Automated Planning in Repeated Adversarial Games
This work addresses the challenge of automated planning in multi-agent adversarial environments, offering a novel approach for agents to collaborate implicitly, though it appears incremental as it builds on existing RL methods like R-max.
The paper tackled the problem of achieving above-Nash equilibrium performance in repeated adversarial games by introducing TeamUP, a model-based RL algorithm that learns to form tacit collaborations with heterogeneous adversaries, and demonstrated its effectiveness by outperforming the winning strategy in the Lemonade Stand Game Tournament.
Game theory's prescriptive power typically relies on full rationality and/or self-play interactions. In contrast, this work sets aside these fundamental premises and focuses instead on heterogeneous autonomous interactions between two or more agents. Specifically, we introduce a new and concise representation for repeated adversarial (constant-sum) games that highlight the necessary features that enable an automated planing agent to reason about how to score above the game's Nash equilibrium, when facing heterogeneous adversaries. To this end, we present TeamUP, a model-based RL algorithm designed for learning and planning such an abstraction. In essence, it is somewhat similar to R-max with a cleverly engineered reward shaping that treats exploration as an adversarial optimization problem. In practice, it attempts to find an ally with which to tacitly collude (in more than two-player games) and then collaborates on a joint plan of actions that can consistently score a high utility in adversarial repeated games. We use the inaugural Lemonade Stand Game Tournament to demonstrate the effectiveness of our approach, and find that TeamUP is the best performing agent, demoting the Tournament's actual winning strategy into second place. In our experimental analysis, we show hat our strategy successfully and consistently builds collaborations with many different heterogeneous (and sometimes very sophisticated) adversaries.