Esports Prediction

Authors

• Benjamin Lofo Follo
• Arman Jahanpour
• Jialin Yang

Introduction

Purpose

This repository holds the code related to the study compiled in the 2021 report Predictive Analysis in League of Legends e-Sports Matches, written by the same authors towards the requirements of the COMP 432 Machine Learning class, given to students at Concordia University Montreal's Fall 2021 semester.

Deliverable Contents

The deliverable collection contains the following documents:

Folders

• data/: Resources and files stored from the scripts for data collection
• data/json: Example Endpoint Responses for the Riot API that are analyzed by extract_game_data.py
• data/processed: Final CSV training data set
• data/raw: Unprocessed files created by extract_game_data.py
• data/csv: List of match ids generated by fetch_game_ids.py
• src/: Executable Python files
• src/models/: Joblib files for the models that have been trained by our ipynb file
• src/parsers/: Parsers used within extract_game_data.py to prepare digestable tuple instances
• src/services/: API scripts to communicate with third-party data services

Files

• src/fetch_game_ids.py: Connects to the Riot API to get all game ids from the available tiers and divisions.
• src/extract_game_data.py: Scans all .csv files inside the output folder, and calls the Riot API to fetch all information related to each match ids, which includes overall game statistics and game timeline.
• src/processing.ipynb: Interactive Jupyter Notebook used to Preprocess the data collected inside the data folder.
• src/training.ipynb: Interactive Jupyter Notebook used to train models against the preprocessed data inside the data folder.

Requirements

The following packages are required to ensure all scripts run properly

• sklearn: Provides k-Means library for clustering and vectorizing text
• os: Read the files in the folder specified

A Riot API key is also needed for the API Client to work, which you can get here: https://developer.riotgames.com/

Instructions

1. Fetch the game ids with the following command

python src/fetch_game_ids.py

2. Extract the game data using the following command (NOTE: this step may take very long (ETA:20 hours)

python src/extract_game_data.py

3. Run the processing.ipynb to pre-process the data prior to analysis
4. Use the training.ipynb file to train the machine learning models

Thank you

We have used these sources for bringing the project to life:

AlphaStar: Grandmaster level in StarCraft II using multi-agent reinforcement learning (For formatting layout of our report) https://deepmind.com/blog/article/AlphaStar-Grandmaster-level-in-StarCraft-II-using-multi-agent-reinforcement-learning

Riot Developer - Collecting data (Best practices on the Riot API) https://riot-api-libraries.readthedocs.io/en/latest/collectingdata.html