Focus Session: Games on Complex Systems

Introduction

Modern game-theoretic frameworks underpin our understanding of strategic behaviour across economics, biology, computer science, and complex engineered systems. As these systems scale and become increasingly mediated by computation, data, and learning algorithms, traditional equilibrium-based approaches alone are no longer sufficient. Concepts from complexity science now offer a powerful lens for analysing how macroscopic phenomena—such as cooperation, polarisation, self-organisation, instability, or sudden transitions—emerge from microscopic strategic interactions.

Evolutionary and learning-based game dynamics have become central to explaining adaptation in social and biological populations, as well as in artificial multi-agent environments. In parallel, advances such as mean-field game theory provide tractable methods for modelling very large populations, linking stochastic control, partial differential equations, and behavioural modelling at scale. Rapid progress in algorithmic game theory, mechanism design, and computational equilibrium analysis has broadened the field's reach to robotics, communications networks, distributed optimisation, and data-driven systems.

Machine learning has further propelled game theory to the forefront of modern research. Multi-agent reinforcement learning now routinely models competition and cooperation; generative and large language models increasingly display emergent strategic behaviours; and adversarial learning has strengthened ties to robustness, cybersecurity, and Stackelberg-type interactions. Across these domains, ideas from complexity science—phase transitions, metastability, chaos, and multi-stability—provide unifying principles for understanding when systems stabilise, fragment, or respond dramatically to small perturbations.

This focus session aims to unite researchers working at the intersection of game theory, complexity science, and computational modelling. We welcome contributions that introduce new theoretical frameworks, algorithms, or empirical findings; highlight cross-disciplinary connections; or showcase emerging applications in economics, biology, social systems, artificial intelligence, and complex engineered infrastructures. By fostering dialogue across these communities, we aim to deepen our understanding of modern strategic systems and chart new directions in the study of large-scale multi-agent interactions.

Submission and Registration

All contributors must be registered for APCNCS 2026 to participate in the focus session. Submission details, key dates, and organizer information will be announced soon.