F1 Race Results Predictor

I’m building a Formula 1 race predictor 🏎️ that pulls in historical data with FastF1, cleans and normalizes results, and updates driver rankings after every Grand Prix to forecast finishing orders. It’s a work in progress, and I’m constantly adding new features and improvements! 🚀

---

Key Points

• Data ingestion: Pulls 2018–2024 race results via FastF1 API
• Data cleaning: Filters DNFs and standardizes finishing positions
• Elo engine: Initializes all drivers at 1500, updates after each GP, configurable K-factors for Sprint/Feature races
• Performance: Delivers ~10% accuracy gain over last season’s standings baseline
• Track Heatmap

---

Installation

Install the required Python packages with pip:

pip install -r requirements.txt

---

Roadmap (WIP)

• Integrate qualifying session ratings
• Build an interactive dashboard (Streamlit or React)

---

Future Work

• Refine K-factor dynamics based on track characteristics
• Incorporate weather data to adjust prediction confidence
• Add live API endpoint for real-time race forecasts
• Improve accuracy (by a lot)

---

Connect

• 🌐 Website: [https://sites.google.com/psu.edu/]
• 🔗 LinkedIn: [https://www.linkedin.com/in/meeravshah/]

---

License

This project is licensed under the MIT License.