P3 can identify six of the next ten knee injuries in the NBA

P3 research, injury risk: 6 of the next 10 traumatic knee injuries in the NBA, identified before they happen.
Athlete landing a hurdle hop during P3 training

The question

Can traumatic knee
injuries be identified
in advance?

The study at a glance: 400+ NBA players assessed, about 500 variables per assessment, two-year prospective follow-up, AUC 0.75 with sensitivity greater than 0.6.

The model

Machine learning
built on movement data

Athlete performing a lateral bound while P3 coaches observe

The signal

What the model
finds important

Athlete decelerating during a lateral movement at P3

The approach

Biomechanics,
not load management

P3 coach cueing an athlete through a movement drill

The intervention

From identification
to intervention

P3 staff in front of the P3 logo wall

For teams

What this means
for teams

Methodology note

P3 research: measure what matters, then build from it.