correct if I'm wrong(Im sure someone will,lol) but the pointspread is not meant to predict team A will win by X pts over team B. In perfect world books would have even split on games and be guaranteed a profit due to juice. Yes, they still come out ahead most of the time even though games may run 70-30 60-40 etc on money splits.
the oddsmaker(s) may feel SF beats Dallas handily in week one this year but they aren't going to put out SF -11. instead they put out a number that hopefully gives everyone pause i.e. should I take the 4 or 5 or give it.
since you do have lopsided outcomes 35-10 etc then things get skewed especially if the line was -3 or so in a 25 pt win.
I dont make my own line in NFL but estimate what I think the line will be. If my thought is off or way off then I start looking to see what Im missing.
just my two cents on trying to sound smart.