Part of this relied on the professors distributing lecture slides afterwards, which they always did. Helped jog my memory later and remind me of what to study for the final.
The strategy of only writing down the important details that everyone advocates does not work very well in math. I can't know what the important details are until days or even weeks after the lecture, when I've had enough time to digest and work through the material while completing the assigned problems. Of course, this is much more true at the beginning of a course (when I've had no prior exposure to the topic) than it is at the end.
I went back and forth between pencil and paper and LaTeX for assignment work. For some courses the proofs were fairly short so it was easy enough to work on paper. For others, the proofs could get several pages long and small mistakes could lead to lots of erasing. It doesn't take very much erasing/rewriting to annoy me enough to switch over to LaTeX. I can't think of a bigger waste of time than erasing and/or rewriting half a page of material because of some mistake.
The poor student followed him all the way out of the room and down the hall, begging not to be reported to the associate dean's office!