Humans and Robots (Intelligent agents) shall interact more in near future than they have ever before. Human-Robot Interaction (HRI) and Robot-Robot interaction (RRI) inevitably involves scenarios when one is unsure of the intention of the other. For successful cooperation, algorithms are needed that take in to account uncertainty in information known to each agent, and yet still achieve team tasks. An algorithm which recursively learns intentions of others by blaming all agents for possible inaccuracies in intention estimation is being implemented. The research particularly aims to bridge the gap between proposed algorithms to its practical implementation.