Why is rolling friction less than the maximum static friction?

I know that rolling friction is almost always less than the maximum static friction, but why is this? In particular I am researching why ball bearings have much less friction. It has been said that the ball bearings come in contact with much less of the metal, but this does not make sense to me as friction is not proportional to surface area. So if it has nothing to do with surface area, then why is rolling friction generally much less than sliding friction?

• Im no expert, but I thought static friction was "rolling friction". I never heard of rolling friction as a term. Static friction is what keeps the contact point from sliding; sliding would fall under kinetic friction. This is the only time "rolling" as opposed to sliding is possible. – CogitoErgoCogitoSum May 12 '17 at 19:46
• The surface area must be involved. A perfect spherical surface would only touch a flat surface at 1 singular point at each angle of rotation, whereas a cube surface would have an entire area touching the flat surface upon sliding. Electrical forces and bumps in non perfect surfaces are involved here. I also belive a rolling motion of a perfect sphere only has surface friction similar to a single point touching a surface and then being interchanged with another point a small distance dx, further "down the line", therefor the static friction. – Gjert May 12 '17 at 21:17