Continuous Blackjack: Equilibrium, Deviation and Adaptive Strategy

TL;DR

This paper introduces continuous blackjack, analyzing its Nash equilibrium, deviations, adaptive strategies, and applying reinforcement learning to develop practical, model-free strategies for the game.

Contribution

It presents a comprehensive study of continuous blackjack, including equilibrium analysis, deviation behavior, adaptive strategies, and reinforcement learning applications.

Findings

Nash equilibrium characterized for continuous blackjack

Deviations from equilibrium analyzed and modeled

Reinforcement learning yields effective model-free strategies

Abstract

We introduce a variant of the classic poker game blackjack -- the continuous blackjack. We study the Nash Equilibrium as well as the case where players deviate from it. We then pivot to the study of a large class of adaptive strategies and obtain a model-free strategy. Finally, we apply reinforcement learning techniques to the game and address several associated engineering challenges.