Football is a very expensive sport, played intensively during youth by far more white than black people in the US. The pro level hires out of college - you have to get into college to pursue a career in football. It is also a major and attractive road out of poverty for more white than black people in the US (there are more low income white folks than there are black folks total, in the US). It pays high wages, provides large social perks, and is honored as a young man's career choice among most white people in the US. It provides entry to careers at the very highest levels of US society - Senator, Judge, Corporate Executive.
The fastest runners and jumpers on the US pro football field are almost all members of the "black" race. The positions in which running speed and jumping power dominate the initial physical requirements are almost entirely manned by young men with melanistic skin. In the defensive backfield, say, even at safety (the slowest DB position) a white face that can play at a high level is unusual enough to be famous, like this guy:
http://en.wikipedia.org/wiki/Harrison_Smith_(American_football) He was a terror in college, at a position in demand, scoring far above average on most attributes for the job, but still dropped to 29th in the NFL draft (almost out of the first round) because of doubts about his running speed (his scout camp 4.57 40 is what is called "adequate"). Like the other white guys in the defensive backfields of the NFL, he was hired in spite of his running speed, not because of it.
This is almost certainly genetic. The human species does harbor geographically established genetic variation, geographical pockets of people with some characteristic genetic differences. High range short distance running speed appears to be one of them - characteristic of people from a genetic pocket in the Congo and West Africa.
But it is not racial. The sociological races do not reflect the genetic variation patterns. The "US black" race, especially, is a grab bag of widely disparate genetic groups (as widely as humans get, anyway) unlikely to resemble each other in any respect other than species norm (not even the melanistic skin is from a common genetic pool). By circumstance of history (resistance to yellow fever and vivax malaria, coastal access) the US drew almost all of its plantation slave population from that one genetic pocket, which seems to have created a frequently misleading coincidence of skin color and foot speed/jumping ability within the US, too often translated into a coincidence of foot speed and sociological race.
So the question is not why black people run faster than white - that just confuses the matter. They don't, is the simple answer. The question is (or would be) why people from Central and West Africa tend to on average run faster over short distances than everybody else on the planet, including all the other "blacks".
And when considering it, one might also inquire into why the slower short distance northeastern genetic pocket up by the horn, the Kenyan and Ethiopian pocket, is so good at long distance running - to the point that an English college level cross country coach (and England respects the sport) bringing his team of elite, trained, well fed, well equipped, young men at the peak of their abilities on tour through Kenya, might come to the conclusion that the average high school in the region could choose on the spot and field a team from its regular classrooms capable of beating his.