Superforecasting: The Art and Science of Prediction

Knowledge Graph: Superforecasting: The Art and Science of Prediction (Philip E. Tetlock & Dan Gardner, 2015)

Editorial spotlight: ↑ the superforecaster advantage: 30% better than intelligence analysts with classified data

Concepts

• Tetlock's superforecaster (importance 5): Top 2% of forecasters who consistently outperform both crowds and experts. Not defined by credentials but by sustained accuracy across diverse questions.. Source: (from training memory of book).
• Brier score (importance 4): Mathematical measure of forecast accuracy. Ranges 0.0 (perfect) to 2.0 (maximally wrong). Squares the difference between forecast probability and actual outcome.. Source: (from training memory of book).
• Perpetual beta (Tetlock's term) (importance 4): Treating beliefs as provisional hypotheses requiring constant updating as new evidence arrives. Core superforecaster mindset.. Source: (from training memory of book).
• Perpetual beta mindset (Tetlock synthesis) (importance 4): Combining intellectual humility, active open-mindedness, granular updating, and continuous learning. The psychological core of superforecasting.. Source: (from training memory of book).
• Granular probability judgment (importance 3): Using precise percentages (63%, not 'likely') forces clearer thinking and enables accurate scoring. Superforecasters think in percentages, not vague words.. Source: (from training memory of book).
• Dragonfly-eye perspective (importance 3): Aggregating many different viewpoints like a dragonfly's compound eye. Teams of diverse forecasters see patterns individuals miss.. Source: (from training memory of book).
• Actively open-minded thinking (AOT) (importance 3): Psychological trait measured by surveys: willingness to consider contrary evidence and update beliefs. Strong predictor of superforecaster performance.. Source: (from training memory of book).
• System 1 / System 2 (Kahneman) (importance 3): Fast intuitive thinking (System 1) vs slow deliberate reasoning (System 2). Superforecasters toggle between them strategically.. Source: (from training memory of book).
• Tip-of-your-nose perspective (importance 3): Tetlock's term for epistemic humility: recognizing how little you can see of the world from your limited vantage point.. Source: (from training memory of book).
• Foxiness (Tetlock's cognitive style) (importance 3): Intellectual humility, eclecticism, self-criticism, tolerance of ambiguity. Contrasted with hedgehog certainty. Cognitive style of superforecasters.. Source: (from training memory of book).
• Black swan events (Taleb) (importance 3): Rare, high-impact events that are retrospectively explainable but prospectively unpredictable. Challenge to any forecasting program.. Source: (from training memory of book).
• Calibration (forecast-outcome alignment) (importance 3): When you say 70%, it happens 70% of the time across many forecasts. Superforecasters are well-calibrated; overconfident experts are not.. Source: (from training memory of book).
• Distinguishing luck from skill (importance 3): Requires many forecasts over time. One correct prediction could be luck; sustained performance over 100+ forecasts indicates skill.. Source: (from training memory of book).
• Bayesian updating (importance 3): Formal method for revising probabilities given new evidence using Bayes' theorem. Superforecasters approximate this intuitively.. Source: (from training memory of book).
• Cognitive reflection (importance 2): Tendency to override intuitive answers with deliberate analysis. Measured by CRT test; correlated with forecasting accuracy.. Source: (from training memory of book).
• Planning fallacy (Kahneman-Tversky) (importance 2): Systematic tendency to underestimate time, costs, and risks of future actions while ignoring past similar cases. Classic inside-view failure.. Source: (from training memory of book).
• Hindsight bias (importance 2): Tendency to see past events as more predictable than they were. Recording probabilistic forecasts in advance prevents retrospective distortion.. Source: (from training memory of book).
• Illusion of validity (importance 2): Kahneman's term for unwarranted confidence in predictions despite low actual accuracy. Narrative coherence creates false certainty.. Source: (from training memory of book).
