On the optimality of expert-adjusted forecasts by