What kinds of challenges do networked Independent (non-communicative) Learners (IL) face in an evolutionary game? How to deal with the exploration/exploitation dilemma when the agents’ actions can cause their elimination and subsequent extinction of all agents in the network through a cascading failure effect? An approach could be to put the agents’ elimination on hold for a short period. But how does having agents apply reinforcement learning (RL) techniques compare to activating them with a certain probability? We were intrigued by an evolutionary game adaptation focused on modeling the evolution of cooperative behavior in realistic systems, as it suggests that rational agents can both survive and profit through cooperation. The adaptation consists of adding a vulnerability to each agent in the network (the lowest payoff needed for it to survive) and an elimination mechanism resulting from the agents’ vulnerability. Here, we show three groups of experiments, all to investigate how agents’ elimination unfolds while networked independent learners play the Prisoner Dilemma Game (PDG) with neighbors in multiple trials – and their interactions cause the elimination of agents, making the game and network co-evolve. We ran various settings to investigate high temptation to defect, and in future work, we will investigate defectors’ added resilience in distinct network topologies.