A data-driven approach to identify factors correlated to online games performance

Abstract

In the online gaming industry, the use of engagement features has been an increasingly used strategy to expand business results. This practice is even more important in those business models in which the company's revenue is directly proportional to the level of user engagement, like in free-to-pay and gambling games. In this context, knowing the engagement features and other success factors like technical characteristics and distribution channels that result in more profitable games is a decisive factor for online game developers. This work aims to contribute to this challenge by proposing a data-driven approach based on Machine Learning (ML) models to identify which factors are correlated with the performance of online games, in order to contribute to improving development decision-making and business revenue.