Field notes · 8 Oct 2026
First court test
One player, one session, the first prototype. We filmed it and lined the video up with the sleeve's own log.
Video, 37 seconds, no speech. Real footage from the first court test on 8 October 2026: one player shoots free throws wearing the Betta sleeve, filmed from the floor under the hoop, so the rim is out of frame. The sleeve's own log runs beside him, lined up with the video to within about 0.05 seconds. Rep 6, in real time and then at half speed: pulses while his hand is close but not yet set, one firm buzz when he is set, then the shot, the hold, and one long buzz on both motors about 0.8 seconds after the flick, a good rep. Then rep 18, paused at the release: his hand came no closer than 12 degrees to his set point, and set means within 9, so there is no set buzz and no good-rep buzz. It ends on a key to the three cues: pulses, close but not set yet; one firm buzz, set; one long buzz, a good rep. Each vibration he felt plays as a low tone.
What he feels
What it can't tell yet
- Makes and misses. The rim was out of frame. The buzz checks his form (set point, flick, hold), not whether the ball went in.
- Whether the ball is in his hand. The sleeve can't see the ball.
- One player, one session. Too few to say how it works for anyone else.
What we're changing
The player and coach said the weight was no issue. What they want fixed:
- Motors held in place. They slipped out of position.
- Sensors you can reach for repair. One stopped working before this session, so these reps ran on the hand sensor alone. That one dropped out 4 times while he shot, for 4.5 to 6.8 s each.
- Lower on the palm. It sat too high.
- Either arm. No thumbhole, so one sleeve fits left or right.
Next test
- The rim in frame, so makes and misses can be counted.
- Filming from before calibration.
- The live feed kept running to the last shot.
- More players.
Method: phone video at 30 frames a second, filmed from the floor under the hoop. Sleeve log at 20 samples a second. We lined them up to within about 0.05 s by matching the ball leaving his hand with the sleeve's peak hand speed.