Description

Send up to five twenty-second clips of yourself running — from the side, from behind, or a mix. Get back a measured analysis of your mechanics, a strength programme built from what it found, and a date to film again.

Most form analysis is someone watching your video and telling you what they think. This measures it. Ground contact time, stride and cadence, how far ahead of your body your foot lands, how much your pelvis drops, whether your knee falls in, and whether any of it holds up as you tire — read off the video frame by frame, with the numbers printed next to the frames they came from.

Then I read it and write the verdict. The measurements are the evidence; the coaching is mine.

Your report is emailed to you within 7 days of sending your clips, and usually sooner.

How it works

1. Film it. Twenty seconds per clip, on a phone, after ten minutes of running. From the side, from behind, or a mix — up to five clips.
2. Send it. One short form: your pace, your shoes, your height, anything that hurts. The clips upload straight from your phone.
3. It gets measured. Ground contact, stride, overstride, pelvic drop, left against right, and whether any of it holds up as you tire — read frame by frame.
4. I read it. Every finding traces back to a number I can check against the frame it came from. I write the verdict and the plan.
5. You get it. Emailed within 7 days, with your strength programme and a date to film again.

What you get

• The measurements, with the still frames they were taken from and the measured angle drawn on. You can see where every number came from.
• What stood out — the two or three things worth your attention, in priority order, not a list of nineteen numbers to worry about equally.
• Left against right. Where the two sides differ enough to be real rather than camera angle.
• Whether it holds up tired. The first half of your clip against the second. A fault that only appears when you fatigue is the one that decides races, and an average over the whole clip hides it by construction.
• My assessment. Written by me, after reading the above.
• A full strength & conditioning block built from what the video found — not a generic runner’s routine. Every exercise comes with what it targets, why a runner does it, how to perform it, where to start, and what to do when it gets easy or stays too hard. The exercise library is included: this is a programme you can follow, not a list of names to go and look up.
• A date to film again, and what that retest will actually be able to tell you.

The retest is the point

Anyone can measure you once. The harder question is whether anything changed, and most tools answer it dishonestly — a two-millisecond difference gets reported as progress when it is indistinguishable from filming you twice on the same day.

Every comparison here carries a floor: the smallest change this filming setup can actually tell apart from its own noise. A change smaller than that is reported as unresolved, not as an improvement. Sometimes that means the honest answer is “nothing has moved yet”, and you get told that.

It also knows how long each thing takes. Cadence responds inside a month. Frontal-plane hip control does not show up in two. So a measure that has not moved and was not due to is reported as on schedule rather than as a failure — you can see what is still coming.

Every finding is graded

Running has a lot of confidently-taught convention with very little behind it. So each finding carries a grade:

• A — the direction of the effect is not seriously disputed.
• B — a real signal, but it moves with your speed, your build and the surface.
• C — widely taught, weakly evidenced. Worth a conversation, not worth rebuilding your running around.

Nothing graded C will ever be given to you as a priority. You will know which parts of your report to act on and which parts are a talking point, because saying so is more useful than sounding certain about all of it.

One deliberate omission: there is no verdict on your foot strike. Heel, midfoot or forefoot has not been shown to predict injury or economy in either direction, and changing it moves load between tissues rather than removing it. It is reported and left alone.

The filming protocol

This part is half the accuracy, and it is the half you control. A camera that is not square to you does not add random noise — it adds a bias in a consistent direction that looks exactly like a finding. Ten degrees off perpendicular manufactures several degrees of pelvic drop in a runner who has none.

Full instructions come with your form. The short version:

Where	A treadmill if you have one. This is not a compromise — it is genuinely better. It holds you in the middle of the frame where the lens distorts least, it fixes your pace so two months can be compared, and it gives forty consecutive steps instead of the four you get running past a fixed camera. Set 1% gradient. Outdoors works, with more care.
Camera	Level with your hip, square on, three to four metres away, on something solid. A person holding it is fine; a person walking around with it is not.
Frame rate	Slow motion if your phone offers it. 120fps is what matters, 240 is better, 60 is the floor. A ground contact lasts about 220 milliseconds — that is seven frames at 30fps and fifty-three at 240. Resolution barely matters by comparison: 1080p at 120fps beats 4K at 30 every time.
When	After at least ten minutes of running. The first minute of anyone’s run is not their running.
How long	Twenty seconds per clip.
What to wear	Fitted. Shorts above the knee, top tucked in or close-fitting. Loose clothing hides the joints being measured.

Film each one twice if you can. Clips get spoiled and re-filming means another session.

If you want the most useful optional extra: a side clip at threshold pace. Form under load is the form that decides races, and it is often not the same form.

Both angles, one analysis

Film from the side, from behind, or both. They see completely different things, and neither can see the other’s:

• From the side — overstride, ground contact, knee flexion at landing, hip extension at toe-off, trunk lean.
• From behind — pelvic drop, the knee falling in through stance, how wide your feet land. The only view that can see any of it.

Send up to five clips, in whatever mix you like. One form, one set of questions, one report covering the lot — not five half-reports to read side by side. One clip is enough if that is all you can film, and the report says what the missing angle would have added.

Two or three of each is the sweet spot. A blend beats five of the same angle by a wide margin: a second side clip sharpens numbers you already have, where a first clip from behind adds three mechanics the side simply cannot see.

Extra clips of the same angle are not padding — they are pooled. Three side clips are three times the strides, and the smallest change a retest can resolve shrinks as the square root of that: measured, five clips cut it by around half on some measures. Which means extra clips do not make your report longer. They make next month’s comparison sharper.

Why it takes a few days

The measuring is quick — that part is done within minutes of your clips arriving. What takes the time is me reading it.

Every report is read and signed off by a person before it goes anywhere. Nothing is emailed to you automatically off the back of a measurement. That is the whole difference between this and an app, and it is worth the wait of a few days.

Within 7 days, always. Usually sooner.

What happens to your video

Your clips are used to extract body-position coordinates — a set of numbers describing where your joints were in each frame — and are then deleted. Usually within minutes of the analysis running, and never later than 30 days.

The coordinates carry no image of you, and every future comparison is rebuilt from them. So nothing is lost by deleting the video: you keep a comparable history for as long as you keep sending clips, and no footage of you sits on anybody’s server.

If you are under 18, a parent or guardian consents on the form rather than you.

Honestly, what this is not

This is coaching analysis of running mechanics. It is not a medical or physiotherapy assessment: it does not diagnose injury and it does not screen for it. If something hurts, is numb, or is not settling, that is a clinician’s question and not a camera’s.

The reference ranges are for trained runners at a steady submaximal pace. They are not population norms and not pass/fail cut-offs. Several of them — cadence above all — move with your speed and your leg length, which is exactly why the form asks what pace you filmed at.

And it is two-dimensional video, not a laboratory. It is very good at the things a single camera can see honestly, and the report tells you when your clip was not good enough to measure something rather than measuring it anyway.