What's described here is more of a framework than an exact forecasting method. DeMesquita's methods would fit quite well, actually. > You can reliably bet on t

What's described here is more of a framework than an exact forecasting method. DeMesquita's methods would fit quite well, actually. > You can reliably bet on the effect of the upset being inversely proportional to the predicted probabilities. Broadly, yes, if you think of it in an information-theoretic sense. Rare signals carry more information than common ones.