A Family of Turn Based Strategy Games with Moose

The game that is the focus for this study is a simple turn-based strategy game used for a competition in an AI class at Innopolis University for two years. In this study we continue to investigate the game with evolved game-playing agents. The space of the moose competition games is expanded and the hypothesis that changing the resource profiles of the game will lead to the evolution of different types of agents is tested. The fields that the moose forage in may have enhanced rate of plant growth or enhanced richness of plant biomass. Increasing these parameters is found to decrease conflict between moose, after a minimum level is passed, and enhancing richness is found to modify moose behavior in intuitive ways. In some parts of the resource model space, the agents exhibit negative density dependent selection that increases the rate of confrontation. This study demonstrates that the family of moose games provides a rich palette of different games for testing students and agent AIs.