f1 intelligence
./docs

Documentation

Technical reference for the F1 prediction platform

Check on GitHub
01 / overview

What is this?

F1 Intelligence is a Formula 1 race winner prediction platform. It ingests real session data from every grand prix weekend from the free OpenF1 and Jolpica APIs, scores each driver on seven features with weights fitted to every grand prix since 2010, and applies a softmax to produce win probabilities.

Sprint weekends run a separate 8-feature model tuned for the shorter format: heavier weight on grid position and car pace, no tyre degradation or long-run pace signals since there are no pit stops in a sprint.

The platform also features a detailed **Circuits Directory** mapping track footprint layouts, overtake rates, and safety car likelihood, along with **Head-to-Head Comparison Tools** for analyzing driver and constructor performance side-by-side.

What data is used?

OpenF1 + Jolpica : Lap times, sector splits, tyre data, qualifying times and weather from OpenF1; the official race classification from Jolpica
Historical data : 2018–present with full lap telemetry; 1980–2017 with race/qualifying results only
Static circuit data : Overtake rate and safety car probability seeded once per track: never changes at runtime
Nothing paid : Both sources are free and need no account (FastF1 now needs an F1TV subscription). Everything else is computed or static
02 / how it works
01
OpenF1 + Jolpica
free F1 session data
02
ETL Jobs
Python / Render
03
Neon DB
PostgreSQL
04
Hono API
Cloudflare Workers
05
Astro UI
Cloudflare Pages
After qualifying
OpenF1 supplies Q1/Q2/Q3 lap times, sector splits, and grid positions. FP2 long-run stints are also ingested and shown on the race page, though the model does not use them.
Feature computation
compute_features scores each driver on seven features: grid slot (and how hard the track is to pass on), qualifying pace, team qualifying and race pace, recent form and the gap to their team-mate.
Prediction
Each feature is weighted by a model fitted to past winners, and a softmax turns the scores into win probabilities. Only races before each one count, so the scorecard is out-of-sample.
After the race
compute_season_stats re-aggregates driver and team season totals. DNF counts, win rates, and car performance scores update for the next prediction.
Sprint weekends
Run an extra ingest + prediction cycle on Friday for the sprint session. The sprint model uses SQ qualifying times and grid position as the dominant signals.
Historical accuracy
Past predictions are stored alongside actual race results so you can see how the model performed over the season.