Implicit Strategic Optimization: Rethinking Long-Horizon Decision-Making in Adversarial Poker Environments

TL;DR

This paper introduces Implicit Strategic Optimization (ISO), a prediction-aware framework for long-horizon adversarial decision-making, demonstrating improved long-term performance in complex strategic environments like poker and Pokemon.

Contribution

The paper proposes ISO, combining a strategic reward model with optimistic learning, providing theoretical guarantees and practical improvements over existing methods in long-horizon strategic games.

Findings

ISO achieves sublinear regret and equilibrium convergence.

Experiments show improved long-term returns in poker and Pokemon.

Framework degrades gracefully under prediction noise.

Abstract

Training large language model (LLM) agents for adversarial games is often driven by episodic objectives such as win rate. In long-horizon settings, however, payoffs are shaped by latent strategic externalities that evolve over time, so myopic optimization and variation-based regret analyses can become vacuous even when the dynamics are predictable. To solve this problem, we introduce Implicit Strategic Optimization (ISO), a prediction-aware framework in which each agent forecasts the current strategic context and uses it to update its policy online. ISO combines a Strategic Reward Model (SRM) that estimates the long-run strategic value of actions with iso-grpo, a context-conditioned optimistic learning rule. We prove sublinear contextual regret and equilibrium convergence guarantees whose dominant terms scale with the number of context mispredictions; when prediction errors are bounded, our bounds recover the static-game rates obtained when strategic externalities are known. Experiments in 6-player No-Limit Texas Hold'em and competitive Pokemon show consistent improvements in long-term return over strong LLM and RL baselines, and graceful degradation under controlled prediction noise.