This paper presents a proposition of a method for adjusting game difficulty to the current level of player's skills in one-on-one games. The method is based on recognition of human player's actions and recording of those actions in the form of behavioral trees. Such trees are later used to drive behaviors of computer-controlled opponents so that human player has beat hit own strategy and improve on it, to win subsequent games. The basic concept of the method is presented on the simple example of Tic-tac-toe. The paper discusses required enhancements and modifications needed for the use of this method in more refined games. Finally, additional aspects of modern games related to the application of presented method, e.g. continuity of game universe (in contrast to discrete board games) and vague and hard to recognize player's actions, are brought into consideration.
Authors
Additional information
- Category
- Aktywność konferencyjna
- Type
- publikacja w wydawnictwie zbiorowym recenzowanym (także w materiałach konferencyjnych)
- Language
- angielski
- Publication year
- 2013
