LolWin

A League of Legends Data Analytics Application

Click here to see the code

A Data Mining Project

The idea for LolWin was born during a group project in a Data Mining course at the University of Victoria. I and colleagues, Rolin Buckoke and Cooper Mountford thought it would make an interesting project to analyze various elements of high-level League of Legends matches to see if we could find any non-obvious gameplay events or conditions that tend to result in victory. Originally we were going to make a Django web application to display our data but later decided this was out of the scope of the project and I decided to take it on as a personal project.
Click here to see the report

My contributions to this project include data collection, data preprocessing, and data visualization. For data collection, I wrote a data scraper that collects all of the games from the top 200 players in North America over a specified interval. The data scaper code can be found here . I worked on the data preprocessing with Rolin Buckoke and he laid out the groundwork to clean all of the data by both specific champions and individual games. Originally the data was output as a JSON file but we decided that it was more work to parse JSON then a CSV so I converted the preprocessor to output CSV data. While doing so I also added in a Pearson correlation generator. The code for the preprocessor can be found here .

The Web Application

The web application is written using the Django python framework and consists of two major functions. A weekly match data analysis and a champion specific data analysis. The objective is to provide a page weekly updated page consisting of an analytics report similar in style to that of the project report that we made for class. Furthermore, the champion specific data preprocessor that ended up being redundant in our project would be used to create similar weekly reports that relate specifically to individual champions in the game.

The application is in the process of being deployed on google cloud's app engine. The data scraper, as well as the preprocessors, are planned to be run as weekly cron jobs.