# Superforecasting [[Philip Tetlock]] and Dan Gardner's 2015 book — full title *Superforecasting: The Art and Science of Prediction* — presenting the findings of Tetlock's IARPA forecasting tournament work (Good Judgment Project, 2011–2015). The headline finding: a small minority of volunteer forecasters — about 2% of the participant pool — consistently and dramatically outperform everyone else, including intelligence-community professionals with access to classified information. The book characterizes this group, identifies the cognitive habits that produce their accuracy, and argues that those habits are *learnable* rather than innate. The book is the popular extension of Tetlock's earlier *Expert Political Judgment* (2005), which had documented that aggregate expert political prediction is barely better than chance. *Superforecasting* is the hopeful complement: yes, most experts are poorly calibrated, *and* there are identifiable practices that produce better calibration. Both findings matter. ## What the Book Argues ### The Tournament Setup IARPA — the U.S. intelligence community's research arm — funded a four-year tournament to compare prediction methods. Multiple research teams competed; Tetlock's Good Judgment Project consistently outperformed the others, including the in-house intelligence-community baseline. The book describes the tournament structure, the cohort of volunteer forecasters who participated, and the specific cognitive practices that distinguished the top performers. ### Characteristics of Superforecasters The top 2% of forecasters (the "superforecasters") share identifiable habits: - *Open-minded but rigorous* — willing to update on new evidence without abandoning structured analysis - *Probabilistic thinking* — comfortable with degrees of confidence rather than binary "will happen / won't happen" assessments - *Decomposition* — breaking complex questions into estimable sub-components rather than guessing the whole - *Outside view first* — anchoring on base rates and reference classes before considering case-specific details - *Incremental updating* — small adjustments to probability estimates as evidence accumulates, rather than dramatic flips - *Aggressive perspective-shifting* — actively seeking views that contradict their initial assessments - *Calibration tracking* — keeping records of their predictions and recalibrating against actual outcomes ### Foxes vs. Hedgehogs Building on [[Isaiah Berlin]]'s distinction (which Tetlock had used in *Expert Political Judgment*): hedgehogs know one big thing and apply it everywhere; foxes know many small things and integrate them. Foxes outperform hedgehogs at forecasting because they update on disconfirming evidence and resist overgeneralization. Most media-prominent pundits are hedgehogs; most superforecasters are foxes. The implication: the people most rewarded by attention markets are not the people most accurate at the predictions they're being asked to make. ## KB Significance **1. The diagnostic frame for political punditry.** *Superforecasting*'s findings imply that audience trust in confident political pundits is structurally misallocated. The pundits with the highest confidence are usually less accurate than calibrated forecasters; the pundits with the largest media platforms are typically hedgehogs whose forecasting record is weak. This is a *structural* critique of the punditocracy that goes deeper than partisan critique — the entire infrastructure of confident political prediction is operating on selection criteria nearly orthogonal to accuracy. **2. Calibration as cognitive defense.** The book's most useful contribution for RWKB readers may be the discipline of probabilistic thinking itself. Right-wing media trades in absolute certainty (the election was stolen, the pandemic was overblown, climate alarmism is fraud, this candidate will win in a landslide). Disciplined probabilistic thinking — "this claim has X% likelihood given current evidence" — is structurally incompatible with the absolute-confidence mode the audience has been trained into. Distributing calibration literacy is a cognitive-defense intervention. **3. Hopefulness about decision-skill.** Most decision-science findings are pessimistic about ordinary human reasoning. *Superforecasting* is the rare counterweight: a finding that *some* people do beat the dismal baseline, and that what distinguishes them is *practice* rather than innate talent. The book's quiet argument is that civic decision-making could be substantially improved by widespread training in the cognitive habits superforecasters use. This is one of the few normatively constructive findings in the field. ## See Also - [[Philip Tetlock]] — Lead author and the researcher behind the tournament - [[Daniel Kahneman]] — Tetlock's findings extend the Kahneman-Tversky biases program into expert prediction - [[Nassim Nicholas Taleb]] — Friendly disagreement: Taleb is more skeptical that any forecasting beats noise; Tetlock has the empirical evidence that some does - [[Falsifiability]] — The discipline calibration tracking operationalizes - [[Decision Science]] — The hub where this work lives