process is frequentist

(aka, why frequentists will always make more money)

In the "real" (corporate/governmental) world, most high-level decision making is frequentist, because being senior and influential means you have control over processes (general rules) more than individual decisions.

Frequentist statistics is all about designing processes:

• given a procedure, how well does it work?
• What is the best procedure under some criterion?
• If we do the same thing over and over again, on average how many successes will we see?

By contrast, Bayesian approaches are about particulars: given this set of circumstances and these assumptions, what's the right conclusion to draw?

As a senior decision-maker, you're not doing the science yourself; you're setting up the process by which scientists are trained. It's not your job to write a good paper; it's your job to have a set of rules and a process that produce good papers. It's not your job to make the perfect business decision every time; it's your job to have a process in place that generally produces good business decisions.

Any pursuit that requires coordination of multiple people requires process. Science 'writ large' is inherently frequentist, because the community needs a standard of evidence. There needs to be a process that, if people follow it, generally produces knowledge about the world. Publishing p-values might not be the right process---it might be that the right process is for every paper to specify a Bayesian model and report likelihoods---but ultimately it's still a process and has to be evaluated on frequentist terms.