About the project
Finding patterns in player engagement
For my final-year Computer Science dissertation at Oxford Brookes University, I'm investigating whether machine learning can predict player engagement using data about games on Steam.
I'm interested in the relationship between things like player counts, reviews and game characteristics, and how those could help us understand engagement.
What I'm working on
I'm currently exploring datasets, checking data quality and considering suitable measures of engagement. The plan is to prepare the data, compare predictive approaches and evaluate their limitations, including the risk of data leakage.
Where things stand
This is still a work in progress, so I'm sharing what I'm exploring rather than presenting conclusions I haven't reached yet. I'll update this page as the research develops.