Overwatch League (OWL) Data Analysis Project
Course: Business Analytics (BUS 462)
Executive Summary
An executive at an Overwatch League (OWL) franchise/team is convinced that their team is performing very poorly compared to the other 19 teams. After investing $30-60 million in franchise fees alone, there’s a strong desire to make some of that back in prize money (by winning tournaments) . The fundamental problem analyzed in this report is what player performance metrics can be relied on to predict the OWL players who are most likely to increase the team’s profitability through prize money earnings. Multilinear regression is used in SAS Enterprise Guide to predict the earnings of the Total (League), with the support of KNIME regression tree to identify and include critical explanatory variables in the model. Total (League) is defined as the total prize money that the player has earned from OWL tournament/stage prize money.
Based on our analysis, we concluded predicting prize money earnings based on publicly available player performance data is hard. The predictive models can only be used as reference for making decisions, as the R2 for both methods only explain ~27% of Total (League) revenue. More specifically, there are significant challenges in missing values, inconsistent data collection, inherent complexity in the domain (roles, metas), and changes in the game's fundamental structure (OW1 to OW2, 6v6 to 5v5).