Evidence posture
We read the same literature
you do.
Most screening products oversell what one test can know. Ours is built on the opposite premise: say exactly what the evidence supports, deliver prevention to everyone, measure change honestly, and put a licensed professional in the loop the moment one is warranted. Clinicians tend to trust the platform because of what it refuses to claim.
The position
What the literature supports
Multicomponent neuromuscular prevention programs meaningfully reduce lower-extremity injury risk in youth athletes — especially young female athletes — when delivered consistently. That is why every participating athlete receives the universal program, not just flagged ones.
What it does not support
No single field test — drop jump, phone-video angle, or composite score — reliably predicts which individual athlete will sustain an ACL tear. We know this literature well, and we built the product so it never has to pretend otherwise.
What follows from both
Screening is for finding modifiable deficits and establishing baselines; prevention is for everyone; monitoring over a season beats any single snapshot. The product is the model — screen, train, measure change, refer when warranted — not the screen alone.
Measurement conventions
Distance-based frontal metrics
Frontal-plane control is read with distance-based, event-normalized measures rather than raw video angles, which are the least robust markerless output. Where a measure is lower-confidence, it is labeled lower-confidence.
Change vs. measurement error
A score difference is reported as change only when it exceeds the measure’s established error. Otherwise the report says “no confident change” — improvement is earned, not asserted.
Asymmetry, banded
Interlimb differences are reported in screening bands with direction preserved — a conversation starter for a clinician, not a diagnostic cut-point.
Versioned everything
Protocols, algorithms, and reference data are versioned; every result records which versions produced it. Threshold changes go through documented governance, never silent recalculation.
Governance
Research-first verification
Clinical parameters are verified against multi-source, peer-reviewed literature before deployment.
Licensed advisory review
A clinical advisory bench — licensed physical therapy and biomechanics expertise — reviews protocols, thresholds, and athlete-facing language.
Deterministic safety path
Scoring is rule-based and auditable. No generative AI produces a tier, suppresses a red flag, or touches the safety path.
Red flags stop the machine
Pain, instability, swelling, or acute symptoms suppress scoring entirely and route the family to licensed care. The platform never coaches an athlete through a red flag.
Better Athlete is not a medical device and does not diagnose, predict individual injuries, or clear athletes to play. Full references and the current algorithm documentation are available to network members and research partners on request.
Continue
The provider packet →
What arrives with a referred family — and why every number in it is traceable.
Recovery episodes →
Post-op milestones, adherence, and the return continuum — anchored to the pre-injury baseline.
Join the network →
Membership tiers by practice size, and the compliance posture behind them.