• Belief perseverance (importance 2): Tendency to cling to initial beliefs despite disconfirming evidence. Superforecasters resist this by treating beliefs as testable hypotheses.. Source: (from training memory of book).
• Confirmation bias (importance 2): Seeking evidence that supports existing beliefs while ignoring contradictory data. Classic cognitive trap superforecasters actively counter.. Source: (from training memory of book).
• Hedgehog (Tetlock's cognitive style) (importance 2): Viewing world through single big idea, high confidence, resistance to belief change. Makes for good TV but poor forecasts.. Source: (from training memory of book).
• McLaughlin Group Syndrome (importance 2): Tetlock's term for experts who make confident, extreme predictions that prove wrong but face no accountability. Named after political TV show.. Source: (from training memory of book).
• Affect heuristic (importance 2): Letting emotional reactions determine probability estimates. 'Terrorism feels scary' translates to 'terrorism is likely' without quantitative check.. Source: (from training memory of book).
• Groupthink (Janis) (importance 2): Consensus-seeking that suppresses dissent and critical evaluation. Destroys wisdom of crowds by eliminating independence.. Source: (from training memory of book).
• Clairvoyance fallacy (importance 2): Judging forecasts by whether they got the outcome right, ignoring whether the reasoning was sound. A 90% forecast that fails isn't necessarily wrong.. Source: (from training memory of book).
• Resolution (discriminating ability) (importance 2): Distinguishing outcomes: giving higher probabilities to events that happen, lower to events that don't. Complementary to calibration in Brier score.. Source: (from training memory of book).
• Iraq WMD intelligence failure (importance 2): 2003 invasion based on faulty certainty about weapons. Motivating example of expert overconfidence and groupthink consequences.. Source: (from training memory of book).
• Regression to the mean (importance 2): Extreme outcomes tend to be followed by more moderate ones. Forecasters who ignore this overreact to recent data.. Source: (from training memory of book).
• Availability heuristic (importance 2): Judging likelihood by ease of recall. Recent, vivid, or emotional events feel more probable than base rates warrant.. Source: (from training memory of book).
• Anchoring (adjustment heuristic) (importance 2): Initial values influence final estimates even when anchor is arbitrary. Superforecasters recognize anchors and adjust away from them deliberately.. Source: (from training memory of book).
• Scope insensitivity (affect-driven) (importance 2): Failure to scale responses to magnitude. Saving 2000 birds vs 200,000 birds elicits similar emotional response and similar probability estimates.. Source: (from training memory of book).
• Growth mindset (Dweck) (importance 2): Belief that abilities are developable through effort. Superforecasters exhibit this; see forecasting as learnable skill.. Source: (from training memory of book).
• Dunning-Kruger effect (importance 2): Unskilled people overestimate their ability; skilled people underestimate. Superforecasters resist both via calibration practice.. Source: (from training memory of book).
• Accountability-forecasting gap (importance 2): Political pundits face no consequences for wrong predictions. Incentivizes overconfidence and entertaining narratives over accuracy.. Source: (from training memory of book).
• Identifiable victim effect (importance 1): People respond more strongly to individual cases than statistics. Affects probability judgments when vivid examples are available.. Source: (from training memory of book).
• Plausible deniability (institutional) (importance 1): Vague predictions ('tensions will rise') allow retrospective claims of correctness regardless of outcome. Blocks accountability and learning.. Source: (from training memory of book).
• Radical transparency (Bridgewater model) (importance 1): Ray Dalio's firm records all meetings, tracks prediction accuracy, and makes track records visible. Tetlock cites as forecasting accountability exemplar.. Source: (from training memory of book).

