"Easy" way of doing this: see what would have happened if everyone believed that you were using a model where liars don't prosper (ie random dictator), but actually use a Pareto method.