Judging forecasting accuracy: How human intuitions meet theoretical models