Open-source motion capture data of elite-level baseball pitchers on GitHub
github.com
github.com
What's unclear is how long-lasting the performance improvements are (can the pitcher maintain spin rates without constantly going back to the data and coaching staff?).
One very positive thing I've seen is applying the analytics to young athletes as a future injury prevention tool. We now can, with high probability, identify kids who are on the path to permanently ruining their arm long before they cause lasting damage and give them an intervention - "You can dominate in Little League, but you'll never be a pro if you keep doing it this way. You must either change the way you throw or stop trying to be a baseball pitcher"
[0] https://www.basicbooks.com/titles/ben-lindbergh/the-mvp-mach...
Bauer. I had a bit of a meltdown about him - I was momentarily worried that his name alone would bring a huge negative reaction to people associated with Driveline training methods. That seems to have not come to pass.
I very much like the Driveline "Skills that Scale" training approach. I took several brand-new-to-softball middle schoolers into 50+ MPH bat speeds with overload/underload training and bat sensors. One great athlete hit 60+ in training in her very first year of playing the game. Course, she looks like a D1 athlete as well, so she may have gotten there just by putting a bat in her hand . . .
Mocap is a step beyond what I want to do with young athletes but having a data-driven approach is a coaches dream. Measure things that matter, build a constraint-based training environment, and let exterior outcomes shape players.
Look lower down at the pictures or pitchers (heh). The legs themselves aren't symmetrical.
So they probably tweaked the points so that the most common range of angles to make the math easier to do.
Like I said, just a guess. But with one leg bent and the other stretched out, they probably form a line close to parallel with the ground for right-handed pitchers at the point in time when the pitcher is transitioning from wind up to release.
Source: Am a biomechanist with several years of mocap experience.
Because of this, if you placed markers with bi-lateral symmetry, it would be much more difficult to solve joint rotation angles, as the overall facing of the point cloud would be undetermined. So, you have additional markers placed on none moving (not a joint) locations that ruin symmetry and allow facing to be easily solved. Typical locations are outside of the thighs, upper arms, and occasionally the back.
I built mocap studios for games companies.
But I like this answer better. It's a simple and clever solution to a real problem that you might not readily see as you're developing the solution.