Dark Mode
Carlsen Online Blitz and Bullet Data
Data Science and Analytics
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
This collection details all of World Champion Magnus Carlsen's online games played on the Lichess chess server between 2017 and 2021. The original Portable Game Notation (PGN) files were extracted via the Lichess API and processed using the custom Python library pgn2data, which transformed them into easily accessible CSV tabulated datasets. This processing step added supplementary descriptive features to offer greater insight into the events of each game. The data offers a consolidated perspective on Carlsen's performance, accounting for the different usernames he employed during that period. The data is available because Lichess, a free and open-source Internet chess server run by a non-profit organisation, makes its data available for extraction.
Columns
The information is divided into two primary files linked by the unique identifier game_id. The main file, "carlsen games," contains 31 top-level data points, while the "carlsen game moves" file contains the step-by-step sequence of plays.
Key fields in the "carlsen games" file include:
- game_id: A unique identifier generated during processing, not sourced from Lichess. There are 13,442 unique values.
- event: Describes the match type. 'Rated Bullet game' is the most frequent event, making up 73% of the records.
- date_played: The date the match occurred, covering a period from 14 December 2017 through 23 December 2021.
- white and black: The usernames of the players. 'DrNykterstein' is the most common username for both White and Black (36% for each).
- result: The final match outcome. Results of 1-0 (White wins) account for 49% of the games, while 0-1 (Black wins) accounts for 43%.
- white_elo: The White player's Elo rating at the time of the match. The mean rating is approximately 2,960, ranging from 968 to a maximum recorded value of 3,377.
Distribution
The information is delivered in a tabular CSV format. The primary file, carlsen_games.csv, is 3.97 MB and contains data for 13,442 unique games. The repository also includes the original PGN files and an additional file detailing ECO codes (Encyclopedia of Chess Openings) to help identify game openings. The dataset is static, with the expected update frequency stated as never.
Usage
This data product is highly suitable for detailed statistical analysis of elite chess performance, particularly in high-speed online formats. It can be used for training machine learning models or artificial intelligence agents specifically on grandmaster-level rapid and blitz play. Researchers can track ELO fluctuations over the four-year coverage period or study opening choices using the supplementary ECO codes file.
Coverage
The temporal scope extends across four years, covering all recorded activity of World Champion Magnus Carlsen on the Lichess platform between 14 December 2017 and 23 December 2021. The demographic focus is singular, covering the entire gaming history of the champion, consolidated across his various known usernames on the server.
License
CC0: Public Domain
Who Can Use It
- Data Scientists: For developing predictive models related to competitive gaming results and feature engineering based on time control and ELO data.
- Chess Researchers: To gain detailed insights into the strategic evolution and opening choices of the World Champion in fast-paced internet chess.
- Chess Enthusiasts: To filter and visualise the complete history of Magnus Carlsen’s matches on Lichess.
Dataset Name Suggestions
- Magnus Carlsen Lichess Games 2017–2021
- Carlsen Online Blitz and Bullet Data
- World Champion Lichess Game Log
Attributes
Original Data Source: Carlsen Online Blitz and Bullet Data
QUALITYLoading...