Form score
How close your mechanics sit to the references calculated for you.
Every scored metric earns continuous partial credit for how far it sits from its target — full credit inside the band, then a smooth fade, never a cliff at the edge. Metrics the camera could not see clearly are weighted down, and the remaining ones are renormalised, so you are judged on what was actually visible rather than penalised for a bad angle.
What it does not tell you: It moves with the profile you pick, because the profile changes how much tolerance each band gets. Use the comparable score to track yourself over time.
Comparable score
The same run against one fixed rubric, so it means the same thing every time.
Identical measurements, scored at a fixed tolerance and using only the side-view metrics. That makes it independent of the profile you chose and of whether you happened to film a rear view — which is exactly what a progress chart needs. This is the number History stores.
What it does not tell you: It ignores the rear-view metrics by design, so it is not a fuller picture than the form score — it is a more stable one.
Injury Load Index
How your mechanics compare on patterns published cohorts associate with injury.
A weighted read of six measurements, each weighted by the effect size reported in the source study and discounted by how reliably it can be measured from video. Only the contributors actually measured are counted, and the weights are renormalised over them.
What it does not tell you: It is an index of load patterns, not a probability, and not a diagnosis. From a side view alone the two strongest contributors are invisible, so its confidence is capped. It says nothing about your training load, history, or tissue tolerance — the things that actually decide whether you get hurt.
Economy & speed
The four kinematic terms that predicted race performance in the largest study of its kind.
Braking, vertical oscillation, duty factor and shank angle at touchdown. In a study of 97 runners from recreational to elite, those four together explained 31% of the variance in season's-best time and 39% of running economy.
What it does not tell you: 31% is a lot for running technique and still leaves most of performance to physiology, training and genetics. A high score does not make you fast on its own.
Across the run
How much your stride varied, and whether it drifted.
Stride-to-stride variation in contact and stride time, plus the change from the first third of the clip to the last.
What it does not tell you: These are observed statistics with no claim attached. There is no published 'correct' amount of stride variation, so nothing here is a target — it is a description of what your run did.