The collection of recruit's individual ratings are summed, but not in a uniform way ( which it looks like many people are assuming). They are weighted using a Gaussian distribution, which may or may not be more valid. If 2 teams have the same number of recruits, and an identical average composite rating....they would be equal if the distribution was uniform. Since it is Gaussian, a team with more highly rated recruits is rewarded more than they are punished for having lower rated recruits.
My best attempt at an explanation.