Domino-Inspired Optimization (DIO): A Game-Mechanics Metaheuristic and Its Empirical Comparison With MBO and PSO

Abstract

We propose Domino-Inspired Optimization (DIO), a population-based metaheuristic derived from the mechanics of domino play: matching, toppling cascades, and gap filling. DIO models “chain reactions” by coupling local perturbations with decaying, neighbor-propagating updates over a dynamic permutation of decision variables (the “domino chain”). We formalize DIO’s operators, analyze time complexity, and benchmark it against Migrating Birds Optimization (MBO) and Particle Swarm Optimization (PSO) on five standard test functions (Sphere, Rastrigin, Rosenbrock, Ackley, Griewank). Under a common budget (10-D, population 40, 200 iterations, 3 runs), PSO led on Sphere, Rosenbrock, and Ackley; MBO led on Rastrigin and Griewank; DIO was consistently competitive—often second—while providing interpretability and strong exploitation on smooth basins. Results and critical commentary are reported. We discuss sensitivity, limitations, and research directions in hybrid domino cascades and adaptive chain topologies.