README.md

metadata

title: Lichess Insights Analyzer
emoji: ♟️📊
colorFrom: blue
colorTo: green
sdk: streamlit
app_file: app.py
pinned: false
license: apache-2.0

Lichess Insights Analyzer ♟️📊

Dive deep into your Lichess chess game statistics! This Streamlit web application analyzes your rated Bullet, Blitz, and Rapid games directly from the Lichess API.

Key difference from the original Kaggle notebook: While the notebook analyzed pre-downloaded data for Alireza Firouzja, this web app allows you to analyze the games of any Lichess user simply by entering their username!

🚀 Try the live app here: https://huggingface.co/spaces/Andolinism/lichess-insights

✨ Features / Analysis Sections

Explore various aspects of any Lichess user's performance, including:

• Overview & General Stats: Win rate, total games, W/L/D ratio, average opponent Elo.
• Performance Over Time: Rating trend, games per year, win rate per year.
• Performance by Color: Detailed win/loss/draw percentages for White and Black.
• Time & Date Analysis: Performance based on day of the week, hour of the day (UTC), and day of the month. Includes time control category breakdown.
• ECO & Opening Analysis: Frequency and win rates for openings, presented using both standard Lichess API naming and a custom ECO code mapping (from eco_to_opening.csv).
• Opponent Analysis: Most frequent opponents and performance based on Elo difference against the field.
• Games against Titled Players: Filter and analyze performance specifically against players with official FIDE titles (GM, IM, FM, CM, WGM, WIM, WFM, WCM, NM).
• Termination Analysis: Breakdown of game endings, including a detailed analysis of wins and losses due to time forfeits, broken down by time control.

🎮 How to Use

1. Navigate to the app using the link above: https://huggingface.co/spaces/Andolinism/lichess-insights
2. Enter the Lichess Username you want to analyze in the sidebar settings.
3. Select the Time Period for the analysis.
4. Choose the Game Type (Bullet, Blitz, or Rapid).
5. Click the "Analyze Games" button.
6. Wait for the data to be fetched and processed.
7. Explore the analysis sections using the sidebar.
8. Interact with the plots (hover, use modebar buttons). Direct drag-to-zoom is disabled for better mobile scrolling.

🛠️ Technology Stack

• Language: Python
• Web Framework: Streamlit
• Data Manipulation: Pandas
• Plotting: Plotly
• API Interaction: Requests
• Hosting: Hugging Face Spaces

💾 Data Source

• Game data is fetched directly from the Lichess API based on the provided username and filters.
• Custom opening name mapping uses the eco_to_opening.csv file.

🙏 Acknowledgements

This web application is heavily inspired by and expands upon the analysis performed in the Kaggle notebook: Chess Insights: Alireza Firouzja's Games Analysis by mynameiskaveh.

📜 License

This project is licensed under the Apache 2.0 License (or the license you selected).