Claims

• Superforecasters beat intelligence analysts (importance 5): Top 2% of GJP forecasters were 30% more accurate than intelligence analysts with access to classified information. Ordinary people with good habits outperform credentialed experts.. Source: (from training memory of book).
• Good judgment is a teachable skill (importance 5): Core thesis: forecasting isn't mystical talent but learnable through practice, feedback, and cognitive toolkit application. Anyone can improve.. Source: (from training memory of book).
• Prediction tournaments improve accuracy (importance 4): Competitive forecasting tournaments with scoring and leaderboards produce systematic accuracy improvements over individual expert judgment.. Source: (from training memory of book).
• Foxes beat hedgehogs (Tetlock 2005) (importance 4): From Expert Political Judgment: generalists (foxes) who know many things forecast better than specialists (hedgehogs) who know one big thing.. Source: (from training memory of book).
• Experts barely beat dart-throwing chimps (importance 4): Tetlock's 2005 finding: credentialed political experts forecasting geopolitical events were only marginally more accurate than random chance.. Source: (from training memory of book).
• Practice matters more than raw intelligence (importance 4): Superforecasters weren't distinguished by IQ but by deliberate practice, feedback-seeking, and cognitive style. Good judgment is a trainable skill.. Source: (from training memory of book).
• Clear, fast feedback drives learning (importance 4): Forecasting tournaments provide unambiguous outcomes on defined timelines. Most real-world predictions lack this, preventing skill development.. Source: (from training memory of book).
• Superforecaster teams beat solo forecasters (importance 3): Teams of superforecasters discussing forecasts were 15-20% more accurate than individuals working alone, even controlling for wisdom-of-crowds aggregation.. Source: (from training memory of book).
• Training raises Brier scores 10-20% (importance 3): One-hour training modules on probability, bias recognition, and Fermi estimation produced measurable, lasting accuracy improvements in GJP participants.. Source: (from training memory of book).
• Superforecasters update frequently in small increments (importance 3): Top forecasters made updates 4x more often than average, typically in 1-5 percentage point adjustments. Constant recalibration.. Source: (from training memory of book).
• Media-famous experts forecast worse (importance 3): Inverse correlation between expert media prominence and forecast accuracy. Being quotable and being accurate are different skills.. Source: (from training memory of book).
• Accountability improves forecasts (when specific) (importance 3): Forecasters held accountable for precise probability estimates improve. Vague accountability ('predict the future') makes things worse.. Source: (from training memory of book).
• Wisdom of crowds (when conditions met) (importance 3): Averaging independent forecasts improves accuracy, but only with diversity, independence, and aggregation mechanism. Based on Surowiecki's work.. Source: (from training memory of book).
• Forecasting works better for near-term questions (importance 3): Accuracy degrades with time horizon and complexity. Superforecasters excel at 6-18 month geopolitical forecasts; don't claim 30-year predictions.. Source: (from training memory of book).
• Smart aggregation beats simple averaging (importance 3): Weighting forecasters by track record and extremizing consensus forecasts both improve on naive crowd average by 10-15%.. Source: (from training memory of book).
• Superforecasters sustain accuracy across years (importance 3): Top forecasters maintained their performance advantage in Year 2, 3, and 4 of GJP. Not regression to mean; genuine skill.. Source: (from training memory of book).
• Epistemic humility predicts accuracy (importance 3): Forecasters who rate themselves as less confident tend to perform better. Confidence and accuracy are inversely correlated.. Source: (from training memory of book).
• Policy depends on prediction quality (importance 3): Better forecasting could prevent policy disasters (Iraq, financial crisis). Democratic accountability requires distinguishing good from bad predictions.. Source: (from training memory of book).
• Institutions resist rigorous scoring (importance 2): Government agencies, think tanks, and media avoid tracking prediction accuracy. Preserves plausible deniability but prevents learning.. Source: (from training memory of book).
• Tournament model is scalable (importance 2): GJP methodology works with thousands of forecasters on hundreds of questions. Could be applied across government and policy domains.. Source: (from training memory of book).

