Game Theory/ Iterated Prisoner’s Dilemma Game/Agents with limited memory

In classical Iterated Prisoner’s Dilemma Game, a group of agents play against each other many times. With introduction of memory, they can remember opponent’s previous moves and based on that they can change their strategy (see. https://arxiv.org/pdf/1605.08682).

In a recent work (https://arxiv.org/pdf/1911.05282), we let agents to get recommendations from their neighborhood. It was interesting that in some cases getting recommendation is not the best strategy. In this cmpe492 project we would like to further investigate this unexpected behavior.