Systematic Overconfidence Taxonomy
Humans exhibit systematic overconfidence across domains, with subjective confidence consistently exceeding objective accuracy in three distinct forms—overestimation of absolute performance, overplacement relative to others, and overprecision of confidence intervals—that behave differently across task difficulty levels.
Why This Is an Axiom
This represents a robust, cross-culturally replicated empirical finding that characterizes a fundamental feature of human metacognition. The systematic nature (not random error) and the taxonomic structure (distinct forms with different moderators) make this a foundational constraint on human judgment.
Key Evidence
Moore and Healy's (2008) taxonomy distinguished three forms with different behavioral signatures: overestimation is strongest on hard tasks, overplacement shows the Dunning-Kruger pattern (inversely related to competence), and overprecision (too-narrow confidence intervals) is robust across difficulty. Meta-analyses show 60-80% of drivers rate themselves above average, experts in competitive domains systematically underestimate competitors, and confidence intervals are typically calibrated for 98% coverage when people request 80% intervals. The effect sizes are large and consistent across domains (finance, medicine, forecasting, general knowledge).
Curriculum Connection
This axiom motivates the curriculum's detailed treatment of confidence calibration, comparative judgment, and interval estimation as separate skills requiring separate training. It explains why reducing overconfidence requires domain-specific practice with immediate feedback rather than general debiasing instruction. The taxonomy helps instructional designers target the specific form of overconfidence most relevant to each learning domain—overestimation in skill acquisition, overplacement in competitive contexts, overprecision in uncertainty quantification.