You obviously don't have an understanding of statistics. Things like confounding variables can make things seem like they are correlated when they are not or not as strongly as they suggest. For example a treatment for weight loss may seem more effective than it is because the researcher neglected to take into account lifestyle changes. A dishonest person could inflate the effectiveness of the treatment by ignoring this confounding variable there by lying with statistics.
The raw statistics like passing yards or td's may suggest that Dobbs is the second best qb but there are other variables to consider, such as supporting cast, era, offensive system, etc.