Choosing the Right Crowd

This thread traces how the book distinguishes useful diversity and expertise from the conformity and status that can distort group judgment.

increasingBearing stren­gth in each eventdecreasing

Audience Against Expert

The Studio Audience Two Kinds of Help On Who Wants to Be a Millionaire?, contestants could ask a chosen friend or poll a studio audience. The friends answered correctly about 65 percent of the time; the audience did so about 91 percent of the time. The comparison was imperfect because the questions and the experts were not controlled. Even so, the result offers a vivid instance of a group outperforming the individual whom a contestant specifically selected for intelligence. What Aggregation Can Do The point is not that expertise is useless. Rather, a large set of varied judgments can correct errors no one participant recognizes in isolation. This example prepares the reader for the book’s more rigorous cases, while cautioning against assuming that the most impressive individual is automatically the most reliable guide.

The Narrow Expert

Expertise Has Boundaries Knowledge of a Domain Chess experts can remember realistic board positions because they recognize familiar patterns. When pieces are placed randomly, that advantage largely disappears, revealing how tightly expertise is tied to a particular domain. The book uses this distinction to question the assumption that excellence in one field transfers to broad forecasting, strategy, or judgment. A specialist may know a great deal while still being poorly equipped to predict an unfamiliar future. A Pool, Not a Prophet Studies reviewed by the author find experts inconsistent with one another and sometimes inconsistent with their own earlier assessments. Confidence alone does not reliably identify whose forecast will prove right. The practical alternative is not to discard specialists, but to combine informed judgments rather than entrust every consequential decision to one supposed seer.

Value of Difference

Diversity Adds Possibilities Beyond Similar Talent Computer simulations by Scott Page suggest that groups mixing different problem-solving approaches can outperform groups consisting only of individually high-performing agents. The less capable members may contribute tools or perspectives missing from a homogeneous team. The claim is not that ignorance helps by itself. The benefit arises when people have enough competence to contribute, but do not all approach a problem in the same way. Exploration and Learning James G. March’s distinction between exploration and exploitation explains the organizational risk: a group can become efficient at its familiar work while losing the ability to discover alternatives. Newcomers may know less on average yet still improve a team by bringing nonredundant information. Diversity therefore matters not as decoration, but as an input into the group’s capacity to learn.

The Silenced Dissent

Consensus Before Evidence A Closed Circle The Kennedy administration’s Bay of Pigs planning proceeded without serious consultation with relevant skeptics in intelligence and the State Department. The same circle both designed the operation and assessed its prospects. Concerns were muted, while assumptions about Cuban support, military strength, and American deniability went insufficiently challenged. The group’s apparent agreement made contrary views seem improbable before they were truly tested. Groupthink’s Cost Irving Janis’s analysis shows how cohesion can produce an illusion of invulnerability and a habit of rationalizing away warning signs. Discussion in such a group can reinforce certainty instead of opening inquiry. The episode demonstrates that conversation alone does not create intelligence. Without dissent and outside perspectives, deliberation can harden a bad premise into a collective plan.

One Ally Against Conformity

The Pressure to Agree A Visible Falsehood In Solomon Asch’s experiments, participants judged which line matched a model while planted confederates gave an obviously wrong answer. Many subjects followed the group at least once, though they could see the discrepancy. Their conformity often reflected a desire not to stand apart, rather than a genuine change of belief. The experiment makes social pressure visible in a task whose correct answer is unusually clear. The Power of One Dissenter When even one confederate gave the correct answer, conformity dropped sharply. A lone ally made it easier for the participant to state what he actually saw. The finding links diversity to independence: a different voice adds information, but it also gives others permission to speak honestly.

Chapter 1, opening example

Chapter 2, discussion of expertise

Chapter 2, after the Olds history

1961, during the planning of the Bay of Pigs operation

During Solomon Asch’s line-judgment studies