Datasets
Standard Dataset
League of Legends Esports Player Game Data (2019-2024)
- Citation Author(s):
- Submitted by:
- Maxime De Bois
- Last updated:
- Thu, 01/16/2025 - 10:00
- DOI:
- 10.21227/5evv-jk25
- Data Format:
- License:
- Categories:
- Keywords:
Abstract
This dataset contains detailed player end-game statistics (e.g., number of kills) and in-game events (e.g., player kills) from all professional League of Legends matches held between September 15, 2019, and September 15, 2024. It encompasses a total of 37,388 matches across 392 tournaments, featuring 4,927 unique players. Matches come from all regions and tiers of play. Initially released as part of the research paper PandaSkill: Player Performance and Skill Rating in Esports – Application to League of Legends, the dataset supports a wide range of applications, including performance analysis, esports analytics, and skill rating modeling.
The dataset is made of three files: game_metadata.csv, game_player_stats, game_events. Fields game_id and player_id are shared keys between the files.
game_metadata.csv
General information about the games.
- game_id: ID of the game
- date: date in format YYYY-MM-DD HH:MM:SS.ssssss
- match_id: ID of the match (e.g., a BO5 is a match of max 5 games)
- tournament_id: ID of the tournament
- tournament_name: name of the tournament (e.g., Playoffs)
- series_id: ID of the serie
- series_name: name of the series (e.g., LCK Summer 2024)
- league_id: ID of the league
- league_name: name of the league (e.g., LCK)
Note: every game can be included in a tree structure such that: Game ⊆ Match ⊆ Tournament ⊆ Series ⊆ League
game_player_stats.csv
Player statistics at the end of a given game.
- game_id: ID of the game
- player_id: ID of the player
- player_name: name of the player
- team_id: ID of the player's team
- team_name: name of the player's team
- team_acronym: acronym of the player's team
- role: role of the player (e.g., Mid)
- win: whether the player has won the game or not
- game_length: length of the game in seconds
- champion_name: name of the Champion played by the player
- team_kills: total number of champion kills of the player's team
- tower_kills: total number of tower kills of the player's team
- inhibitor_kills: total number of inhibitor kills of the player's team (destroying an inhibitor that has respawned is counted as a kill)
- dragon_kills: total number of Drake kills of the player's team
- herald_kills: total number of Rift Herald kills of the player's team
- baron_kills: total number of Baron Nashor kills of the player's team
- player_kills: player's number of champion kills
- player_deaths: player's number of deaths
- player_assists: player's number of assists
- total_minions_killed: player's number of minions killed
- gold_earned: player's total amount of gold earned
- level: player's final level (max 18)
- total_damage_dealt: damage dealt by the player, disregarding the target
- total_damage_dealt_to_champions: player's damage dealt to Champions
- total_damage_taken: player's damage taken, disregarding the source
- wards_placed: player's number of wards placed
- largest_killing_spree: player's largest killing spree
- largest_multi_kill: player's largest multi-kill (max 5)
game_event.csv
Events occurring in a given game, often involving players.
- game_id: ID of the game
- timestamp: game timestamp in seconds
- event_type: type of the event (e.g., player_kill)
- killer_id: ID of the killer
- killed_id: ID of the killed if it exists
- assisting_player_ids: list of ID of the assisting players
- drake_type: type of the drake (e.g., infernal)
Dataset Files
- game_players_stats.csv (42.23 MB)
- game_metadata.csv (3.29 MB)
- game_events.csv (97.57 MB)