I have been trying to figure out if, and if so, just how is the shape of the molecules in a gas is taken into account by the Maxwell-Boltzman distribution. I know the assumption is perfect spheres, but still, a square hit and a glancing blow on a pool table yield different results. My intuition is also telling me that repeated off center collisions would be required to get those situations where atoms are traveling many-fold faster than the mean. It seems like SIN would have to be in there somewhere. Any insights?