Technique is still judged by eye.

The tools that measure movement properly live in labs. The athletes who need the numbers have a phone. We built for the phone.

For decades the only way to measure a human in motion with real precision was a room of infrared cameras and reflective markers, run by a biomechanist. That is still where the gold standard lives — and it is why almost nobody who trains ever sees a number about their own technique.

Most coaching happens from the bank, the sideline or a phone screen. It is expert and it is honest, but it is an estimate: which stroke, how many degrees, better or worse than last week. The research says the details matter — in one study of injured runners, each extra degree of pelvic drop raised the odds of injury by about 80 %. A few degrees is not something the eye reliably catches.

Mocta-X started as a rowing club's problem: a coach with a squad of athletes, one launch, and no way to give everyone a same-day read of their stroke. The answer had to work with the phone already in the kit bag.

What's broken

Technique is judged by eye

A coach on the bank sees a handful of strokes from a moving launch. A runner films themselves and squints at a phone. The eye is good at noticing something is off and bad at saying by how much, or whether it is better than last month.

Video without numbers is just video

Slowing a clip down shows the fault. It does not measure the catch angle, count the strokes where it happened, or separate a bad rep from a bad session. The important patterns — a pelvic drop of a few degrees, a finish that fades late in a piece — are small and cumulative.

Feedback arrives late, or not at all

Athletes in a club queue for a coach's attention; athletes on their own get none. By the time a habit is pointed out, it has been rehearsed for weeks. What is missing is a same-day read that names the one thing to change.

The access gap

Measured technique is not rare because nobody wants it. It is rare because the equipment, the operator and the visit have stayed expensive, and because consumer tools have mostly sold confidence they could not back up.

Markers, cameras, staff
What a motion-capture lab needs

Optical capture is the gold standard, and it stays in research institutions and elite programmes because it needs a room, a technician and a booking.

Weeks, then a PDF
What an assessment usually returns

Even where a lab exists, a single visit produces a report about one day. It cannot tell you whether the change you made on Tuesday held on Thursday.

One phone
What the athlete already has

A side-on clip from any phone carries enough information to place 33 landmarks on every frame and grade the movement — if the software is honest about what it can and cannot see.

/// Why this is solvable now

Three things changed.

1. Pose estimation got good enough.
Single-camera models such as MediaPipe Pose now place 33 landmarks per frame, and independent studies put their gait-timing error at tens of milliseconds against marker-based capture. Not a lab — but enough to coach from, if you are honest about the limits.
2. The phone became the camera.
A side-on clip from any modern phone is enough. The web app needs nothing installed, and the iPhone app films, trims and uploads in one flow.
3. Language models learned to use tools.
A frontier model can now query the graded data — strokes, angles, history — before it writes, and cite the frame behind each claim. That turns a score into a coach's read instead of a spreadsheet.

That's what we built Mocta-X Dashboard for.

Rowing, running and golf graded stroke by stroke from a phone clip, with a coach that shows its work. Free plan, nothing to install, and an iPhone app on TestFlight.

Open Mocta-X Dashboard