Supervised Statistical Machine Learning to Predict LoL Game Results

The task involved aggregating multiple CSV files, cleaning and processing the data to develop custom metrics and applying binomial linear regression machine learning algorithm to predict match outcomes with 92% accuracy in the game of League of Legends, a complex esports title with over 60,000 professional games and 300,000 champion instances. The process involved defining independent variables such as champion and player strength on a given patch, drafting strategies, and opponent analysis. The resulting model can be used to make informed decisions in the esports betting industry, with the potential to generate significant returns.