About MyPitwall

How the model works and how to read the numbers

How It Works

MyPitwall uses Monte Carlo simulation to predict Formula 1 race outcomes. For each race, the model runs thousands of simulated races — each with randomized variables like qualifying pace, race incidents, pit stop timing, and weather — then aggregates the results into probabilities.

When you see “Win: 34%” next to a driver, it means that driver won in 34% of the simulated races. This isn’t a confidence level — it’s the model’s estimate of how often that outcome would happen if the race were run many times.

Data Sources

The simulation model is built on historical performance data from the following open-data projects:

Jolpica F1 API — historical race results, qualifying times, pit stops, and season schedules (Apache 2.0)
TracingInsights — enriched lap data including tyre compounds, stint information, and sector times (Apache 2.0)
FastF1 — session telemetry and timing data used to fill gaps and import pre-season testing data (MIT)

These sources feed into the following model parameters:

Driver skills — qualifying pace, race pace, racecraft, tyre management, wet skill, and consistency. Calibrated from historical performance data and updated after each race weekend.
Team performance — car pace relative to the field, pit stop speed and reliability, strategy quality. Updated as new data becomes available throughout the season.
Circuit characteristics — overtaking difficulty, tyre degradation, safety car probability, pit loss time, and lap count for each circuit.

Reading the Numbers

Win / Pole probability

How often the driver achieved that outcome across all simulations. Higher is better. These are not betting odds.

Move Up

The probability that a driver or team overtakes the entry directly above them in the final championship standings.

Form dots

Colored circles showing recent performance relative to the model’s prediction. Green = exceeded expectations. Grey = as predicted. Red = underperformed. Dark grey = DNF.

Skill profile (radar chart)

Each axis shows how the driver ranks across the grid for that skill dimension. Values are normalized to 0–100 where 100 is best on the grid.

Prediction deltas (▲▼)

How far the actual result differed from the model’s prediction. ▲ means the driver beat the prediction (finished higher than expected). ▼ means they finished lower than expected.

About the Project

MyPitwall is an independent, non-commercial project exploring probabilistic sports prediction. It is not affiliated with, endorsed by, or connected to Formula 1, the FIA, Formula One Licensing B.V., Formula One World Championship Limited, or any F1 team or constructor.

Formula 1, F1, and related marks are trademarks of Formula One Licensing B.V. All team names and logos are the property of their respective owners.

Built with Python (FastAPI) for the simulation backend and Next.js for the frontend. The model is refreshed after each race weekend with updated parameters.

Made by Adriel Frederick.