Dark Mode
Rated and Unrated Online Chess Data
Product Reviews & Feedback
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
Explore the strategies and nuances of online chess games, a classic board game revitalised by the digital community. This rich collection details a large number of matches, including essential metrics such as the specific opening used, the total number of turns played, and the final outcome. Enthusiasts and analysts can delve into this data to gain a deeper understanding of game flow and explore the central question of how to predict the best chess opening. This resource is a treasure trove for anyone interested in game theory or sports analytics.
Columns
The dataset includes 17 distinct columns detailing game information:
- game_id: Unique identifier assigned to each recorded game.
- rated: A boolean field indicating if the game was officially rated (81% of games are rated).
- turns: The recorded number of turns played, ranging from 1 to 349, with a mean of 60.5 turns.
- victory_status: The condition that ended the game. 'Resign' is the most frequent status, accounting for 56% of results.
- winner: Identifies the winning colour; White is the winner in 50% of the recorded games.
- time_increment: The time control used for the match, with '10+0' being the most common setting (38%).
- white_id / black_id: The unique usernames of the players handling the White and Black pieces, respectively.
- white_rating / black_rating: The Elo rating of the White player (mean 1.6k) and the Black player (mean 1.59k).
- moves: The precise sequence of moves recorded for the game (nearly 18,920 unique sequences).
- opening_code: The standardised code used to classify the opening (e.g., A00, C00).
- opening_moves: The number of moves defining the opening sequence (mean 4.82).
- opening_fullname: The full descriptive title of the opening, such as 'Van't Kruijs Opening'.
- opening_shortname: A commonly used shortened name for the opening, with 'Sicilian Defense' being the most frequent at 13%.
- opening_response / opening_variation: Fields detailing the subsequent responses and variations within the opening phase.
Distribution
The data is provided as a single data file,
chess_games.csv, weighing 7.69 MB. The structure consists of 17 columns across approximately 20,100 individual game records. This dataset is noted as having an expected update frequency of 'Never'.Usage
This data is perfectly suited for classification challenges. It can be used to uncover insights that help improve competitive gameplay and gain a strategic edge. Ideal use cases involve exploring the relationships between different openings and the likelihood of winning, building predictive models for game outcomes, and statistical analysis correlating player ratings with game length or victory status.
Coverage
The focus is exclusively on online chess games. The data includes detailed metadata about competitive play, including player ratings that span from 784 up to 2723. The metrics capture key game aspects such as victory conditions (Resign, Mate) and the specific time controls employed during play. The data's usability score is rated 10.00.
License
The data is governed by the Attribution 4.0 International (CC BY 4.0) license.
Who Can Use It
- Data Scientists and Analysts: For developing machine learning models aimed at classification or prediction based on strategic initial moves.
- Chess Enthusiasts: To statistically test the efficacy of different openings and gain a strategic advantage in their own games.
- Educational Researchers: To study the dynamics of competitive online performance and the statistical likelihood of success based on opening choice.
Dataset Name Suggestions
- Online Chess Opening Strategy Data
- Annotated Online Chess Game Records
- Chess Game Outcomes and Openings Analysis
- Rated and Unrated Online Chess Data
Attributes
Original Data Source: Rated and Unrated Online Chess Data
QUALITYLoading...