The method of normalisation combined with peer ratings is utilised to provide the solution to the biased rating problem of mapping group work marks to individual marks. We critically evaluate the method of normalisation following the findings of an article which argued against the use of self and/or peer rating mechanism. We demonstrate that the findings of that article also hold for the method of normalisation as the influence of human behavioural factors are not accommodated in the designs. Additionally, we argue that the method (and its variants) is rather complicated, where all possible contingencies are not pre-specified. It makes the arrangement between tutors and students in conducting peer assessments incomplete and unverifiable by a third party.