A steady rule fits van running costs over five years, but for year fifty it predicts a figure nobody would trust. What is this failure called?

1. The rule was built from five years of readings, so it describes that stretch well. 2. Year fifty lies far outside that stretch, where none of the readings can vouch for the rule. 3. The point at which a rule stops describing the situation is called model breakdown. 4. Interpolation would mean predicting inside the five years of readings, which is not what happened. 5. Regression names the method used to fit the line, not the failure that follows. 6. Correlation measures how closely the readings sit to the line, again not the failure itself. _Source: OpenStax College Algebra (CC BY 4.0), Ch 4 "Linear Functions", section 4.3 Fitting Linear Models to Data_