Why (not) frequentist inference (too)

It was very kind of Prof. Gomez Villegas to invite me to write an article with the title “Why Frequentist Inference”. However, when thinking things over, it occurred to me that there is no reason to take the position of defending frequentist inference, or in trying to convince anyone that frequentist inference is the only correct way to do statistics. To take such a position is like trying to convince Raul Gonzalez, a star of the Spanish World Cup Soccer Team, to kick goals with only his right foot. The position that I am comfortable with is to view frequentist inference as one part of the tool kit of the complete statistician. Some things are best done using Bayesian tools, some things are best done using frequentist tools, and the complete statistician uses all of the available tools. (We also do not forget that there are likelihood tools, nonparametric tools, etc.)