Reciprocal Scoring

From Forecasting Wiki
This article is work in progess. The author is working on it and it is not yet ready for review.

Reciprocal Scoring[1] is an approach to forecasting topics that would be extremely difficult resolve conventionally (i.e. by observing the resolution criteria). Some common examples include unresolved scientific debates and predictions of human extinction. The approach works by framing the problem as a Keynesian Beauty Contest--that is, rather than predict the probability of the outcome directly, forecasters are directed to predict what the aggregated outcome of their forecasts will be.

See Also[edit]


  1. Karger, Ezra and Monrad, Joshua and Mellers, Barb and Tetlock, Philip, Reciprocal Scoring: A Method for Forecasting Unanswerable Questions (October 31, 2021). Available at SSRN: or

<comments />