Wisdom of Crowds

The Paradox

Francis Galton (1906) ran a contest at a country fair: 800 people guessed the weight of an ox. No single person got it exactly right. But the average of all guesses was 1,197 lbs—the EXACT weight! Galton was trying to prove democracy was foolish, but accidentally proved crowds are wise. Surowiecki (2004) showed this works for predictions, forecasts, and decisions—IF the crowd is diverse and independent.

Why Does This Work?

Error Cancellation: Each guess = signal + noise. When you average many guesses, the noise cancels out (errors are random in both directions), leaving pure signal.

Conditions for Wisdom:

• Diversity: Crowd includes experts and novices with different knowledge
• Independence: People guess without seeing others' guesses
• Decentralization: Each person draws on their own local knowledge
• Aggregation: There's a mechanism to combine guesses (averaging)

When It Fails: If people start copying each other (herding) or all have the same biased information, the crowd becomes dumb. Social proof kills crowd wisdom.