Applying artificial intelligence algorithms in MOBA games

Abstract

Multiplayer Online Battle Arena games focus mainly on struggles between two teams of
players. An increasing level of cyberbullying [1] discourages new players from the game and they often
chose a different option, that is, a match against opponents controlled by the computer. The behavior of
artificial foes can be dynamically fitted to user’s needs, in particular with regard to the difficulty of the
game. In this paper we explore different approaches to provide an intelligent behavior of bots basing on
more human-like combat predictions rather than instant, scripted behaviors.