On the Convergence of Bounded Agents