Abstract: The volume of research on Artificial Intelligence (AI) is a motivating factor for researchers in AI to solve a number of difficult problems in real life situations. One of such problems is game playing which has become one of the most interesting AI applications to the public. Ayo game belongs to the family of the board game called mancala and it is one of the oldest games of strategy known among the Yorubas in Nigeria. Some attempts have been made to solve the game but due to its irregular pattern as the game progresses, the exact solution had hitherto not been found. We have used refinement-based heuristic approach which instead of exploring the entire search space of Ayo, pruned part of the space that are not relevant and thus reduce the amount of computation to evolve Ayo game player with a view to suggesting move strategies. In this research we evolve Ayo game player through simulation which was tested against Awale shareware. For the purpose of evaluation we designed a paper-based questionnaire to harvest users impression about the performance of our prototype simulation through a series of game playing experiment with Awale shareware and found that the prototype simulation is computationally efficient and a better user satisfaction results were obtained.
O. Akinyemi. Ibidapo, Longe. Harrison O. Dele and O. Olugbara. Oludayo, 2016. On the Performance Evaluation of Refinement-Based Heuristics Strategy to Evolve Ayo Game. Asian Journal of Information Technology, 15: 3624-3630.