Methods

• Fermi-style decomposition (importance 4): Break hard questions into tractable sub-questions with rough estimates, then combine. Named after Enrico Fermi's 'how many piano tuners in Chicago' technique.. Source: (from training memory of book).
• Kahneman's outside view (importance 4): Start with base rates and reference classes before adjusting for case specifics. Counteracts inside-view narrative bias.. Source: (from training memory of book).
• Tetlock's Ten Commandments for Forecasters (importance 4): Codified best practices from GJP: triage, break problems down, strike balance of inside/outside views, update beliefs, recognize errors, bring diverse perspectives, synthesize, express uncertainty precisely, distinguish luck from skill, learn from mistakes.. Source: (from training memory of book).
• Probabilistic thinking (vs binary) (importance 4): Expressing uncertainty as probability distributions rather than yes/no. Enables scoring, calibration checks, and continuous improvement.. Source: (from training memory of book).
• Balance inside/outside views (Commandment #3) (importance 4): Start with base rates (outside), adjust for case specifics (inside). Most forecasters over-weight inside view; superforecasters strike balance.. Source: (from training memory of book).
• Express uncertainty numerically (Commandment #8) (importance 4): Use percentages, not vague words. 'Likely' means 55% to one person, 75% to another. Precision enables accountability.. Source: (from training memory of book).
• Inside view (the trap) (importance 3): Focusing on case-specific details and narratives while ignoring base rates. Default human mode; leads to overconfidence.. Source: (from training memory of book).
• Tetlock's extremizing algorithm (importance 3): Mathematical technique to adjust crowd probabilities: move consensus forecasts away from 50% toward 0% or 100% to correct for shared biases.. Source: (from training memory of book).
• Reference class forecasting (importance 3): Identify similar past cases (reference class), calculate base rate, adjust for case specifics. Developed by Kahneman & Tversky for planning fallacy.. Source: (from training memory of book).
• Keeping rigorous score (importance 3): Tracking forecasts with Brier scores, maintaining prediction journals, and conducting post-mortems. Essential for learning from mistakes.. Source: (from training memory of book).
• Cultivating diverse perspectives (importance 3): Reading widely across ideological spectrum, seeking out disagreement, actively recruiting contrary viewpoints. Superforecaster teams practice this systematically.. Source: (from training memory of book).
• Question decomposition (Commandment #2) (importance 3): Break big vague questions into smaller tractable sub-questions. The Fermi approach applied to geopolitics.. Source: (from training memory of book).
• Update beliefs incrementally (Commandment #4) (importance 3): Bayesian-style updating: small adjustments as new evidence arrives. Prevents both belief perseverance and whipsawing.. Source: (from training memory of book).
• Recognize your errors (Commandment #5) (importance 3): Conduct post-mortems on wrong forecasts; identify systematic biases; update your method. No defensiveness.. Source: (from training memory of book).
• Synthesize inside/outside + other views (Commandment #7) (importance 3): Don't just list perspectives; integrate them into coherent probability estimate with explicit weights.. Source: (from training memory of book).
• Learn continuously (Commandment #10) (importance 3): Treat every forecast as experiment; extract lessons; refine technique. Growth mindset applied to prediction.. Source: (from training memory of book).
• Scope-sensitivity check (importance 2): Test whether probability estimates scale properly with question scope. Avoids 'affect heuristic' where emotional intensity overrides quantitative reasoning.. Source: (from training memory of book).
• Premortem (Gary Klein) (importance 2): Imagine a future failure, then work backward to identify causes. Reduces overconfidence by forcing consideration of failure modes.. Source: (from training memory of book).
• Consider-the-opposite technique (importance 2): Deliberately generating arguments against your current forecast. Forces confrontation with contrary evidence and weakens confirmation bias.. Source: (from training memory of book).
• Red-teaming (adversarial testing) (importance 2): Assigning someone to attack your forecast, identify weak assumptions, generate contrary scenarios. Military technique adapted for forecasting.. Source: (from training memory of book).
• Structured analytic techniques (SATs) (importance 2): Formal methods used in intelligence analysis: analysis of competing hypotheses, devil's advocacy, scenario planning. Varying effectiveness.. Source: (from training memory of book).
• Analysis of Competing Hypotheses (ACH) (importance 2): Matrix method: list hypotheses as columns, evidence as rows, score consistency. Helps avoid confirmation bias by forcing evaluation of all hypotheses.. Source: (from training memory of book).
• Devil's advocate (importance 2): Formally assigning someone to argue against consensus. Less effective than genuine belief diversity; people don't take assigned roles seriously.. Source: (from training memory of book).
• Question triage (Commandment #1) (importance 2): Distinguish forecastable questions from impossible ones. Don't waste effort on inherently random or too-distant questions.. Source: (from training memory of book).
• Bring out contradictors (Commandment #6) (importance 2): Actively recruit people who disagree. Don't surround yourself with yes-men.. Source: (from training memory of book).
• Distinguish luck from skill (Commandment #9) (importance 2): Requires statistical thinking and large sample sizes. One success doesn't prove ability; sustained performance does.. Source: (from training memory of book).

