Fox vs. Hedgehog Thinking
Know many things broadly instead of one thing deeply
Philip Tetlock's landmark twenty-year study of 82,361 expert predictions revealed a fundamental divide in thinking styles. Borrowing from philosopher Isaiah Berlin, Tetlock identified 'hedgehogs' who know one big thing and 'foxes' who know many little things. Hedgehogs fashion tidy theories of how the world works through the lens of their specialty and bend every event to fit. Foxes draw from an eclectic array of traditions, accept ambiguity and contradiction, and are willing to change their minds.
The results were devastating for hedgehogs. Their predictions were roughly as accurate as random chance. Worse, they actually became less accurate as they accumulated credentials and experience, because more information gave them more material to construct convincing but wrong narratives. Foxes significantly outperformed, especially on long-term predictions. They treated their own ideas as hypotheses to be tested rather than truths to be defended.
In the IARPA forecasting tournament, Tetlock's team of foxes, drawn from the general public with no access to classified information, beat professional intelligence analysts by margins that remain classified. The key traits were active open-mindedness, intellectual curiosity, willingness to update beliefs, and the ability to aggregate perspectives from diverse sources.
- Breadth of perspective beats depth of expertise for predictions in complex, uncertain domains
- The more famous an expert, the worse their predictions tend to be, because fame rewards confident hedgehog-style pronouncements
- Active open-mindedness, the willingness to seek and genuinely consider views that challenge your own, is the key differentiator
- Good forecasters update their beliefs continuously based on new evidence, treating each belief as a hypothesis
- Aggregate perspectives from diverse sources rather than relying on a single expert or theory
- Adopt the fox mindsetApproach problems with the assumption that you are probably partly wrong. Seek out perspectives from different disciplines and viewpoints. Be willing to say 'I don't know' and change your mind when evidence warrants it.Pro tipThe best forecasters described themselves as 'genuinely curious about, well, really everything.' Cultivate wide-ranging curiosity as a strategic asset.WarningThis feels uncomfortable. Hedgehog-style confidence feels good and impresses others. Fox-style uncertainty is psychologically harder but more accurate.
- Aggregate perspectives like a dragonflyCollect viewpoints from many sources, especially from outside your domain. Tetlock described the best forecasters as 'foxes with dragonfly eyes,' synthesizing tens of thousands of different perspective lenses into a coherent picture.Pro tipSuperforecaster Ellen Cousins said of narrow experts: 'I try to take facts from them, not opinions.' Use specialists as information sources, not as oracle decision-makers.
- Practice active open-mindednessWhen you encounter information that contradicts your current belief, do not dismiss it. Seek out contrary evidence proactively. Two-thirds of people in one study would rather not even look at counterarguments. Be in the other third.Pro tipCharles Darwin made a point of copying into his notes any fact or observation that contradicted his current theory. He relentlessly attacked his own ideas until he found one that withstood the assault.
- Track and learn from your predictionsWrite down your predictions with specific probabilities and time frames. Review them systematically. When you are wrong, do not explain it away; analyze why your thinking was flawed and update your approach.Pro tipSuperforecasters 'flip-flop like crazy.' Updating beliefs rapidly in response to new evidence is not weakness; it is the hallmark of good judgment.WarningHedgehogs remain 'undefeated while losing constantly' by reframing every outcome as confirming their theory. Resist this temptation.
Biologist Paul Ehrlich predicted global catastrophe from overpopulation; economist Julian Simon predicted human ingenuity would prevail. They bet on metal prices as a proxy. Simon won, but both were wrong in important ways. Ehrlich underestimated innovation; Simon underestimated environmental degradation. Each dug deeper into his hedgehog position rather than learning from the other.
Tetlock's Good Judgment Project recruited volunteers from the general public, identified the foxiest thinkers, and organized them into teams of twelve. They were given hard geopolitical questions with no special training or classified data.
Tetlock began his study at a 1984 National Research Council meeting on Soviet relations. He observed renowned experts making perfectly contradictory predictions with equal confidence. Over the next two decades, he collected predictions from 284 experts and found that the average expert was a terrible forecaster. But he also found a subgroup, the integrators (foxes), who consistently outperformed. The IARPA tournament from 2011 to 2015 confirmed and extended these findings, showing that foxes could be identified, trained, and organized into teams that dramatically outperformed even classified intelligence analysts.