Deterministic Versus Adaptive Pathfinding: A Comparative Analysis of Depth-First Search and Q-Learning Using Game Refinement Theory