Who is the best team in the world? Regardless of the sport, all fans love to ask this question. But how do we determine it? Analysts are often wrong and use metrics based on their own biases. Here, my collaborators and I attempt to answer that question for League of Legends using a data-driven approach. This project was sponsored by Riot Games and AWS.
My collaborators and I created an award-winning (2nd place) power ranking for Global Esport teams. In this project, we were tasked with creating a method that ranks the top global League of Legends Esports teams using official Riot Games data and AWS services. Here, I worked with two colleagues from varying backgrounds (data science, full stack developer and research and data analyst (myself)) to create this power-ranker. We used the Riot data provided (3 years of competitive data) to create new features which allowed us to establish the model that would best predict how likely a team would win. We conducted exploratory data analysis to understand what features may be important – and more importantly, why they should matter given the context of professional League of Legends.