Inspired by the desire to find an objective selection criterion which might be useful notably in borderline cases of hip dysplasia, 1265 sets of radiographs, taken on 1128 pedigree Rottweilers, have been assessed for 12 particulars of the joints; three subevaluations as well as a final score (FS) have been obtained for each individual involved. The data so obtained from a representative subpopulation have been analysed for their contributions to the 12 assessments leading towards the value of the final score. This article shows that FS, most notably in the borderline cases, is due to a subjective assessment, which disqualifies FS as a selection criterion. In pursuit of this line of thought, it was considered desirable to investigate whether FS could be replaced by some measurable (hence objective) particular assessment, such as the Norberg-Olsson Value (N). In conclusion, we find that N on its own, though objective, fails to meet the requirements.