You're doing it wrong.
It's not about who scores the most points, it's about who scores the most points relative to other players at the same position. Generally the top tier of RBs severely outperform the 2nd and 3rd tiers of RBs by a tremendous margin. That's not to stay a statistical aberration QB year can't do the same, it's just not as common. It's not unusual for a team to have two RBs scoring 15 and 18 points on average while their QB scores 16 or 17 per week (a mid-round QB, such as Tony Romo), while another team has a top tier QB averaging 22-23 points but their best RBs are averaging 8 or 9 points. When you look at the scoring differential in this instance you have a much better shot with the two RBs.
Also, it's easier to predict where QBs will generally land in the first two tiers. If your RB picks come too late, you're taking guys like Tim Hightower and Reggie Bush and finding yourself up a creek when they get bunched by the new guys or get stuck in ugly RBBCs. Top RBs generally only lack production when they're injured (Chris Johnson notwithstanding).
Assuming guys like Aaron Rodgers and Tom Brady come back down to earth a bit (they can still be way ahead of the pack but if you look at their projected numbers for the season they'd smash all-time records by miles so they're likely to regress a bit), they may average 28 points per game which is outstanding, but if you can get 20 points per game with Tony Romo and have McFadden and Fred Jackson, while the owners of Rodgers and Brady started the season using Reggie Bush and Shonn Greene as one their RBs, then you're probably way ahead of them.