Thanks for the explanation, Roger. I think it will be valuable for those making
and evaluating forecasts and predictions to look this procedure over before
doing so. I hope Canie can put this somewhere that it is easy to find - hint, hint.