in any big cohort of students, it is safe to say that the bell shape curve will apply.....

and in comparison with different cohorts separated by years of birth, the cohort ability and the curve will be about the same.....

the use of marks for grades suffer from the fact that questions are not equivalents, and hence of different difficulty, hence giving rise to different absolute marks in different exams/years....

but if you plot the grassian curve, it will almost exactly overlap each other, for different years....

which method of expressing results to use thus depend on the reasons for such stratification.....

if it is the determine the amount of knowledge, then absolute marks is needed....eg, how many right answers you need to pass a driving test.....

but if it is purely to select the best students for a particular course, then relative grading (ie results expressed as relative rather than absolute performance) gives better differentiation.....eg selecting student for a medical course.....

there are lots of combinations of the two extremes in reality, depending on other priorities that may exist.....