We are often more certain than the evidence really allows.
More quotesThirty seconds in painfully cold water.
Or the same thirty seconds followed by another thirty as the water becomes one degree warmer.
People chose to repeat the longer ordeal.
Experience had become objectively worse by duration and subjectively better in memory. The ending did enough repair to change the verdict.
Businesses should find this result both useful and morally hazardous. It explains why a difficult process can be remembered kindly when the final uncertainty is resolved with care. It also tempts people to leave the misery intact and add a cheerful flourish. “Great news—your complaint is now closed!” after six weeks of incompetence is not behavioural design. It is a small insult in upbeat punctuation.
Endings deserve attention because they are disproportionately remembered, especially after an intense peak. The final explanation, confirmation or reduction in effort can influence whether the whole episode feels complete. That effect is real even when the preceding minutes remain unchanged.
Reality and memory therefore need separate measures. Duration, failures and harm show what happened. Later recall shows what the person carries forward into the next choice. Optimising either alone creates nonsense: a technically smooth process nobody trusts, or a memorable recovery used to excuse preventable failure.
The warmer final thirty seconds did not remove the cold. They altered the shape of remembering it.
One degree. Twice the duration. A preferred repeat.
Human memory has never shown much respect for operational averages.