Matt_Simpson comments on Raising the forecasting waterline (part 1)