The Angel Problem: Computational Simulation and Analysis of Emergent Behavior in Pursuit Games on Infinite Boards

The Angel Problem, formulated by John Horton Conway in 1982, consists of a two-player game on an infinite board in which an angel, capable of moving up to k squares per turn, attempts to escape indefinitely from a devil who blocks one square per turn. This work presents a computational simulation of the problem using Python, with power k = 2 and a board of size 40 × 40, in which the angel adopts a random movement strategy while the devil also blocks squares randomly. Over 200 simulated turns, the angel survived without being captured, consistent with theoretical results demonstrating that an angel of power 2 or higher can always escape. The results are analyzed in light of combinatorial game theory and recent mathematical literature, highlighting the relationship between the angel’s movement power, block density, and the connectivity of the state space. The simulation visually illustrates the angel’s trajectory and the spatial distribution of the blocks, revealing patterns of local confinement and emergent evasion strategies.