Dave Berri, my favorite sports economist, has even shown that there is no correlation between draft position for quarterbacks and pro performance: in other words, a quarterback taken in the first round is no more likely to be a great pro than a quarterback taken in the third or fourth round. He argues -- and I agree -- that the college game is just so different from the pro game that any kind of prediction is impossible. So what should we do? Well, if I were a NFL GM, looking for a quarterback in the next two drafts, I wouldn't even try to figure out in advance who would be the best. I'd put Colt McCoy, Graham Harrell, Chase Daniel, and Sam Bradford's names into a hat, pick one out, sign him to a short term contract and try him out. The only way to tell whether someone can play pro quarterback is to actually let them play pro quarterback. That's consistent with a lot of what I argue in “Outliers,” which is the real reasons for success are so complicated and so unknowable that we have to switch from trying to predict performance to being patience and content with measuring performance.
http://www.cnbc.com/id/28813545
BTW: Berri proved no relationship by excluding all the low round picks who never played in the NFL