Chapter 4 of 11All chapters
Chapter 4 of 11
Common fallacies
Named mistakes are easier to spot.
The frequent ones
Affirming the consequent reads a rule backwards. Denying the antecedent assumes that no P means no Q. Both feel natural and both are wrong.
- Correlation is not causation, and a third cause explains many pairs.
- Survivorship bias looks only at what made it through.
Base rates
A test that is 99 percent accurate for a condition that one person in ten thousand has still produces mostly false positives. Ignoring how rare something is beforehand is the most expensive error in this list.