I have trouble with the V in P. By that I mean the judgmental weighting of whatever values one might be dealing with. If these were written on the backside of the Decalogue, the golden sheaths of President Smith or by normal, normative, consensual or singular interpretation, I’d still be sceptical of the prescribed values in the Bayesian inferential process. From my perspective whatever the result, it will serve as superficial justification for a structure void of substantive content, i.e. value.

The Bayesian is a method of determining the existence of specific, pre-assigned, pre-determined forms of causes. But does this secure us a decision between universal causality and non-causality at all?

Causality may exist in nature, but it exists only in our theories in attempts to explain (predict) observations. Because of data like that addressed in the Simpson's rule paper, I have come to want to examine the raw measurement data rather than to trust someone's statistical analysis.

In my own case, I tend to look for equations that fit the data with less variance. For example, in examining data about rotation velocity the data suggested (at least to me) several relations with luminosity of galaxies. However, I found some papers where another option was to eliminate some of the data. Over 80% of the total dat points favored a V^2 vs. luminosity relation with large uncertainty. By having several relations with a constant factor of 2 between them, ALL published data could be accounted and at a much lower uncertainty. So, ignoring some data seems preferable to the community than the uncomfortably several related slopes. So, the next issue might be to find the ``cause'' of the differing types of relations to luminosity.