Review of in-game behavior of esports players with machine learning methods

In this thesis, machine learning methods were used to analyse the data of League of Legends players in different leagues during a specific season. The players' professional status was predicted using machine learning classification algorithms, and the performance metrics of the algorithms used were compared. The study aims to contribute to Esports professionals by helping them determine the level of players based on their behaviour in the game. The study is based on a comprehensive dataset regarding the game performance, strategic choices, in-game decisions, and overall skills of players in different leagues. Machine learning algorithms were used to process this extensive dataset to determine the future potential of players. The results obtained demonstrate that this approach has achieved significant success in determining the likelihood of success at the professional level. This study aims to make a valuable contribution not only to competitive games such as League of Legends but also to talent discovery and the identification of potential professional players in the field of Esports in general. As a result, the potential of machine learning methods in the field of Esports has been highlighted, and a solid foundation has been laid for future research in this area. This study aims to offer a new approach to talent management and the discovery of promising players in the Esports industry. A notable aspect of this study is the flexibility of machine learning algorithms and their ability to effectively process large data sets. Considering the differences between leagues and seasons, the adaptability provided by these algorithms is important in determining players' performance and potential. In addition, the findings obtained in this study contribute significantly to the existing knowledge on Esports talent management, while also emphasising the need for further research and development in this field. On the other hand, with the increasing application of machine learning methods in the Esports industry, issues such as ethics and privacy are also gaining importance. The correct and fair use of players' data is vital for the sustainability of research and applications in this field. Finally, this thesis aims to provide a new perspective on the process of discovering and developing professional talent in the Esports industry, thereby making a valuable contribution to the process of identifying and developing the talents of future Esports players. It is hoped that the findings of this study will serve as a foundation for the further growth and development of the Esports industry, contributing to future research and applications.