My understanding:
It is the relative gaps between internal marks within a cohort. That is all. None of these "marks" and none of these "ranks". These internal marks are moderated according to the external marks. (ie they use how the cohort did in the externals to determine how they should do in...