Learn More
Heuristic search has been very successful in abstract game domains such as Chess and Go. In video games, however, adoption has been slow due to the fact that state and move spaces are much larger, real-time constraints are harsher, and constraints on computational resources are tighter. In this paper we present a fast search method — Alpha-Beta search for(More)
Monte-Carlo Tree Search (MCTS) is a successful algorithm used in many state of the art game engines. We propose to improve a MCTS solver when a game has more than two outcomes. It is for example the case in games that can end in draw positions. In this case it improves significantly a MCTS solver to take into account bounds on the possible scores of a node(More)
Alpha-Beta pruning is one of the most powerful and fundamental MiniMax search improvements. It was designed for sequential two-player zero-sum perfect information games. In this paper we introduce an Alpha-Beta-like sound pruning method for the more general class of " stacked matrix games " that allow for simultaneous moves by both players. This is(More)
We present a first attempt at the compilation of the Game Description Language (GDL) using a bottom-up strategy. GDL is transformed into a normal form in which rules contain at most two predicates. The rules are then inverted to allow forward chaining and each predicate will turn into a function in the target language. Adding a fact to the knowledge base(More)
Many games display some kind of material symmetry. That is, some sets of game elements can be exchanged for another set of game elements, so that the resulting position will be equivalent to the original one, no matter how the elements were arranged on the board. Material symmetry is routinely used in card game engines when they normalize their internal(More)
→ no single day will suit everybody. → no single day will suit everybody. → no single day will suit everybody. → no single day will suit everybody. → no single day will suit everybody. → no single day will suit everybody. → select 2 seminar days, for even weeks and odd weeks 2 / 24 Motivation Holding weekly research seminars in a department. → no single day(More)
breakthrough is a recent race based board game usually played on size 8 × 8. We describe a method to solve 6 × 5 boards based on race patterns and an extension of Job-Level Proof Number Search (JLPNS). Race patterns is a new domain specic technique that allows early endgame detection. The patterns we use enable positions as far as 7 moves from the end to be(More)
In this paper we present a framework for testing various algorithms that deal with transpositions in Monte-Carlo Tree Search (MCTS). When using transpositions in MCTS, a Directed Acyclic Graph (DAG) is progressively developed instead of a tree. There are multiple ways to handle the exploration exploitation dilemma when dealing with transpositions. We(More)