Leveraging Sentiment Analysis of Steam Player Reviews to Predict Game Popularity

In this study, we propose a model that can predict the popularity of computer games based on the sentiment analysis of Steam player reviews. We use natural language processing (NLP) methods to classify Steam player reviews into positive, negative, and neutral sentiments. Focusing on Steam player sentiments, we utilize user reviews from various gaming titles and apply machine learning to explore the correlation between popularity indicators of the games and how Steam players feel about the game. The dataset includes thousands of reviews on diverse games from Kaggle. We examine sentiment analysis patterns across various types of game genres and their impact on the success rate of games. This study also investigates the influence of user-generated material on the evolution of game trajectories by studying the temporal shifts in sentiment patterns. To offer insights into how early feedback can serve as a leading indicator for game developers and marketers.