Metric VBT validation & accuracy
Three independent peer-reviewed studies have tested Metric against the lab-grade tools coaches and researchers use to measure barbell velocity — 3D motion capture and linear-position transducers. Here's what they found.
Velocity accuracy vs lab-grade hardware
A correlation coefficient (r) is a 0–1 score of how closely two measurement methods agree on the same reps. 0.7+ is considered strong in sports science; 0.9+ is excellent. The numbers below are from Šagovac 2024, Metric v4.5 on the bench press, against a Vitruve linear transducer across 150 reps.
Validated against 3D motion capture
The reference our peer-reviewed studies use is a calibrated Vicon 3D motion-capture rig (twelve infrared cameras at 200 Hz) or a research-grade linear-position transducer. Both cost in the tens of thousands of dollars and don't leave the lab. The point of comparing Metric to them isn't to match them rep-for-rep — it's to show that a phone-and-camera setup that fits in your pocket lands close enough that velocity-based programming decisions hold up.
How Metric compares to VBT devices
Renner et al. (2024, PLOS ONE) tested three smartphone VBT apps against a Vicon mocap rig and a RepOne linear transducer across 589 reps. Lower error is better — root mean squared error (RMSE) is how far off each reading was, on average, in metres-per-second.
Validity vs reliability in VBT
Validity is whether the number is close to the truth: if a Vicon system says 0.98 m/s and Metric says 0.99 m/s, that's high validity. Reliability is whether the number is consistent: same lift, same setup, same reading tomorrow as today.
For training decisions — picking today's working weight, knowing when the bar slows enough to stop the set — reliability is the load-bearing property. A tool that's slightly off but consistently off is useful; a tool that's right on average but jumps around between reps isn't. Across the three studies above, Metric's intraclass correlation coefficient (ICC) for mean velocity sits between 0.79 and 0.98 — strong to nearly-perfect reliability across all three lifts.
Peer-reviewed validation studies
Summary articles for each validation study live on the blog with the raw correlation tables, study design, and links out to the original PDFs.
- Šagovac, A. — University of Zagreb, 2024
Validity & reliability of Metric in the bench press
Metric v4.5 vs Vitruve linear transducer · 150 reps · r = 0.93 mean velocity
Read summary ↗ - Renner, Mitter, Baca — PLOS ONE, 2024
Concurrent validity of three smartphone VBT apps
Metric v2.3.1, Qwik VBT, MyLift vs Vicon mocap + RepOne LPT · 589 reps
Read summary ↗ - Taber et al. — Int. J. of Strength & Conditioning, 2023
Validity and reliability of a computer-vision system
Metric v0.5.4 vs 3D motion capture · 800+ reps · ICC mean velocity 0.79–0.98
Read summary ↗ - Trowell et al. — 2024
Validation of a commercially available mobile application
Metric v0.6.0 vs 3D motion capture
Read summary ↗ - Metric in-house — 2022
Reliability & validity of the Metric VBT beta
Internal validation of an early beta. Baseline before the published peer-reviewed studies.
Read summary ↗
Running a study and want to use Metric as a study tool? We support raw-data export, methodology consultation, and the engineering team is available for technical questions. See the research page for the open invitation and contact details.