Entities

• Good Judgment Project (GJP) (importance 5): IARPA-funded tournament with 20,000+ forecasters making predictions on geopolitical events 2011-2015. Created the empirical foundation for superforecasting research.. Source: (from training memory of book).
• IARPA (Intelligence Advanced Research Projects Activity) (importance 3): US intelligence community research arm that funded forecasting tournaments to improve geopolitical prediction accuracy.. Source: (from training memory of book).
• Expert Political Judgment (Tetlock 2005) (importance 3): Tetlock's prior 20-year study tracking 82,000+ predictions by 284 experts. Established that expert forecasts were barely better than dart-throwing chimps.. Source: (from training memory of book).
• Daniel Kahneman (importance 3): Nobel Prize-winning psychologist whose outside view, planning fallacy, and System 1/2 framework underpin forecasting best practices.. Source: (from training memory of book).
• Jonathan Baron (psychologist) (importance 2): University of Pennsylvania researcher who developed the AOT scale used in GJP to measure open-mindedness.. Source: (from training memory of book).
• Cognitive Reflection Test (Frederick 2005) (importance 2): Three-question test with intuitively appealing wrong answers. Bat-and-ball problem is famous example.. Source: (from training memory of book).
• Gary Klein (decision researcher) (importance 2): Developer of premortem technique and naturalistic decision-making research. His work complements Kahneman's bias focus.. Source: (from training memory of book).
• James Surowiecki (Wisdom of Crowds) (importance 2): Author of 2004 book documenting conditions under which crowd aggregation beats individual experts.. Source: (from training memory of book).
• Nassim Taleb (importance 2): Author of Black Swan; critic of forecasting who argues many events are inherently unpredictable. Tetlock engages with this critique.. Source: (from training memory of book).
• CIA (and intelligence community) (importance 2): Primary customer for forecasting research. GJP was designed to improve IC prediction accuracy after Iraq WMD and other failures.. Source: (from training memory of book).
• Good Judgment Open (public platform) (importance 2): Free public forecasting tournament launched after GJP. Allows anyone to practice and develop forecasting skills.. Source: (from training memory of book).
• Isaiah Berlin (fox-hedgehog metaphor) (importance 1): Philosopher who introduced the fox-hedgehog distinction from Greek poet Archilochus. Tetlock adapted it for forecasting research.. Source: (from training memory of book).
• Irving Janis (groupthink researcher) (importance 1): Psychologist who identified groupthink dynamics in Bay of Pigs and other policy failures.. Source: (from training memory of book).
• Rob Johnston (CIA methodology critic) (importance 1): CIA analyst who documented analytical failures and advocated for structured techniques. Collaborated with Tetlock on IC reform.. Source: (from training memory of book).
• Richards Heuer (ACH developer) (importance 1): CIA analyst who created Analysis of Competing Hypotheses method and wrote Psychology of Intelligence Analysis.. Source: (from training memory of book).
• Thomas Bayes (probability theorem) (importance 1): 18th century mathematician whose theorem provides mathematical foundation for belief updating given evidence.. Source: (from training memory of book).
• Carol Dweck (growth mindset researcher) (importance 1): Stanford psychologist who identified fixed vs growth mindset distinction. Superforecasters lean heavily toward growth.. Source: (from training memory of book).
• Ray Dalio (Bridgewater Associates) (importance 1): Hedge fund founder who implemented radical transparency and systematic belief-testing at his firm. Cited as forecasting culture model.. Source: (from training memory of book).

