Probtus

Probtus is a theoretical concept used in probability theory and forecasting to capture the joint notion of a predicted probability and its associated uncertainty. It is a compact descriptor that appears in some teaching materials and theoretical discussions as a way to discuss probabilistic forecasts that carry explicit confidence information.

In a probtus framework, a forecast is represented by a two-part descriptor: a probability p in [0,1] of an event, and an uncertainty u in [0,1] reflecting the confidence or dispersion around p. The pair (p,u) can be used to assess calibration and sharpness of forecasts and can be extended to multi-category predictions by associating a vector of probabilities with a vector of uncertainties.

The term emerged in educational and theoretical contexts during the 2010s as researchers sought a simple way to talk about forecast probability together with uncertainty. It is commonly used in classroom examples, simulations, and conceptual discussions about how to evaluate and compare probabilistic predictions.

Probtus relates to established ideas such as probabilistic calibration, proper scoring rules, and uncertainty quantification. Critics note that introducing a separate uncertainty parameter can complicate interpretation and require precise definitions of how p and u interact in scoring and decision making.