[pa7x1]

I think, at a minimum, we should update our Bayesian priors based on past evidence.