"Will I do it?": on Manifold, the crowd systematically overestimates people's chances of achieving their own goals

The analysis of 6,150 predictive markets created by users on their own future shows persistent collective optimism, which only the individual concerned ends up correcting.

More and more people are sharing their goals online and inviting others to bet on their success. Psychology teaches that observers are usually less optimistic than the person themselves, and prediction markets are supposed to aggregate the wisdom of crowds. Zihan Chen, Lei Nico Zheng and Di Zhu reconstructed all the orders of 6,150 markets on the Manifold platform in which 1,660 users asked a question about their own future, and compared them to markets for external events.

Overconfidence that starts from day one

These "self-deals" end with a 4.4-point yes vote less often than their final price would have suggested, compared with a difference of only 0.9 points for external events. Absent at the opening, this bias appears in the first twenty-four hours, even when the person does not bet: where observers bring the prices of the external markets back to the base frequency, they push those of the personal markets towards the yes. The only exception is questions about an outcome that the person does not want.

The correction comes late and comes mainly from the person concerned, whose orders provide 38% of the final price movement, compared to 9% on foreign markets. Since prices do not distinguish successes from failures, a recalibration learned from past markets improves subsequent forecasts by only 1.5%. The authors' conclusion: the crowd believes in the future that is described to them, and the information that would allow them to be deceived remains in the possession of the person himself.