90% of the T Distribution

A quiet post about old-school beer math somehow turned into a delightfully messy comment-section variety show. The article itself is simple enough: a Guinness statistician, forced to publish under the fake name Student, figured out that when you only have a few data points, your “90% sure” range should be wider than people casually assume. In plain English: if you don’t have much data, don’t act so confident. That part was almost wholesome.

But the community, naturally, had other plans. One camp jumped straight into correction mode, arguing the post mixed up “standard deviation” with “standard error” — basically accusing the author of using the wrong label for an important idea. Another commenter pounced on the phrase “number of samples,” insisting it should really say “sample size.” Yes, the stats post instantly became a word-policing arena, and readers seemed weirdly thrilled about it.

Then came the plot twist no one saw coming: a commenter got distracted by the author’s about page and asked what on earth a line about his wife meant — specifically, that he could never repay her love “in degree nor kind.” Suddenly the thread wasn’t just about confidence intervals; it was about romance, language, and whether this mathematician accidentally wrote the most intense wife compliment on the internet. So while the article tried to teach caution with small numbers, the comments taught a more timeless lesson: online, someone will always debate the wording, and someone else will absolutely derail the room with poetry.