score
0
streak
0
precision — 0%
// reaction model log-normal fit
bias = — ms  (systematic tendency) σ_log = —  (log-normal spread) window = — ms  (cycle-fraction capped) hit rate target ≈ 85% (empirical quantile, n≥10)
// timing error distribution
// session stats
samples (n)0
mean bias—
log-spread—
window size—
regimeslow
difficultycalibrating
adaptive reaction timing instrument
PULSE
Tap as the ring closes. The window adapts in real time using nonparametric quantiles of your reaction time distribution — keeping you at exactly 85% accuracy, always at your personal limit.