FPL Forecaster

Viewing your team's expected score
See who is top player for each position
Zoomed in version of the main pag
Compare your players with others available for purchase
Best team to play based on the model's predictions
Main Page

Inspiration

Our teammate's love for Fantasy Premier League (FPL) and data availability inspired us to create a tool that helps managers make informed, strategic decisions effortlessly.

What it does

Our project uses XGBoost to predict Fantasy Premier League points, recommending optimal transfers, displaying the best team possible, and how well your team is expected to do via a Chrome extension for smarter FPL decisions.

How we built it

We integrated historical player performance data with real-time FPL API feeds, training an XGBoost machine learning model to accurately predict player points. A Chrome extension built in React provides intuitive visualizations and actionable recommendations, and the model and extension are connected by a Flask backend.

Challenges we ran into

Challenges included efficiently processing large datasets and ensuring seamless integration between the XGBoost model and Chrome extension for real-time predictions. We also faced issues in properly processing the outputs for display in the extension.

Accomplishments that we're proud of

We're proud of creating a robust predictive model that significantly simplifies decision-making, transforming complex statistical analyses into clear, actionable insights for FPL managers.

What we learned

We gained valuable experience in advanced machine learning techniques, API integration, and developing user-friendly extensions to effectively communicate complex predictions.

What's next for FPL Forecaster

Future developments include refining the predictive accuracy further, expanding to other sports with similarly large datasets, and experimenting with using this data to predict real life game performance.