Modeling Real Life Situations as Multi-Agent Games with Imperfect Information Against Nature

Sammanfattning: Our research explores Multiplayer Games with Imperfect Information, short for ”MAGIIAN”. We specifically focus on two aspects: firstly, identifying real-world problems that can be appropriately modeled as MAGIIANs, and secondly, synthesizing and interpreting strategies for these models using existing tools and algorithms. Potential problems that were identified include emergency situations, shutdown protocols, and pick-up situations. In strategy synthesis, we utilized a tool developed in a previous study, aiming to formulate a strategy in a form easily understandable to humans. All the identified scenarios could be modeled as MAGIIANs, and the tool successfully produced a winning strategy for two out of the three models. The study did not identify any major patterns determining which real-world problems could be modeled as MAGIIAN games. However, it introduced techniques such as the ”passive” action and adapted the concept of simultaneity to the situation, both of which proved beneficial for the modeling of MAGIIANs. Additionally, the study discovered the potential of ”limit tests” to expand the field’s application beyond merely strategy synthesis and modelling.