Anomaly Detection in Player Performances in Multiplayer Online Battle Arena Games

Authors:

Abstract
Esports are digital video games that are played professionally. In recent years there has been a growing need to improve the broadcast experience by incorporating real-time data-driven analytics. In these same games, when played by the general public, there is a growing issue of cheating. Using the popular esport and video game DOTA 2 as a case study, we present a novel application of Archetype Analysis that can be used for anomaly detection in player performance. We show how these anomalies can be utilised for both esports broadcasting and cheat detection.
Funding Information
  • Engineering and Physical Sciences Research Council (EP/M023265/1)

This publication has 22 references indexed in Scilit: