Strategy Synthesis for Multi-Agent Games of Imperfect Information

Abstract: This project focuses on strategy synthesis for multi-agent games of imperfect information. The aim ofthis project is to identify strategies for two games of interest. We utilize a program from a previousbachelor thesis to transform the games. In the transformed games, a winning strategy can be found,which then can be translated back to the original game. For the method of strategy synthesis, weproposed a naive approach to provide an intuitive understanding of strategy synthesis and also used astandard approach to showcase both results. Our results demonstrate the existence of winningstrategies within the original game using both methods. We found that the standard approach presentsthe strategies in a more formal and compact way, while our naive approach is more intuitive and easyto grasp. Moreover, we presented an informal proof, showcasing that the strategy in the transformedgame is preserved when translated, providing valuable insights into strategy synthesis. However,strategy preservation remains to be proven formally.