Method
Most programs are built forward from a template. Ours are built backward from a model. We test the qualities that matter, the model tells us which one is holding you back, and the program is written against that limiter, in the block where it responds.
What explains performance
Sprinting, jumping and change of direction are the physical outputs most field sports come down to. We model each one against the gym-floor tests that predict it. The share the model explains is the share a program can act on. The rest is reported as unexplained rather than sold back to you.
The split of the 50.9% between the two predictors is illustrative in this comp. The total is the published figure.
Reverse-engineered programming
Once we know which measurable quality explains the gap for an athlete in your position, the program stops being a guess about what to emphasise. Strength, reactive strength and speed each respond to different work at different points in a block, so the order matters as much as the exercises.
Receiver. Reactive strength is the limiter.
Strength is not the problem here. Adding a strength block would spend twelve weeks on the quality that is already where it needs to be.
Reactive block first. Speed exposure alongside it.
Illustrative athlete. In the platform the limiter, the bands and the block emphasis come from your own test.
The model in the real world
For Football Canada's flag football program we built a selection model from the testing battery and validated it the hard way, leaving each athlete out in turn and predicting them from the rest. It is the same machinery that finds your limiter, pointed at a different question.
National team athlete data is used only in aggregate. No individual's results are shared or shown.
Who we coach
Wording on CFL and Team Canada is yours to confirm before publish.
What a test actually says
A countermovement jump is usually reported as a height. The height is the least useful thing in it. The shape of the force trace tells you whether you are slow into the ground, weak out of it, or landing in a way that will cost you later.
The two marked in colour are the ones driving this athlete's programming. Height is along for the ride.
Return to performance
Clearance too often rests on one percentage. We report the joint angle, the absolute value and the limb difference together, from ordinary single-camera video. What the method cannot recover, it does not report.
Single-leg hop, markerless 2D pose, knee flexion on the camera-side limb.
Your first filmed sprint, your first logged sets. The model reads them, and the block is written against what it finds.
Build your program