About the project

The data lab
of motorsport.

data-box.io is a site for motorsport, data science and artificial intelligence enthusiasts. We use real race data to show how data analysis can be applied โ€” inside and outside Formula 1.

Our Mission

๐ŸŽ
Data as a product

We turn race data into navigable insights. The same logic used by a Formula 1 team applies to products, finance, healthcare and logistics.

๐Ÿ“Š
Applied Analytics

Each section shows a type of analysis: descriptive (what happened), diagnostic (why it happened), and predictive (what will happen).

๐Ÿค–
Artificial Intelligence and Machine Learning

Prediction models based on relative scoring, recent form, circuit characteristics and safety car probability.

As if you were Head of Data at a Formula 1 team

Formula 1 telemetry is one of the richest real-time data use cases in the world. A 2026 car has around 300 sensors generating data at up to 1,200 measurements per second โ€” tyre temperature, pressure, battery usage, steering angle and G-forces.

As Head of Data, you would use this information to: optimise pit stop strategy in real time, predict tyre degradation per lap, model battery usage per sector, detect engine issues before they happen, and compare driver performance with the ideal lap model.

This site applies equivalent principles with public data โ€” showing that the same methods work for any company that wants to extract value from its operational data.

Data Sources

F1 Official
FIA Results
Wikipedia
Historical data
PlanetF1
Race reports
The Race
Technical Analysis
GPFans
Standings
RacingNews365
Stats and standings
RaceFans
Lap charts
Total Motorsport
Full classification

Techniques Used

๐Ÿ“Š
Descriptive Analysis

Results, final positions, gaps, lap times โ€” what happened. Data verified by multiple sources.

๐Ÿ”Ž
Diagnostic Analysis

Tyre strategy, outliers, position changes โ€” why it happened. Root causes: VSC timing, ERS drain, strategic errors.

โšก
Predictive Analysis

ELO + recent form + circuit + SC probability โ€” what will happen. Educational model based on public data.