What you're describing sounds like an outcome generator (which outputs the results of various possible actions) being fed into a utility function evaluator. As I understand, there is reasonable evidence for this being a functional model of part of human decision-making.