Skip to content
Athlete archiveMeasure my first jump free

Measure → remember → compare → recommend → test → learn

Follow one real Dallin Thornton attempt from source frame to session pattern, then see how the next practice can make the recommendation more personal.

Follow the jumpSee the limits
  1. 01 · Capture

    One side-on phone video

    No markers, manual tracing or second camera. Dallin’s standing height and crossbar height establish the real-world scale.

    Dallin Thornton during the approach phase of jump 1326
    IMG_1326
    approach27.4 FT/S
  2. 02 · Measure

    The jump becomes evidence

    Jump 1326 resolves to 27.4 ft/s approach speed, 155° top-arm extension at plant, 20° takeoff, 169° inversion and 8′3″ push above grip.

    Dallin Thornton during the plant phase of jump 1326
    IMG_1326
    plant155°
  3. 03 · Remember

    One attempt joins the archive

    The video, context, measurements and confidence stay together. Nineteen attempts become a record memory cannot smooth over.

    Dallin Thornton during the inversion phase of jump 1326
    IMG_1326
    inversion169°
  4. 04 · Compare

    The pattern becomes visible

    Eight attempts completed the rock-back. Eleven stalled. Grip and top arm stayed stable while inversion varied 46%.

    Dallin Thornton during the push phase of jump 1326
    IMG_1326
    push+8′3″
  5. 05 · Recommend

    The evidence ranks one test

    Protect the stable grip and top arm. Test a complete rock-back. The recommendation resolves to Dallin’s own finished attempts.

    Dallin Thornton during the inversion phase of jump 1326
    IMG_1326
    inversion169°
  6. 06 · Learn

    The next session answers back

    VaultLab checks whether inversion improved without sacrificing approach speed or plant position, then refreshes the athlete model.

    Dallin Thornton during the push phase of jump 1326
    IMG_1326
    push+8′3″

The session, not the highlight

A best jump proves capability.
The archive reveals repeatability.

Dallin reached 169° on jump 1326. Across the complete session, only 8 of 19 attempts finished the rock-back. That difference is the product.

Dallin · 19 measured attempts8 finished / 11 stalled
01146°
0257°
0355°
04140°
05170°
06170°
0752°
0864°
0946°
1066°
1154°
1256°
13169°
14171°
15161°
16171°
17174°
18168°
19107°

Every column is a real jump. Hover or focus to read the measured inversion.

How Adaptive learns

No instant “personal AI.”
Evidence first.

Every athlete starts with measured evidence. As the archive grows, VaultLab can distinguish a one-off miss from the athlete’s recurring pattern.

  1. 01
    Measured attempt

    Five phases, source frames, confidence and one ranked finding.

    MEASURE
  2. 12
    Personal baseline

    Normal ranges and repeatability unlock across at least two sessions.

    BASELINE
  3. 30
    Adaptive archive

    Matched successes, drift and a confidence-backed next experiment.

    ADAPT
MEASUREDFeet, degrees, timing and repeatability

Directly tied to the uploaded attempt and its capture confidence.

MODELEDSimilarity, drift and athlete-specific relationships

Unlocked only when the archive supports the inference.

HUMANSafety, pole selection and coaching judgment

Always belongs to the athlete and coach.

Get started

Build the record your next PR can use.

Start with one measured jump. Adaptive learns only when your own archive earns the claim.

Measure my first jump freeSee a real session