09/07/2026
New Driveline Baseball research is out:
Adam Bloebaum pulled 28,307 game-speed fastballs from 1,722 athletes, youth through pro, to test one question:
Is elite movement less complex, more complex, or just more repeatable?
The preprint is up on SportRxiv, free to read.
The answer is repeatability, but not the kind youโre picturing. Nobody repeats a delivery down to every angle and position. Bernstein named that problem sixty years ago: repetition without repetition.
What elite throwers repeat is the nonlinear signature of the movement, not the trajectory that produces it. Average complexity at the trunk and pelvis is roughly flat across velocity tiers. What changes is how much that complexity changes pitch to pitch. Trial-to-trial variation in pelvic spectral entropy was 57% lower in the 90+ mph group than in the sub-85 group.
It also lives in the proximal engine, not the arm. Trunk and pelvis carried the largest effects; the throwing shoulder sat further down the list. The arm reads as a downstream consequence of a stable pelvis and torso rather than the skill itself.
The assessment implication is the sharp one. A single throw cannot see any of this. Standard point-of-interest numbers off one pitch predicted velocity at Rยฒ = 0.21. Nonlinear features from the full movement time series hit 0.41, and the trial-to-trial variability of those features tracked velocity harder than their averages did.
Bottom line: elite pitchers do not repeat their trajectories. They repeat the structure underneath them. If youโre evaluating a delivery off one throw, youโre missing the thing that separates the top of the pool.
