Repeatability problems involve the whole process, from sampling consistency to lab measurement variances. That's why I don't put a lot of stock in these UOAs showing which oil is better unless you see a magnitude of difference over several OCIs and samplings.