Two statisticians - professors of Bayesian statistics - write in the New Yorker.
...the Bayesian approach is much less helpful when there is no consensus about what the prior probabilities should be.
That will be a bust for the climatologists then.