Historical Data Ingestion
We pull every qualifying, sprint, and race result from the last three seasons. Driver performance is decomposed into skill, consistency, and track-specific factors — and fed into the rating prior.

A complete F1 Fantasy toolkit built on real data — a free team builder, price-flip planner and top-500 chip board, 20,000-run race simulations, per-circuit track stats, a league analyzer that reads your rivals, and an AI coach that explains every call. Four tools are free; the rest cost less than a coffee.
Everything below runs on the same prediction engine — the five-stage pipeline further down this page.
Assemble your five drivers and two constructors against the live $100M cap, set a captain, and see projected points for every combination — with the optimizer one click away.
Open Team Builder →The exact price every driver and constructor must hit to flip. Plan transfers around the market instead of reacting to it.
Open Budget Builder →What F1 Fantasy's global top 500 do with their six chips — usage by round, the return each one paid, and how many teams are still holding.
Open Chip Stats →Overtaking difficulty, DRS zones, safety-car history, and tyre wear per circuit — plus the race-weekend weather forecast — and how each trait rewards or punishes fantasy picks.
Open Track Stats →Model your league rivals' likely moves and find the transfers and captain calls that defend or grow your rank.
Open League Analyzer →Ask which swap, captain, or chip — and get the why. Reasons over the same predictions and simulations that power every other tool.
Open AI Coach →Five stages run behind every prediction. No black box — the whole pipeline is open source and auditable.
We pull every qualifying, sprint, and race result from the last three seasons. Driver performance is decomposed into skill, consistency, and track-specific factors — and fed into the rating prior.
Each driver and each car carries a dynamic rating that updates after every session. Bayesian priors prevent overfitting to small samples — rookies and veterans are weighted differently.
We cluster circuits by result-based correlation, not just physical characteristics. That means we know which tracks actually produce similar outcomes — an edge off-the-shelf models miss completely.
For every race weekend we run twenty thousand full simulations with position-specific noise, DNF correlations, weather sampling, and tire-strategy variance. Probability distributions, not point estimates.
Our optimizer finds the budget-constrained team that maximizes expected points across every simulated outcome. It considers covariance between drivers on the same constructor — because correlated risk isn't free.
15+ years of race telemetry, lap times, weather, and driver performance feed our Bayesian ratings, Monte Carlo simulations, and AI predictions.
Ask the AI coach anything, run a simulation, ship your team. From $0.99/month; cancel any time.
Start free. Upgrade when you want it. Pays for itself the first time the simulator saves you a bad transfer.
Everything we've been asked about the engine, the business model, and what you can actually expect when you sign up.
TrueSkill than classic ELO.Free tier available. No credit card required. Setup takes thirty seconds.