MCTS Based Agents for Multistage Single-Player Card Game

Abstract

The article presents the use of Monte Carlo Tree Search algorithms for the card game Lord of the Rings. The main challenge was the complexity of the game mechanics, in which each round consists of 5 decision stages and 2 random stages. To test various decision-making algorithms, a game simulator has been implemented. The research covered an agent based on expert rules, using flat Monte-Carlo search, as well as complete MCTS-UCB. Moreover different playout strategies has been compared. As a result of experiments, an optimal (assuming a limited time) combination of algorithms were formulated. The developed MCTS based method have demonstrated a advantage over agent with expert knowledge.

Optimisation of MCTS Player for The Lord of the Rings: The Card Game

“…Comparing with our previous results [11], the use of expert rules has allowed to achieve a significant increase in the effectiveness of the method at a high level of difficulty. The extension of the algorithm is based on the limitation of analyzed actions.…”

“…This allowed to perform a new analysis for games with high complexity level. We have demonstrated that the relative effectiveness of the mixed strategy proposed in [11] rises as the difficulty of the problem increases.…”

“…This article is directly based on the authors' conference work [11]. The algorithm has been extended by adding more expert knowledge into the standard MCTS implementation.…”

Bulletin of the Polish Academy of Sciences: Technical Sciences

“…This allowed to perform a new analysis for games with high complexity level. We have demonstrated that the relative effectiveness of the mixed strategy proposed in [16] rises as the difficulty of the problem increases.…”

“…Several precautions had been taken to reduce number of allowed moves. Using Action filtering and Obliged Actions are examples of successful domain-specific knowledge incorporation.This article is directly based on authors conference work [16]. The algorithm has been extended by adding more expert knowledge into the standard MCTS implementation.…”