Abstract
In this article we investigate how three multi-player search policies, namely max(n), paranoid, and Best-Reply Search, can be embedded in the MCTS framework. The performance of these search policies is tested in four different deterministic multi-player games with perfect information by running self-play experiments. We show that MCTS with the max(n) search policy overall performs best. Furthermore, we introduce a multi-player variant of the MCTS-Solver. We propose three update rules for solving nodes in a multi-player MCTS tree. The experimental results show that the multi-player variant of the MCTS-Solver is a genuine improvement for MCTS in multi-player games.
Original language | English |
---|---|
Pages (from-to) | 3-21 |
Number of pages | 19 |
Journal | ICGA Journal |
Volume | 36 |
Issue number | 1 |
DOIs | |
Publication status | Published - Mar 2013 |