This study presents a novel approach for predicting the outcome of the League of Legends (LoL) game using topological data analysis (TDA), specifically Vietoris-Rips persistent homology. Unlike traditional methods for analyzing time-series data, such as statistical and machine learning techniques, this method can better comprehend the intricate structure within the data. TDA is applied to the first 10 min of the LoL gameplay data, focusing on damage, experience, and resource metrics, thereby revealing the deep structural patterns that influence game outcomes. This approach enables a nuanced analysis beyond the capabilities of traditional methods, capturing non-linear interactions between players and teams. This study offers a new perspective on time-series data analysis and predictive modeling in digital games, highlighting the potential of TDA in understanding complex systems. Although the study focuses on LoL, the promising results suggest that TDA can be applied to the analysis of different types of complex data, providing valuable insights into dynamic and strategic environments.