How a man and his Ox, and the Delphi Method ended up proving that groups of people harbor wisdom and knowledge.

In the decades immediately following Galton’s ox experiment, the idea quietly seeped into statistics, psychology, and social science. The experiments were repeated: jelly beans in jars. Temperatures in rooms. At first, this work framed collective accuracy as a mathematical artifact rather than a social phenomenon. There was little talk of “group minds” or emergent intelligence; instead, crowds were treated as noise that – under the right conditions could statistically self-correct.

Things began to evolve in the early 1950s, with the Delphi Method. Developed at the RAND Corporation, the Delphi Method arose from a clear failure of traditional expert committees. When military and policy experts were asked to forecast unprecedented futures – nuclear capabilities, technological breakthroughs, geopolitical shifts – their predictions often worsened in face-to-face groups. Hierarchy, reputation, and social pressure pushed participants toward premature consensus, creating confidence without accuracy.

Delphi was designed to remove these social distortions. Experts were separated from one another and asked to respond to the same questions independently and anonymously. Anonymity was central: without names, titles, or reputations attached to answers, participants could express uncertainty or unpopular views without fear of embarrassment or professional risk. The goal shifted from sounding credible to being correct.