Forecasting Downloads for Gacha Games: A Comparative Study of Predictive Models & User Acquisition Strategies

Our research aims to help game developers enhance the quality of Gacha games and increase their popularity among players. We used forecasting techniques and a survey of 177 participants to predict the number of downloads and identify common issues leading to player attrition. Our findings suggest accurate forecasting and issue resolution can boost downloads and player retention rates. Linear regression was the optimal method for predicting the number of downloads for Genshin Impact V3.4, with a predicted count of about 3.85 million. In contrast, exponential smoothing was the most accurate method for Honkai Impact 3rd, with a scheduled download count of 25.96 thousand. We collected data from reputable sources, including Statista, VG Insights, research articles, and YouTube videos. Our research offers practical recommendations for game developers seeking to enhance the quality of their gacha games and their popularity among players.