I know it makes me unlikeable and no fun to be around but I just can't find humor in most of these statistics humor memes that go around applied communities. The attempts at humor are almost always derived from real problems that are having real, harmful consequences.

For every "haha Bayes is slow" and "linear models do everything" and "MCMC sux" meme that claims to be joking there are real reviewers, sometimes the very same meme authors, propagating that misinformation in classes and forcing it onto authors in reviews and job applications.

Humor can absolutely be a powerful pedagogical tool to complement technical learning. I sneak all kinds of bad jokes in my writing. People like @ChelseaParlett tell (slightly) less bad jokes that demonstrate or emphasize particular concepts common in introductory classes.

But many of these stats memes, and much of the discussion around them, isn't coming from that pedagogical perspective. It's coming from people asserting what they think are truths based on irresponsibly little research on their own part.

These memes not only demonstrate serious structural and systematic problems in statistical pedagogy and competency, they propagate those problems by reinforcing the shallow heuristics that have come to replace actual statistics understanding in many fields.

But I'm just a bitter old man who has been broken by the continuous stream of unselfaware researchers, practitioners, and developers who are pretty sure they understand statistics well enough based entirely on the expectation that statistics is easy enough to learn.