- Wenya Cai
- Hao Duong
- Xinran Yu
- Vivi Li
- Alisa Liu
- Seunghyun Park
Data Resource: https://www.kaggle.com/datasets/bobbyscience/league-of-legends-diamond-ranked-games-10-min/data
This project aims to predict the winner of the League of Legends game based on the data collected during the first 10 mimutes of approximately 10,000 ranked games. We also dived deep into exploring the causal relationship between players' actions and the result of the game.
League of Legends is a multiplayer online battle arena (MOBA) game where two teams (blue and red) compete to destroy each other's Nexus. The game takes place on Summoner's Rift, featuring three lanes (top, mid, and bot) guarded by turrets and a jungle filled with neutral monsters and objectives like Dragon and Baron Nashor. Players choose from five roles: top, jungle, mid, bot ADC (Attack Damage Carry), and support, each with unique responsibilities. Success requires teamwork, strategic map control, and coordination to outmaneuver opponents and secure victory.
The recommended strategy for the first 10 minutes for blue team to win is: defensive playstyle and to focus on earning Gold and Experience.
- For professional team: Insights for teams to strategize (team members selection, game strategies, ...).
- For sponsoring companies who is evaluating teams' performance: Insights for the companies to be able to evaluate their team performance, calculating chances of winning and to evaluate to recruiting members.
- For eSport companies who is looking for metrics to recruiting team, members: With the model providing information about team’s performance, to have a competitive sponsoring offer, or updating their sponsoring offers, since the industry is about popularity and performance in matches.
- Gather data and finetuning model for whole game/ tournament.
- Pipeline and production.
- Potential expansion to other eSport games.