Creation of Skill Rating System in 5v5 First Person Shooter Games

Skills

TLDR

The project resulted in the creation of a 5v5 game where 100 bots participated, and the outcomes of each match were recorded. Subsequently, a Skill Rating system was employed to analyze the match results after each game. The bots were then categorized into Skill Groups based on their performance and were given opportunities for rematches within their respective skill levels. This iterative process aimed to enhance the overall gaming experience by continually refining the skill balance among the bots.

Three algorithms, including Elo, TrueSkill, and a proprietary one, adjusted the skill ratings of the bots. The iterative process simulated 50 matches, refining skill accuracy for dynamic gameplay.

The AI use the Unreal Engine Desision Trees and Blueprints to Move and perform the Shooting Actions. The Bots use the EQS system for movement using intrest points on the maps.

YouTube Video

Documentation