If a forecast is wrong, should we discard the tool?
Consider how to revise assumptions and verification methods when outcomes differ from expectations.
ASSETCASTING commentary
An interpretation of the fictional story, not a statement by the original video creator.When reviewing a forecast, separately record the inputs available then, the assumed conditions, and the conditions that actually changed. Check whether the question exceeded the tool's intended scope or omitted contrary evidence. Continually changing the conditions after failure can make an explanation fit every outcome and become difficult to test. Preserve failed cases so that later cases can be compared using the same criteria.