By Anthony Brady MS, CSCS, Director of Sports Science
In 2019, Dillon Tate made his MLB debut. We’ve been fortunate to get to work with him each offseason since then, giving us a unique dataset to work with—longitudinal mocap sessions over 3 years for a big league pitcher. Year over year, his fastball average velocity and peak velocity has consistently increased.
Success Leaves Clues
It’s impossible to summarize all the factors that went into making these plots possible. To go from averaging 92 to 94 mph —> 95 to 96 mph in just one to two years is actually ridiculous. You could write tens of thousands of words attempting to summarize the work that happened consistently—from the weight room, the Plyo Ball ® wall, recovery, as well as programming decisions, drill modifications, and cues from the Director of Pitching, Bill Hezel, at each step along the way.
While we can’t get into everything that happened, we can dissect the biomechanics reports that we’ve collected over the last few years to glean clues that led to this success. And that illustrates the true power of these biomechanics reports at the elite level.
We have actionable and objective data quantifying how Tate moved each offseason. There’s no guesswork here. Having multiple synced high speed camera angles of the pitching motion can tell us a lot. But it can also trick us and it’s easy to let coaches bias happen without having numbers to actually back it up.
Thanks to the mocap lab, we have data points that quantify Tate’s exact positions and movements in space down to the millimeter at 240+ frames per second.
The First Assessment – January 2020
His average velocity during this session was 93.3 mph. Which would qualify him for the elite group in our database that helps build out our normative ranges and programming recommendation algorithms. We knew there was more in the tank and his ceiling was much higher than low to mid-90s. When he was drafted in 2015, his fastball velocity was up to 98 previously.
We don’t know exactly how he moved in 2015, further highlighting one of the biggest value adds behind the motion capture report. Right now, we have multiple reference points of Tate’s mechanics each year to refer back to if something ever feels wrong. We can look back and identify what he was doing when things were going well and attempt to correct based on that objective data.
There’s plenty of low-hanging fruit to pull from his earlier biomechanics reports. Here are a few things that stuck out:
- Creates a good amount of scap retraction or shoulder horizontal abduction (SHA) to start, but there’s more in the tank. In our upper 90s throwers, we’ll sometimes see > 70 degrees of SHA.
- Maximum External rotation is lower than we’d expect to see from someone already throwing this hard.
In the plot below you can see Tate’s SHA overlaid on top of the normative ranges of some of our harder throwers. He’s right in the middle throughout the motion (not surprising since he is one of our harder throwers already):
To illustrate scap retraction even further, here’s an overhead look at his pitching mechanics showing that ~40 degree angle between his torso and humerus around foot plant
Hip Shoulder Separation
- Trunk comes out of counter-rotation early and is too open at foot plant (~17 degs)
- This is making it hard to create hip/shoulder separation
Again, you can see that nearly all of these angles fall within the normative ranges for our elite throwing population. This shouldn’t be a surprise, it’s a big league pitcher throwing 93 mph. But the torso angle and hip shoulder separation angle are both on the fringes of falling out of those normative ranges. Video below show the interaction between the trunk, pelvis and how it creates hip shoulder separation. In this throw specifically, his trunk is open to nearly 25 degrees at foot plant.
We can’t get into everything that showed up in the report biomechanically, and most of it was good anyways. We’re working with an already athletic big leaguer that is throwing low – mid 90s velocity. We’re chasing gains on the margins to find where his ceiling truly is. And the focus around that put simply was:
- Create more scap retraction
- Hold trunk counter-rotation a little longer
- Keep our eyes on that maximum external rotation
These bullet points are all likely related. If the trunk rotates early, it’ll pull the humerus out of scap retraction early, wasting the scap retraction that you are creating. An early trunk and low scap load is also going to make it hard for the arm to rest back deeper into layback (increase MER). As the humerus wraps around the already rotating trunk, the window to get into MER shrinks quickly.
We’ve now taken millions of data points that quantify movement and distilled it down into a plan of attack. This is what it looks like to leverage biomechanics data.
How did we do?
The velocity numbers in-season speak for themselves. I mean seriously, look at these plots again ⬇️
Obviously velocity is not the only thing that matters here, but we know it’s value, we’ve quantified it before. For anyone that hasn’t read that series by Dan Aucoin, here’s one of the most important excerpts:
“We took a hypothetical player, held his IP constant at 75 (a hedge between a starter and reliever), and calculated his WAR before and after a .2443 point reduction in RA9. We found that adding 1 mph to a player’s FBv increased the expected production of said player by ~.25-.35 WAR per season, depending on whether he was a starter or reliever.
Tate has small samples in 19’ and 20’(due to the pandemic) but you can still parse out the value of this velocity increase on RA9, FIP, and WAR.
So things are looking good on Baseball Reference, but what (if anything) actually changed mechanically?
The Retest – January 2021
The velocity increase that Tate had in season showed up in the lab too, running his fastball up to 95.4 mph during the mocap (averaging 94.7 mph). Here are two synchronized pitches from January ‘20 (Red) and January ‘21 (Green). Looks pretty similar right? Hard to tell the differences even when they’re throwing side by side. It gets a little easier when they’re superimposed on top of one another. With biomechanical data, we can quantify these differences.
Arm Action Changes 2019 – 2020
- More scap retraction in total (46° —> 58°) and held more into foot plant (42° —> 55°)
- Maximum external rotation increased as well (171° —> 183°)
Visual comparisons below:
Assessment – Jan 2020 93.3 mph
Retest – Jan 2021 95.4 mph
It’s a even easier to see the difference when having them side by side too:
Hip Shoulder Separation Changes 2019 – 2020
One thing you can notice in the last video is how trunk angle changed into foot plant. The trunk is staying counter rotated longer which is helping create additional scap retraction. This is huge for the hip shoulder separation changes we were looking to make as well:
- Trunk angle stayed more counter-rotated into foot plant and for longer (17° —> 11°)
- Hip shoulder separation values increased at foot plant (20° —> 31°)and maximum (23° —> 33°)
More visual comps below:
Assessment – Jan 2020 93.3 mph
Retest – Jan 2021 95.4 mph
Side by side visual below showing the differences in trunk angle(dashed line) and hip shoulder separation(solid line):
This doesn’t cover everything that changed biomechanically, but these are a few of the most notable comparisons between the sessions. They also highlight some of the most important features of pitching mechanics that we see internally.
Can we definitively say that the increases in scap retraction, better trunk angle, and increased hip/shoulder separation led to his velocity boost? No.
But we can make a pretty damn strong case with just the evidence presented above. It stands to reason that those changes had a significant effect on his arm speeds as well. Look at the increase in maximum elbow extension and internal rotation angular velocities between sessions
Being “data-driven” is easier said than done. The summary at the beginning looking at the initial motion capture report is fairly easy after you’ve done it for awhile. We could have stopped there. “Hey, here’s a plan that we think could work, good luck.” We could have taken the velocity increase from 2019 – 2020 and just ridden on the wave of positive results.
But that’s not really being data-driven. We needed another data point to see what happened between January of ‘20 and January ‘21 that could explain such a significant velocity increase. You have to be willing to be wrong. Regardless of how convicted you are in the assessment and plan that you laid out, if you’re not willing to look at the data again and see if it backs it up or refutes your initial theory, then you shouldn’t have bothered collecting it in the first place.
That’s also the best way to learn more throughout the process. We’ve had biomechanical models and theories for mechanics that have been wrong before and we’ll be wrong again. But because we’re willing to be wrong, we’ll always stay at the front of mechanical analysis and leveraging biomechanical data in a player development setting.