Game Theory, Manipulation, and Preferential Interaction
MIT researchers wanted to analyze how both manipulation and preferential interaction would affect a social situation, so they set up experiments with “the envelope game.” In this game, there are two players and a closed envelope. Player 1 can decide to either cooperate and share the value of the envelope with Player 2 or to defect and keep it all. In addition, only Player 1 has the option to look inside the envelope. Player 2 can decide to continue playing the game or to end it.
Through tests in which each player was assigned a strategic profile to inform their decisions, it was unsurprisingly found that if Player 1 defects, Player 2 ends the game. However, it was also found that even if Player 1 cooperates, if they looked inside of the envelope before deciding, sometimes Player 2 would end the game. When Player 1 decides to open the envelope, they are telling Player 2 that they’re uncertain of their decision, they need the information inside to help decide. Even though only Player 1 knows the value of the envelope, they needed to give information to Player 2 to do so. If Player 2 threatens to end the game if Player 1 looks, they can force Player 1 into a blind decision. This is the power of manipulation, the ability to punish Player 1 for trying to gain information.
Turning now to simulations of “the envelope game” in order to collect large amounts of data, researchers found that looking was not always punished. In versions of the game where there were two variants of Player 1, Player 2 allowed Player 1 to look to determine which version of Player 1 it was and whether or not it was the favorable version. This creates the possibility of a preference for looking. Looking is essentially announcing Player 1’s uncertainty, so Player 2 would prefer Player 1 to either cooperate blind or defect after looking.
“This model is a solid contribution to our understanding of principles of behavior, cooperation, and morality, and more generally fits within a wider literature that is important and insightful which uses game theoretic models and models of learning and evolutionary processes to understand puzzling aspects of human social behavior…” – Moshe Hoffman, Harvard research scientist and lecturer