RatchetRun
A molecular motor does not walk on purpose. Myosin is jostled by heat all the time; burning ATP only makes a forward grab a little more likely than a backward one, so over a whole team of heads the muscle slides on average. Set the ATP level, predict how far a jostled motor drifts, then watch it — and see how noisy a single run really is. Everything runs locally; nothing about you is collected.
Toy model A simplified picture to show one idea. Not a measurement, and not a prediction. Of a molecular motor: a jostled walk with a ratchet, not a real motor protein's numbers.
This is the tool — a free prediction game. The game sets the ATP level each round; you predict, and you never place a step. A single short run often lands nowhere near the average, because each step is random — that is the point. The model and the science are written in full in the guide below (works without JavaScript). A native version, when it ships, adds offline use; it never gates the web.
ATP this round: set each round
Predict the drift after the run:
Loading the game… If it does not start, the full guide below still works.
The motor's shape appears after your first run.
How to play. Press Start. The game sets the ATP level for the round (it changes every round), you pick a drift prediction (buttons or keys 1–4), then press Enter or Run. Watch the walker jostle — some steps go backward. Press Next for another round. There is no way to lose.
Why the ATP slider matters
Each myosin head in heart muscle turns the energy of one ATP into a small stroke that biases otherwise random jostling in one direction; many heads working together pull the thin filaments and shorten the muscle. Single-letter changes in MYH7, the gene for this myosin, are a leading cause of inherited hypertrophic cardiomyopathy, and mavacamten, a drug approved in 2022, works by reducing how many of these heads engage.
Why this is a random walk, not a walk
In a muscle, a huge number of myosin heads reach toward an actin filament. Each head is jostled by heat constantly. Burning ATP does not aim a head; it just makes a forward grab-and-release a little more likely than a backward one. So each head’s motion is a biased random walk: mostly noise, tilted slightly forward. Add up a whole team of heads over many cycles and the filament slides — that is muscle contraction. No head decides where to go, and no single step is placed.
What the game does
Each round the game sets one number — the ATP level — which sets how much the odds tilt forward. You never choose it, so there is no level that makes a sure win. At 0% ATP the step is a coin flip, so the average drift is zero (but a single short run still wanders — sometimes backward!). As ATP rises, forward becomes more likely, so the average drift grows. You predict the drift, then watch one run. The gap between “what usually happens” and “what this one run did” is the lesson: direction is a statistical bias, not a plan.
Methods & about
This is a teaching model, honest about being a simplification. Muscle myosin (myosin II) is not a processive walker — a single head is attached only a small fraction of the time; real sliding is produced by a team of low-duty-ratio heads working as a ratchet, which is why this game’s walker stands for the net motion of a whole team of heads, not one head. (The classic hand-over-hand “walking” motor is myosin V.) The simulation is a transparent biased random walk: each jostle steps forward with probability 0.5 + 0.3 × (ATP ÷ 100), else backward, for 40 jostles. The structure is the real AlphaFold prediction of human MYH7 (P12883), shown for illustration. The game keeps no score history, uses no accounts, and sends nothing anywhere.
Want the deeper experience? See the Fold Commons apps. Everything we make is free and non-profit; learn more on the home page.