Does a Bad Outcome Prove the Decision Was Wrong?

A decision is made before its outcome is known, so its reasonableness cannot be inferred from the result alone. Suppose the forecast gives only a low chance of rain and you arrange an outdoor event, only to be hit by a storm. The failed event is a real loss and deserves review, but the occurrence of a low-probability event does not automatically show that you ignored probability.

Normative decision theory treats choice as reasoning under uncertainty. Expected utility theory recommends comparing the value and probability of possible outcomes, not expecting the chosen option to deliver the best actual result every time. The model is contested: probabilities may be unreliable, and values cannot always be placed on one scale. Still, it exposes a crucial distinction—the quality of a decision concerns reasons available beforehand, while the goodness of an outcome is a fact known afterwards.

Bernard Williams and Thomas Nagel's discussion of “moral luck” also shows why outcomes are hard to exclude from real judgement. Whether the same negligence happens to cause harm can alter our responses, the repair owed and even legal consequences. That does not mean the result alone can rewrite the reasons available at the time.

My view is that a bad outcome should prompt two questions, not one verdict. First ask whether evidence was missed, probabilities misjudged or costs treated too lightly. Then ask what responsibility and repair are now required. If the earlier reasoning had no clear defect, regret can still be fitting without being disguised as “I should have known all along”.

https://plato.stanford.edu/entries/decision-theory/
https://plato.stanford.edu/entries/moral-luck/