A game theoretic approach to evolving and analysing mechanism design in WES

Using a game-theoretic approach, we will design a game that simulates social bugs. In this game, anything that arises should be an equilibrium. The social bugs caused by harmful behaviour that we want to prevent are undesirable equilibria, hence we propose an approach for evolving mechanisms using search-based techniques such as Evolutionary Algorithms that encourage agents to escape `bad' equilibria (i.e., equilibria that contain social bugs). Effective mechanisms are the ones that support cooperation and give incentives to agents to pick the right equilibria. The game-theoretic approach we propose allows us to analyse and explore emergent properties (such as game-theoretic properties) of the system, by running the simulation and evolving mechanisms until an equilibrium is reached. An effective set of mechanisms is one that encourages agents to converge to a `good' equilibrium, which is one that improves the experience of real-users and creates a virtual friction that is experienced by users whose behaviour contradicts Facebook's community standards.