So how did the Books lose?

[BottlesOf]

In today's NYPost, they report that sports books lost a record $563M (estimated) on this past weekend's NFL games, because 12 of 16 underdogs won.

Can someone explain why this would aversly affect the books? Don't betting lines change to keep an equal amount of action on both sides of every wager? I understand that's not a perfect science but they have to come close, don't they? Also, why would it matter that 12 of 16 underdogs (as opposed to favorties) won? Wouldn't it stand to reason that on some games there was slightly more action on the dog, and on others, slightly more on the fav?

Confused,

JBB