Diversity Prediction Theorem

principle

Crowd judgment improves only when people bring sufficiently different perspectives and error patterns. When participants share a bias, aggregation preserves or compounds the mistake instead of cancelling it.

A room full of people can look like a crowd while functioning like a single mistaken mind. If everyone comes from the same background and makes the same kind of error, adding more judgments does not create wisdom; it gives the error more votes.

What matters is the shape of the errors

Aggregation works when individual mistakes point in different directions. Some estimates overshoot, others undershoot, so averaging can cancel part of the noise. Diversity matters because different backgrounds and perspectives make that error pattern more likely.

But shared experience can produce correlated errors: people overlook the same evidence or lean the same way. The average then retains the common bias. This is why the theorem sharpens the wisdom-of-the-crowds effect: the number of participants is less important than whether they are differently wrong.

Where it shows up

Many people, one background

A group drawn from the same background may supply numerous answers without supplying genuinely distinct perspectives. Its headcount increases, but the corrective power expected from aggregation does not.

Difference alone is not accuracy

The theorem does not imply that every kind of diversity improves every judgment, or that a varied crowd must be right. It identifies a boundary: aggregation loses its error-cancelling advantage when mistakes are shared. Independence, highlighted by the Ask the Audience Effect, remains part of the bargain.

Audit errors before adding voters

Before averaging a group’s forecasts, list the participants’ backgrounds, information sources, and reasoning methods. If those columns are nearly identical, recruit someone likely to make a different kind of mistake—or generate a second estimate through a different method, using the crowd within principle.

Episodes that teach this

The Wisdom of the Crowd Effect: Theory, Experiment, Examples and Explanation · explained at 11:47 2,169 views
"if all of them come from the same background and they are all making the same kind of error the wisdom of the crowds isn't going to work"