Engineering @DrivelineBB | Pitcher @Nationals
Joined March 2019
- Tweets306
- Following383
- Followers612
- Likes8.6K
Pinned Tweet
My first patent was published today - super proud of the work @clayton_t22 and I have done to get here. Leading the way @DrivelineBB
Adam Bloebaum retweeted
Capacity for expression of skill should be a carefully measured and periodized application of general, technical, and tactical stimulus that allows the work to always be challenging, but within the athlete’s capacity, driving adaptation.
Anything less is a waste of your time.
Adam Bloebaum retweeted
First Live back from TJ, matched lifetime PR of 94.5, went 18/20 strikes and 4/4 punchouts.
@bl0ebaum has put together some of the best planned and best executed last 13 months of training I’ve ever seen.
Adam Bloebaum retweeted
A term used frequently in elite athletics is "calibration".
Here's how IMUs (Inertial Measurement Units) can teach about what calibration really is, and how to train it⬇️
Adam Bloebaum retweeted
🚀 New research preprint by @bl0ebaum of @DrivelineBB R&D! 🚀
Reproducibility of nonlinear movement complexity, not complexity itself, distinguishes elite throwing
Conclusions:
- Elite throwers do not throw with less complex (or more complex) nonlinear joint trajectories on average.
- They throw with substantially more reproducible nonlinear signatures across trials within a session, and the strongest signal lives at the pelvis and trunk rather than the throwing arm.
doi.org/10.51224/SportRxiv.9…
Adam Bloebaum retweeted
Big research with the dawg @bl0ebaum
sportrxiv.org/index.php/serv…
Adam Bloebaum retweeted
At the range today, another great opportunity to talk Control, Command, Accuracy, and Precision⬇️
Skill development isn't magic. It's a search problem → an optimization problem → an infrastructure problem. The brain explores, converges, then upgrades its physical structures. These processes can be represented quantitatively, giving us a better idea of what's going on during motor learning and what variables can be manipulated. 1/
Once the solution is known, the physical structures must get upgraded to for execution to become automatic. Myelination increases signal velocity, neural outputs synchronize, and metabolic cost decreases. 5/
1
1
284
Good training design uses these principles. Early on, maximize variability — open constraints, diverse contexts, rich feedback. As a solution emerges, narrow the constraints and add repetition and volume to deepen the attractor well. Once it's consolidated, layer in pressure, fatigue, and context switching to stress-test the infrastructure. Always funneling from wide exploration to specific expression, whether within a single session or across a development period. 6/6
1
5
243
Adam Bloebaum retweeted
A 100mph fastball is 144 Joules of kinetic energy.
A pitcher at peak leg lift has 1,204 Joules of gravitational potential energy.
The mound stores more energy than the ball needs. The body just can’t transfer it all cleanly.
Throwing is thermodynamics. 🧵
Adam Bloebaum retweeted
Congratulations @hedgertronic on signing with the Texas @Rangers ‼️
Check out Josh's outing during Pro Day in Tampa, FL.
Adam Bloebaum retweeted
two weeks after releasing RF-DETR segmentation, @DrivelineBB started testing it in their baseball pitch analysis pipeline
it only took ~30 minutes to train this model
link: github.com/roboflow/rf-detr
RF-DETR-Seg outperforms other real-time instance segmentation models, including the YOLO 26 segmentation models released last week
link to my stream covering YOLO26: youtube.com/watch?v=sczVknSv…
Adam Bloebaum retweeted
The ‘99 Spurs were ripping Starcraft LAN parties in the hotel before destroying your favorite team LOL