(PhysOrg.com) -- By combining techniques from game theory and artificial intelligence, computer scientists at the University of Michigan have developed a better way to find the best bidding strategy in a simulated auction modeled after commodity and financial securities markets.