Stephen Bennett comments on Calibration of a thousand predictions