We present a new method, "kairosis", for aggregating probability forecasts made over a time period of a single outcome determined at the end of that period. Informed by work on Bayesian change-point detection, we begin by constructing for each time during the period a posterior probability that the forecasts before and after this time are distributed differently. The distribution of these probabilities is then integrated to give a cumulative mass function, which is then used to calculate a weighted median forecast. The effect is to construct a pool in which those forecasts are most heavily weighted which have been made since the likely most recent change in the forecasts' distribution. Kairosis outperforms standard methods, and is especially suitable for geopolitical forecasting tournaments because it is observed to be robust across disparate questions and forecaster distributions.
翻译:暂无翻译