It is notable that many of the mysteries belong to the shrouded business of examining undergraduates. Far into the summer, rooms full of dons argue passionately about whether Rupert or Fiona deserves a 2.1 or a 2.2. Like too many others in the ancient universities, this process is secret, Byzantine and uncontrolled. When, in Oxford, I came to examining in the school of philosophy, politics and economics in the 1980s, it meant interpreting "sums" of 16 initial marks for each candidate, presented as beta-beta-alpha-query-plus or alpha-beta-minus-minus. It took 20 dons a mind-numbing week to sift through the several hundred candidates. In time, as chairman of this rigmarole, I got the whole thing computerised, but not before meeting passionate opposition from dons who believed that numbers lack the diamantine precision of alpha-beta-minus-query-plus.
