Inteligência artificial adaptativa para ajuste dinâmico de dificuldade em jogos digitais

Abstract

Dynamic Difficulty Adjustment (DDA) consists in an alternative to the static game balancing performed in game design. DDA is done during execution, tracking the player's performance and adjusting the game to present proper challenges to the player. This approach seems appropriate to increase the player entertainment, since it provides balanced challenges, avoiding boredom and/or frustration during gameplay. Therefore, this paper presents a mechanism to perform the dynamic difficulty adjustment during a match of a MOBA game. The idea is to dynamically change the game Artificial Intelligence (AI), adapting it to the player skills in order to make the player's experience less frustrating. We implemented three different AIs to match player behaviors: beginner, regular and expert in the game Defense of the Ancient (DotA), a modification (MOD) of the game Warcraft III. We performed a series of experiments with AI versus AI and, after comparing all results, the presented mechanism was able to keep up with the simulated player's abilities on 85% of all experiments. The remaining 15% failed to suit the simulated player's need because the adjustment did not occur on the right moment. Thereafter, user tests were performed aiming to qualitatively evaluate the mechanism. We can conclude that it behaved as expected, being able to offer a consistent opponent for beginner and intermediate players. For expert players the adaptive AI presented itself not very challenging due to the lack of a more complex strategical behavior. But in contrast to this incompatibility, all users have informed to feel challenged and motivated to fulfill the goal of the game, without feeling anxious or bored during the match.