Odds makers probably don’t know exactly how it will shake out but they should be able to guess accurately enough to be close and then make adjustments as the bets come in.

TV execs certainly have a pretty good idea of what to expect on viewership when they slot teams into time slots and channels for games.

I would guess that based on history and matchups they can be ballpark accurate in terms of total expected dollars bet by game. Within that set you have fans of both teams and then the casual bettors and degenerates. From there, you just have to pick a number that splits the dollars of those groups. Based on history and matchups, they do an amazing job of splitting the money. (If they don’t, they’ll quickly go out of business.)

I think it would be really interesting if sport books revealed their wins and losses by game. I’m sure they miss by a bit on every game (some favorable, some unfavorable) but win in the aggregate.

This is why I don’t think spreads represent what odds makers think will happen other than being in the middle of the betting.

I would also predict that if BYU beats the spread again that more unaffiliated bettors begin to flock to BYU and that it eventually moves the spreads. (Think about those betting tip services. They’d love to be saying, “Take the Cougars, they’ve beaten the spread in every game this year.” One or two probably doesn’t move the needle but several might.) (Maybe this has already occurred. I’m not sure I would know enough to detect when/if it happens. I’m sure I’m also simplifying this a bit.)