects.
The goal of looking into SSW effects on our pitchers and quantifying the different metrics involved is to open up another avenue of data-driven player development and pitch design. The colloquial term in baseball for what is now called SSW is “late life” or “late break” on a pitch. This phenomenon is not necessarily a new discovery, but the quantification of it is and can have large implications. All pitches that effectively utilize SSW effects create a movement profile that is different than the norm for a comparable pitch with only Magnus based movement. Pitchers who may not have elite velocity, lack hop on their fastballs, or have any other knock against their repertoires can start exploiting how to create this “late life” or SSW effect to create better movement on their pitches. Ideally, future pitching development will be able to target this effect relatively easily with the assistance of previously mentioned pitch tracking technologies.
One way to capture how unexpected a pitcher’s movement is would be to compare their horizontal and induced vertical break to the rest of the league.
A very insightful article was recently published by Kevin Goldstein on Fangraphs that talks about the importance of pitch shape, and specifically fastball shape. The link to the article can be found here. In the article, he shows examples of different MLB pitchers and how their fastballs move relative to the norm.
Pitchers that deviate from the norm are usually able to find success because hitters aren’t used to seeing a pitch with that type of movement. While the Fangraphs article doesn’t talk about SSW, it definitely plays a role in creating a unique pitch shape. Based on our current understanding of the effect, we think that certain pitchers will benefit from utilizing SSW because they will be able to get more unexpected movement on their pitches.
One way to examine if a pitcher could improve from using SSW is by creating a plot of overall pitch movement. By mapping out the horizontal break and induced vertical break of every fastball thrown in the NCAA so far in 2021, we can examine how the pitcher’s on our staff compared to the rest of the league. A few examples of these plots can be found below.
A few notes:
- College data for pitch types are not as reliable as MLB data. Each NCAA team that uses a TrackMan unit has someone that manually tags each pitch. While TrackMan tries to predict what the pitch type was based on velocity, spin rate, and movement, it is not that reliable and it gets confused by similar pitches. Despite this, since we are looking at all the NCAA teams that are in the TrackMan sharing network, after doing some data cleaning to remove obvious abnormalities, our sample size was large enough to get a fairly clear picture of what the typical break of a fastball looks like based on where most of the points land on the coordinate plane.
- Even though the y-axis says Vertical Break (VB), it is actually plotting the Induced Vertical Break (IVB) of each pitch. The difference between the two is that IVB takes out the effect of gravity, so a pitch with + IVB would theoretically rise if not for gravity.
Looking at these fastball break charts, you’ll notice how each pitcher’s fastball has a slightly different look. Pitcher 1, who is on the top left, gets more arm-side run on his fastball than average, while Pitcher’s 3 & 4, get more glove-side movement than average on most of their fastballs. The big difference between the fastball shape of those two pitchers is that Pitcher 3 is often able to get more vertical break on his fastball than Pitcher 4, which makes the pitch look like it’s “rising”. Looking at the plot on the top right, you can see how Pitcher 2 has the most “normal” fastball of the 4 Wake Forest pitchers. While his average horizontal break is very close to the average horizontal break of the NCAA, he usually is able to throw fastballs that have a higher vertical break than the average NCAA pitcher.
In order to try and figure out which of the four pitchers, if any, could potentially benefit from SSW, we wanted to get Barton Smith’s opinion on the plots shown above. After looking at them, he mentioned to us how the pitches with a negative horizontal break are SSW four-seamers since the effect is in the opposite direction as a two-seamer, so it moves more glove-side while often having less vertical break.
As we have previously mentioned, SSW is a continuous spectrum, so some pitches will have more SSW than others. The pitches with more SSW will often move more glove-side since SSW often acts as a side-force. While it can act in any direction, Magnus force doesn’t affect side force, so if a pitch is getting more side force, it could be explained by SSW. Based on these characteristics, it looks like Pitcher 4 could already be throwing fastballs with SSW (albeit a minimal amount).
One of the projects we hope to tackle in the future is separating out the total force acting on a pitch into its individual components and calculating the run value that each force contributes. Glenn Healey and Lequan Wang have already done something similar with MLB pitchers using Hawkeye data and they presented on this topic at the 2021 Society for American Baseball Research (SABR) Virtual Analytics Conference in March. A description of their presentation can be found on the SABR website, and the recordings of all SABR presentations will be released later this summer. Below are two examples of their plots that Baron Smith shared in a Twitter thread following Healey and Wang’s SABR presentation. We hope to create something similar with the Wake Forest pitchers based on the data we are currently collecting on SSW effects, existing models of how to separate Magnus-based and SSW-based movements from a pitch’s total movement, and the run value of in-game pitches for our staff relative to the rest of the NCAA.
As you can see in this example, the first force vector brings the pitch almost to the center of the plot, in the blue-greenish area, which has a relatively high intrinsic pitch value (not good for pitchers). After adding the second force vector, the run value of the pitch drastically drops and goes to the dark blue area, which is good for pitchers since they want pitches that have a negative intrinsic pitch value, corresponding to less run produced on that specific movement profile for that type of pitch. For pitchers that end up in a high run value area after measuring the drag + Magnus vectors, utilizing SSW (which is commonly thought of as side force), could help the pitcher get to a low run-value environment, which is the goal for all pitchers.
Here in the second example, we see the opposite case. The first force vector brings the pitch towards the top left corner of the plot, in the yellow-greenish area, which has a relatively low intrinsic pitch value (good for pitchers). After adding the second force vector, the intrinsic pitch value actually increases and goes to the bright yellow area. For this pitcher’s sinker, the side force (SSW effect) reduces the value of the pitch and should be removed to create a more effective pitch that leads to fewer runs.
As we look into this topic more, we will be using in-game TrackMan data for Wake Forest and the rest of the NCAA to calculate the run value of pitches. We will most likely compare pitches with similar movement profiles to the average movement profile of that pitch type within a small range of release heights. This will give us a better idea of how our pitchers’ offerings compare to those that are most similar from the rest of the NCAA, and hopefully allow us to start separating pitches based on potential SSW effects.
While we have only talked about our initial findings, we hope to gain a better understanding of SSW effects using paired tracking technologies on the same pitches. Similar to how some pitchers are better when attacking certain parts of the strike zone, SSW could increase a pitcher’s effectiveness with unexpected movement, especially when located well and paired with their other offerings. If we were able to identify which pitchers would benefit most from trying to use SSW, this ability would create a distinct competitive advantage. The new knowledge we gain from our research will hopefully be used to alter pitches to more effectively extract SSW-based movement, differentiating them from the norm to produce more whiffs, more takes on strikes, and increased success in-game.
