Implementation of user playstyle coaching using video processing and statistical methods in league of legends

• Published in: Multimedia tools and applications Vol. 80; no. 26-27; pp. 34189 - 34201
• Main Authors: Kim, Shinyoung, Kim, Dohyeon, Ahn, HyungGeun, Ahn, Byeongtae
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.11.2021; Springer Nature B.V
• Subjects: Big Data; Computer Communication Networks; Computer Science; Data Structures and Information Theory; Games; Image processing; Machine learning; Market value; Multimedia Information Systems; Special Purpose and Application-Based Systems; Statistical methods; Video; Deep learning; LoL game; Video processing; Coaching education; Playstyle
• ISSN: 1380-7501; 1573-7721
• DOI: 10.1007/s11042-020-09413-4

Recently, the game market is growing rapidly with the growth of e-sports. In particular, the market for League of Legends games continues to grow. Massively Online Battle Arena (MOBA) games have grown to become a massive industry, projected to reach over 140 billion USD worth of market value. Many users learn the League of Legends game, but their skills do not improve. Current analysis of the player’s manual playback through visual media such as video is the most common method. Therefore, this paper extracts data from gameplay videos and analyzes intuitive gameplay “styles” in the popular MMO game League of Legends to provide coaching-specific information. we were able to classify whether the player is cooperative and aggressive, but if additional information which we did not extract nor process like map vision were taken into account, Big Data and Machine Learning could come into play.