LOADING GRID…
Thousands of complete race simulations per Grand Prix — qualifying, tyres, safety cars, retirements — counted into win, podium and points probabilities.
Most F1 race predictor tools rank drivers by a rating and call it a probability. This one races. Every time you press predict it runs the full weekend model over and over: a three-part qualifying session sets a grid, then a complete Grand Prix plays out with fuel burn, tyre degradation, the two-compound rule, pit stops, safety cars, dirty air, DRS, mechanical failures and weather. Win probability is simply the share of those races a driver won.
That is why the numbers move so much between circuits. Pole is worth a great deal at Monaco, where almost nobody passes, and much less at Spa. A powerful engine gains at Monza, Baku and Las Vegas; downforce gains at Hungary and Singapore. A driver with strong tyre management gains wherever degradation is high, like Qatar and Bahrain.
The interesting question is usually not who wins here, it is who this circuit suits. Run any Grand Prix, then pick a second one to compare it against: the same twenty-two cars run the same number of simulations under the same conditions, and the table shows each driver's win and podium chance in both places along with the swing between them. Monza against Monaco moves the high-downforce cars several points of win probability and pushes the power-strong ones the other way, which is the model saying the same thing a paddock would.
This is a model. Car and driver ratings here are our own estimates rather than official data, and no simulation knows about an upgrade that landed last Friday or a grid penalty announced this morning. Read the output as what the model believes, and use it to compare circuits and scenarios rather than as a forecast of a real race result. Nothing here is betting advice.
It does not use a formula or a lookup table. It runs the full race simulation thousands of times over — a real qualifying session, then a complete race with tyre strategy, pit stops, safety cars, retirements and weather — and counts how often each driver wins, reaches the podium and scores. Run 2,000 races and the share of them a driver wins is their win probability.
Yes. It is free, runs in your browser, needs no sign-up, and nothing is sent to a server.
It is a model, not a forecast. Car and driver ratings are our own estimates, so treat the output as what the simulation believes rather than a prediction of a real result. What it is genuinely good at is showing how much a circuit changes the picture: the same grid produces very different odds at Monza and at Monaco.
Each circuit is weighted for engine power, downforce, tyre wear and overtaking difficulty. A powerful car gains at Monza, Baku and Las Vegas; a high-downforce car gains at Monaco, Hungary and Singapore. Overtaking difficulty matters just as much: at Monaco a driver who qualifies badly can almost never recover, so pole is worth far more there than at Spa. After any run you can compare the result against a second circuit, which holds the cars and the conditions fixed and shows you the swing per driver, so the circuit is the only thing that changed.
Yes. Force wet conditions and wet-weather ratings dominate, the field spreads out, mistakes go up and the odds shift substantially toward the best wet drivers.
500 gives a rough shape in under a second. 2,000 is a good balance. 5,000 settles the numbers down to about half a percentage point of noise but takes a few seconds.