Appendix B — A Bestiary of Biases
The rogues' gallery from Part I — the specific ways your mind fakes a deduction. Know their faces.
- Abduction mistaken for deduction — treating a fertile, fallible leap to the best explanation as if it were an airtight proof. The original sin; every other bug rides on it.
- Base-rate neglect — getting mesmerised by a vivid clue and forgetting how common the thing it points to actually is. The disease-test trap; Holmes's permanent condition.
- Confirmation bias — once you have a theory, seeking only the evidence that supports it and never the evidence that would break it. (Cure: go hunting for the "no.")
- The conjunction fallacy — finding a specific, detailed story more believable than a plain one, when extra detail can only make it less probable. Why elaborate "deductions" thrill us exactly when they should worry us.
- WYSIATI ("what you see is all there is") — building total confidence from only the evidence in view, blind to the evidence that's missing. (Cure: observe the absence.)
- Patternicity / apophenia — the pattern-detector, tuned by evolution to over-fire, finding meaning and intent in noise. Genius in a rigged world; conspiracy theory in the real one.
- Anchoring — letting the first number or first idea drag all later judgments toward it, even when you know it was arbitrary.
- The Barnum effect — feeling precisely described by statements vague enough to fit almost anyone. The engine of cold reading, horoscopes, and being "read."
- Overconfidence — the near-universal gap between how sure we feel and how often we're right; strongest, cruelly, in self-styled experts. The trait Holmes maximises and the superforecaster minimises.