OK, you don't apply Bayesian priors to other's behavior based on your own internal knowledge. I'm not sure why you have much confidence in the analysis, then. See the second argument of chroma.