During my physics degree we had relatively fine-grain evaluation - every module was examined and graded (e.g. /100) individually and the sum formed the basis of

During my physics degree we had relatively fine-grain evaluation - every module was examined and graded (e.g. /100) individually and the sum formed the basis of your degree. Some modules were compulsory, some you had one shot to pass (labs), some took an entire semester, some were half, etc. You could work out exactly how much of your degree an exam was worth, to the point that some people knew that they could sit out most of their final exams and still get a good degree. With the exception of the dissertation and occasional paper-writing exercises in labs, there was always a correct answer and marks were normalised using a formula that a previous professor had invented. It was expedient to take the hardest courses (e.g. Gauge Theories of Particle Physics) because you would almost certainly be boosted up from the low class average, rather than "simple" ones where everyone did well. By the third year, homework was optional. You also wouldn't flunk a degree if you bombed a course. Labs were universally hated for marking because it involved qualitative assessment (beyond "did you make the experiment work) and it was run by PhD students who each had wildly inconsistent criteria, despite presumably working from a rubric. I'm somewhat sympathetic because I've taught labs and a lot of students are just inept, but the grading system should really have been overhauled there. Compare that to the English students in my cohort who had what we'd call a "doss" 1st year (worth 0%) and then in their second and third years might sit four essay exams that were worth 12.5% of their degree. As far as I know, some courses were 50/50 assessed/invigilated, but many were 100% examined.