An explanation constructed after the fact to fit the outcome, with no way to have been wrong.
Gould and Lewontin's 1979 critique, borrowing Kipling's title. The test is whether the account could have accommodated the opposite outcome equally well — if so, it explains nothing.
In practice
- An account of a firm's success that names its culture as the cause would have named its culture as the cause had the firm failed.
- A retrospective that explains an outage by "insufficient testing" would have explained a clean quarter by "our testing culture".
- Further reading
- Wikipedia
- Subject
- Science & Epistemology
- Essay
- The ways evidence misleads