MA.5:4.6 - Update from changed conditions and learn from outcomes
Revise the affected assumptions when new evidence or an actual decision changes them. Preserve the earlier forecast at the level needed to explain the change or learn from it; do not overwrite its premises and then claim it predicted the result.
When comparing with actual outcomes, distinguish changed external conditions, estimation error, bias and actions taken in response. Action taken in response to a forecast can prevent the adverse outcome it describes and explain a numerical difference. This does not by itself prove the forecasting model was accurate. Choose a comparison that can test the forecast claim under its stated conditions.
Choose refresh timing from the pace of relevant change and the receiving action. A rolling window can help; refreshing unchanged detail on a calendar alone can consume effort without improving the result.
A target gap can initiate design work through MA.10. Preserve the outlook under the existing arrangement while constructing the proposed alternative’s outlook from its changed offer, work and provision. The target is not a demand observation. If a design change affects both price and service behavior, revise both dependencies rather than changing only a cost assumption. MA.9 supplies a missing arrangement for participants to use and challenge the resulting expectations before commitments are made.