Relations

• Good Judgment Project (GJP) evidences Superforecasters beat intelligence analysts
• Good Judgment Project (GJP) enables Tetlock's superforecaster
• IARPA (Intelligence Advanced Research Projects Activity) motivates Good Judgment Project (GJP)
• Brier score enables Superforecasters beat intelligence analysts
• Brier score enables Keeping rigorous score
• Prediction tournaments improve accuracy supports Good Judgment Project (GJP)
• Expert Political Judgment (Tetlock 2005) evidences Foxes beat hedgehogs (Tetlock 2005)
• Expert Political Judgment (Tetlock 2005) evidences Experts barely beat dart-throwing chimps
• Foxes beat hedgehogs (Tetlock 2005) enables Foxiness (Tetlock's cognitive style)
• Experts barely beat dart-throwing chimps motivates Good Judgment Project (GJP)
• Fermi-style decomposition enables Tetlock's superforecaster
• Kahneman's outside view enables Tetlock's superforecaster
• Inside view (the trap) contradicts Kahneman's outside view
• Daniel Kahneman motivates Kahneman's outside view
• Daniel Kahneman motivates Planning fallacy (Kahneman-Tversky)
• Daniel Kahneman motivates System 1 / System 2 (Kahneman)
• Tetlock's extremizing algorithm evidences Smart aggregation beats simple averaging
• Perpetual beta (Tetlock's term) enables Tetlock's superforecaster
• Granular probability judgment enables Tetlock's superforecaster
• Superforecaster teams beat solo forecasters supports Dragonfly-eye perspective
• Tetlock's Ten Commandments for Forecasters enables Tetlock's superforecaster
• Dragonfly-eye perspective exemplifies Cultivating diverse perspectives
• Practice matters more than raw intelligence supports Good judgment is a teachable skill
• Actively open-minded thinking (AOT) enables Tetlock's superforecaster
• Jonathan Baron (psychologist) motivates Actively open-minded thinking (AOT)
• Cognitive reflection supports Tetlock's superforecaster
• Cognitive Reflection Test (Frederick 2005) enables Cognitive reflection
• Training raises Brier scores 10-20% evidences Good judgment is a teachable skill
• Reference class forecasting exemplifies Kahneman's outside view
• Planning fallacy (Kahneman-Tversky) motivates Kahneman's outside view
• System 1 / System 2 (Kahneman) supports Tetlock's superforecaster
• Premortem (Gary Klein) supports Consider-the-opposite technique
• Gary Klein (decision researcher) motivates Premortem (Gary Klein)
• Tip-of-your-nose perspective exemplifies Foxiness (Tetlock's cognitive style)
• Clear, fast feedback drives learning supports Good Judgment Project (GJP)
• Hindsight bias motivates Keeping rigorous score
• Illusion of validity supports Experts barely beat dart-throwing chimps
• Superforecasters update frequently in small increments evidences Perpetual beta (Tetlock's term)
• Belief perseverance contradicts Perpetual beta (Tetlock's term)
• Confirmation bias motivates Consider-the-opposite technique
• Foxiness (Tetlock's cognitive style) contradicts Hedgehog (Tetlock's cognitive style)
• Isaiah Berlin (fox-hedgehog metaphor) motivates Foxiness (Tetlock's cognitive style)
• Media-famous experts forecast worse supports Epistemic humility predicts accuracy
• McLaughlin Group Syndrome exemplifies Media-famous experts forecast worse
• Accountability improves forecasts (when specific) supports Keeping rigorous score
• Identifiable victim effect exemplifies Affect heuristic
• Affect heuristic motivates Scope-sensitivity check
• Cultivating diverse perspectives enables Tetlock's superforecaster
• Wisdom of crowds (when conditions met) supports Smart aggregation beats simple averaging
• James Surowiecki (Wisdom of Crowds) motivates Wisdom of crowds (when conditions met)
• Groupthink (Janis) contradicts Wisdom of crowds (when conditions met)
• Irving Janis (groupthink researcher) motivates Groupthink (Janis)
• Red-teaming (adversarial testing) supports Consider-the-opposite technique
• Black swan events (Taleb) supports Forecasting works better for near-term questions
• Nassim Taleb motivates Black swan events (Taleb)
• Clairvoyance fallacy motivates Brier score
• Probabilistic thinking (vs binary) enables Granular probability judgment
• Calibration (forecast-outcome alignment) supports Brier score
