He showed slides of a normal distribution for the previous semester's results, vs a bimodal distribution for the current semester - that's the aberration I'm talking about.