AI Strategies and Game Dynamics in Risk: Analyzing AI Personalities and Player Order Effects

This paper explores the development and evaluation of artificial intelligence (AI) strategies in the board game Risk, focusing on algorithmic approaches to decision-making and the impact of player order on game outcomes. By modeling the game as a graph optimization problem, I develop AI personalities with distinct traits in threat perception, priority, placement, and attack strategies, resulting in 16 variations. The study introduces algorithms for troop placement and fortification, aiming to achieve a center-weak border-strong distribution of armies. Analysis of gameplay data reveals trends related to player order, with certain AI personalities exhibiting advantages based on their order of play. Notably, an AI with a defensive and cautious strategy demonstrates a unique advantage when playing second. The findings suggest strategies for human players and highlight areas for future research, including subgraph analysis and dynamic game shifts, to enhance AI decision-making in Risk.