Abstract
A program for playing the three-player tarok card game is presented and the complexity of the game is analyzed. To deal with the unknown distributions of other players’ cards, the program uses a sampling method that accomplishes hierarchical clustering to select representative sets of cards from a host of randomly generated sets. A game tree is searched with the alpha-beta algorithm using several common enhancements and an equivalence transposition table, which groups the positions by strategic similarity instead of storing single positions.
Get full access to this article
View all access options for this article.