Every grid, every gap to pole, one probability.
GridMind turns live OpenF1 timing data into podium and points-finish probabilities, championship title odds, and what-if title scenarios — before and after qualifying.
Loading…
Projected gridHonest numbers, not vanity metrics
AUC & F1, not accuracy
Podium finishes are roughly 15% of the field, so a plain accuracy score would look artificially high. We report the metrics that actually measure a classifier on an imbalanced target.
10,000-run Monte Carlo
Every championship forecast simulates the rest of the season ten thousand times from a Plackett-Luce driver strength model.
Every model ships its own caveat
Training results carry a plain-language note about sample size, so a one-season model doesn't get oversold as more than it is.
OpenF1, live
Free, public F1 timing data from 2023 onward — no paywalled feed.
What GridMind predicts
Race Predictions
Podium and points-finish probability for every driver, from five features: grid position, gap to pole, pit stop count, average pit stop duration, and rainfall — trained per season with ML.NET's SDCA logistic regression.
Championship Forecasting
A Plackett-Luce driver strength model feeds a 10,000-run Monte Carlo season simulation, producing title probabilities for both the Drivers' and Constructors' championships.
What-If Scenarios
See exactly what a contender needs to stay alive and what the leader needs to clinch — points to close the gap, required average finish, the math behind every 'still alive' badge.
How it works
- 1
Ingest
Pull a season's sessions, results, and pit stops from OpenF1.
- 2
Engineer features
Turn raw timing data into the five signals the models train on.
- 3
Train
Fit SDCA logistic regression classifiers for podium and points finishes, per season.
- 4
Predict
Preview the next Grand Prix from recent form before qualifying, then from the real grid once it runs.