The Wisdom of Crowds
The book in brief. James Surowiecki built his book on cases like Francis Galton’s ox. In 1906, 787 villagers at a Plymouth fair guessed the animal’s weight, and the middle estimate landed within 1% of the truth. The book names the conditions under which a crowd is wise: individual errors are roughly independent, point in different directions, and cancel when averaged, leaving the signal.
Key ideas
- Diversity. Opinions within the crowd have to differ.
- Independence. The guesses must be independent of one another, so that no one’s mistake is correlated with anyone else’s.
- Decentralisation. People draw on their own local knowledge.
- Aggregation. A mechanism has to combine the individual guesses.
Why it matters to my work. This is one of the books that shaped how I think about decisions under uncertainty. In The Prediction Market and the Wisdom of the Few, I argue that “the wisdom of the crowd” is the wrong description of a prediction market, which weights by capital and conviction and sets its price at the margin. The crowd supplies the liquidity and the informed few supply the signal, a finding close to that of Superforecasting. An insider destroys a wisdom-of-crowds mechanism by breaking independence, whereas a market wants the insider.
