Jack Soll

Jack Soll is a decision researcher whose collaboration with Richard Larrick helps explain why groups can make better judgments by combining opinions rather than choosing a supposed expert. In The Wisdom of Crowds, their work becomes both a practical warning against “chasing the expert” and a framework for understanding how collective estimates can draw useful information from dispersed knowledge.

Research on Combining Opinions

Soll and Larrick’s work addresses a recurring mistake in decision-making: people tend not to appreciate the value of averaging multiple opinions. The book identifies the phrase “chase the expert” with their research and presents its lesson as relevant wherever an uncertain judgment must be made. Instead of treating one person’s apparent expertise as a substitute for a group’s information, decision-makers can combine judgments and reduce the influence of individual error.

The argument is not that expertise has no value, or that every group is wise. A group’s collective judgment depends on having information to aggregate, and on conditions such as diversity and independence. The book’s broader account of crowd wisdom specifies four important conditions: diversity of opinion, independence of judgment, decentralization, and aggregation. Soll and Larrick’s work supports the practical implication that selecting a single impressive forecaster can discard useful information held by others.

Collective Judgment in Practice

The book illustrates this principle through John Craven’s search for the lost submarine Scorpion. Faced with a vast search area and uncertainty about what had happened, Craven assembled people with different kinds of knowledge, including mathematicians, submarine specialists, and salvage workers. Rather than having them deliberate toward a single answer, he asked each to estimate the likelihood of possible scenarios. He then combined their judgments using Bayes’s theorem to estimate the submarine’s final location. The example makes concrete the alternative to chasing one expert: collect distinct judgments and use a method to synthesize them.

The same logic informs the book’s discussion of corporate decisions. It argues that major choices should be informed by decision markets but not handed over to them, and that important decisions should not rest on one person alone. Alfred Sloan’s group management at General Motors and Peter Drucker’s emphasis on building a management team illustrate how businesses can recognize the limits of an individual executive. The book explicitly connects this lesson to Larrick and Soll: under uncertainty, the collective judgment of a group of executives can surpass that of even a very smart executive.

The Book’s Wider Evidence

The principle appears across settings that differ sharply in scale and purpose. In the account of the 2003 U.S. election markets, the Iowa Electronic Markets’ election-eve prices were, on average, off by 1.37 percent for presidential elections, 3.43 percent for other U.S. elections, and 2.12 percent for foreign elections across forty-nine elections studied from 1988 to 2000. The market generally outperformed major national polls, despite having no more than about eight hundred traders and not resembling the electorate. The example supports the book’s claim that a group need not mirror the population it predicts in order to aggregate useful information.

The book also uses Scott Page’s computer simulations to show why diversity matters within smaller groups. In groups of ten or twenty problem-solving agents with different skills, groups that included both highly effective and less effective agents almost always outperformed groups composed only of the strongest agents. A randomly selected group could do as well as or better than a group assembled by identifying the smartest agents and putting them together. The finding clarifies the limit of expert selection: intelligence alone does not ensure that a group brings different perspectives or skills to a problem.

Other cases show why collective judgment is conditional rather than automatic. The account of the Columbia shuttle’s final flight describes how the Debris Assessment Team had reason for concern after foam struck the shuttle’s left wing, but that concern did not reach the Mission Management Team clearly. In another chapter, Long-Term Capital Management’s financial expertise failed to protect it when Russia defaulted on its debt in 1998 and investors rapidly abandoned risky assets. Together, these cases sharpen the distinction underlying Soll and Larrick’s lesson: having capable individuals is not enough if information is not effectively conveyed, combined, and tested against outcomes.

Significance

Soll’s contribution in the book is chiefly conceptual: his work with Larrick gives a name to the temptation to seek out a single expert and a reason to resist it. Across forecasting, organizational leadership, and problem-solving, the book returns to the same contrast. Individual ability may matter, but it cannot guarantee that one person possesses all relevant information; a well-composed group, using a sound means of aggregation, can bring scattered knowledge into a more reliable judgment.