Touch Target Fitts's Law 2D: Live Regression Model

The 3D version of this simulator visualizes Fitts's Law from the outside: you watch a finger fly to targets and see a hit-rate number update. This 2D companion turns the same physics into a working scientific instrument. Each simulated tap still obeys the identical model — movement time predicted from the index of difficulty log2(D/W+1), landing point perturbed by Gaussian aim error that grows with the motor-impairment slider — but every (index of difficulty, observed move time) pair is now plotted and fed into a live least-squares regression that tries to recover the very constants (a≈150ms, b≈120ms/bit) baked into the simulator, exactly as Paul Fitts's original 1954 study and every HCI evaluation since have done from real human timing data. Watch the fitted intercept and slope converge onto the true values as taps accumulate, and watch the fit degrade — R² dropping, the scatter widening — as you crank up motor impairment or shrink targets below the WCAG 44pt/48dp minimums.