Jan Christian Refsgaard comments on Prediction and Calibration—Part 1