This isn't possible. Consider: There are four divisions in each conference. Each division plays each team in one other division of their conference (i.e. AFC South played AFC East). Now: Consider all divisions mentioned are in the same conference. Suppose team X has 16 wins. This means everyone in their division has at least two losses. This also means that every team in one of the other two divisions has at least one loss. We'll say the divisions are A, B, C, and D. Team X is in Division A and played Division B, thus everyone in Division B has at least one loss and everyone that isn't team X in Division A has at least two losses. Now suppose team Y (which is not team X) has 16 wins. Team Y cannot be in Division A or B, since they are not team X and they have 0 losses. Thus, let's put them in Division C. Since Division A played Division B, Division C must have played Division D. Thus, since team Y has 16 wins, every team that isn't team Y in Division C has at least two losses and every team in Division D has at least one loss. Now suppose team Z has 16 wins and they aren't team X or team Y. Team Z cannot be in Division A, because everyone that isn't team X has at least two losses. They cannot be in Division B, because everyone in Division B has at least one loss. They cannot be in Division C, because everyone that isn't team Y has at least two losses. And they cannot be in Division D, because everyone in Division D has at least one loss. Therefore, team Z cannot exist, thus the maximum number of possible undefeated teams in a given conference is two. Q.E.D. I'm glad my math degree is good for something.

Double digit wins AND improved second halves. We couldn't make any adjustments during a game after halftime. I'm sick of the regression during games.

Belichek is the kinda guy that would push his mother down a flight of stairs or in front of an on coming train for a win. Bad reference Gunny.