The Yerkes-Dodson law (1908) says performance is an inverted-U function of physiological arousal: too little arousal and an athlete is flat and unfocused, too much and they tighten up, over-attend to threat and "choke". This simulator implements it as a Gaussian peak:
Performance(A) = 100 · exp( -(A - A_opt)^2 / (2·σ^2) )
A_opt(C) = 75 − 0.45·C (optimal arousal drops as task gets harder)
σ(C) = max(8, 28 − 0.16·C) (tolerance band narrows for harder tasks)
where A is arousal (0–100) and C is task complexity (0–100). This complexity dependence is the Easterbrook cue-utilisation refinement: a simple task (sprinting, a heavy lift) tolerates — even benefits from — high arousal, while a complex task (a golf putt, a penalty kick, tactical decision-making) has a narrow, lower-arousal sweet spot because rising arousal narrows attention onto fewer cues.
- Arousal slider — moves the athlete's current point left/right along the curve; the marker ball tracks it in real time on the 3D surface.
- Task complexity slider — reshapes the whole performance surface: the peak shifts left and the curve narrows as complexity rises.
- Trial-to-trial noise — adds Gaussian jitter to each "Run Trial" sample, standing in for the real variability nerves and focus introduce between repeated attempts at the same arousal level.
- Run Trial — drops a coloured point at the current arousal/complexity showing one noisy performance sample; run several to see the scatter around the ideal curve.
- Auto-Ramp Arousal — animates arousal from 0 to 100 and back so you can watch the marker climb the curve, peak, and fall into the choking zone.
Real-world relevance: this is the model sport psychologists use to explain why pre-competition routines aim for a specific, task-appropriate arousal level rather than simply "getting pumped up" — and why the same intensity that helps a sprinter can wreck a putter.