[q845712]

i audited university of Chicago's undergrad "hard-track" real analysis (i forget the name, but iirc there was an easier version of the course available).

The problem sets would have several questions you were expected to solve, and then often 1-3 questions marked with 1 or 2 stars. I forget exactly the language, but there was some explanation that you probably couldn't solve the 1-star problems, and solving the two-star questions could be someone's PhD thesis, as it would require novel insight into an unsolved problem.

The point was that it was a class for young aspiring mathematicians many of whom were used to being the smartest person around, and this helped them confront problems they couldn't solve without feeling like total failures. Also at that kind of school, you never know, someone might be able to solve one.