• Resolution (discriminating ability) supports Brier score
• CIA (and intelligence community) motivates Good Judgment Project (GJP)
• Iraq WMD intelligence failure motivates CIA (and intelligence community)
• Rob Johnston (CIA methodology critic) motivates Structured analytic techniques (SATs)
• Analysis of Competing Hypotheses (ACH) exemplifies Structured analytic techniques (SATs)
• Richards Heuer (ACH developer) motivates Analysis of Competing Hypotheses (ACH)
• Regression to the mean contradicts Superforecasters sustain accuracy across years
• Availability heuristic motivates Kahneman's outside view
• Anchoring (adjustment heuristic) motivates Kahneman's outside view
• Devil's advocate supports Cultivating diverse perspectives
• Scope insensitivity (affect-driven) motivates Scope-sensitivity check
• Distinguishing luck from skill exemplifies Distinguish luck from skill (Commandment #9)
• Question triage (Commandment #1) exemplifies Tetlock's Ten Commandments for Forecasters
• Question decomposition (Commandment #2) exemplifies Tetlock's Ten Commandments for Forecasters
• Balance inside/outside views (Commandment #3) exemplifies Tetlock's Ten Commandments for Forecasters
• Update beliefs incrementally (Commandment #4) exemplifies Tetlock's Ten Commandments for Forecasters
• Bayesian updating enables Update beliefs incrementally (Commandment #4)
• Thomas Bayes (probability theorem) motivates Bayesian updating
• Recognize your errors (Commandment #5) exemplifies Tetlock's Ten Commandments for Forecasters
• Bring out contradictors (Commandment #6) exemplifies Tetlock's Ten Commandments for Forecasters
• Synthesize inside/outside + other views (Commandment #7) exemplifies Tetlock's Ten Commandments for Forecasters
• Express uncertainty numerically (Commandment #8) exemplifies Tetlock's Ten Commandments for Forecasters
• Learn continuously (Commandment #10) exemplifies Tetlock's Ten Commandments for Forecasters
• Growth mindset (Dweck) supports Learn continuously (Commandment #10)
• Carol Dweck (growth mindset researcher) motivates Growth mindset (Dweck)
• Dunning-Kruger effect supports Epistemic humility predicts accuracy
• Good Judgment Open (public platform) builds-on Good Judgment Project (GJP)
• Policy depends on prediction quality motivates Good Judgment Project (GJP)
• Accountability-forecasting gap supports Policy depends on prediction quality
• Institutions resist rigorous scoring evidences Accountability-forecasting gap
• Plausible deniability (institutional) supports Institutions resist rigorous scoring
• Tournament model is scalable supports Policy depends on prediction quality
• Radical transparency (Bridgewater model) exemplifies Accountability improves forecasts (when specific)
• Ray Dalio (Bridgewater Associates) motivates Radical transparency (Bridgewater model)
• Perpetual beta mindset (Tetlock synthesis) enables Tetlock's superforecaster
• Perpetual beta mindset (Tetlock synthesis) exemplifies Perpetual beta (Tetlock's term)
• Perpetual beta mindset (Tetlock synthesis) exemplifies Actively open-minded thinking (AOT)
• Question decomposition (Commandment #2) builds-on Fermi-style decomposition
• Balance inside/outside views (Commandment #3) builds-on Kahneman's outside view
• Balance inside/outside views (Commandment #3) builds-on Inside view (the trap)
• Update beliefs incrementally (Commandment #4) evidences Superforecasters update frequently in small increments
• Express uncertainty numerically (Commandment #8) enables Granular probability judgment
• Good judgment is a teachable skill builds-on Practice matters more than raw intelligence
• Good judgment is a teachable skill builds-on Training raises Brier scores 10-20%
• Tetlock's superforecaster evidences Superforecasters beat intelligence analysts
• Foxiness (Tetlock's cognitive style) exemplifies Foxes beat hedgehogs (Tetlock 2005)
• Superforecasters sustain accuracy across years evidences Distinguishing luck from skill
• Forecasting works better for near-term questions motivates Question triage (Commandment #1)

© 2026 LOOMUS. Curation, importance scoring, edge typing, and visualization. Source paper © respective authors, quoted under fair use.