Bayesian interval and non-measureable

Good argued that probabilities are often non-measurable (@good1966a), like in a non-measurable set. However, bounds can be set (He calls them $P^{\ast }$ and $P_{\ast }$) on the probability. Thus, through his ‘black box’ theory, an org has to continually update their beliefs using an internal mental model.

This is very simular to a Bayesian posterior that is updated through iteration. The bounds $P^{\ast }$ and $P_{\ast }$ may just be an interval of higher probability on the posterior.