The basic strategy for sports-book operations is to do its best to balance its books on all games. That means balancing "exposure" on both sides of a game, not balancing ticket counts. One-hundred $10 wagers can be offset by one $10,000 wager, which explains why everyone can be on the Pats -7 vs. the Chargers, but the game closes -6.5.
Also, on weeks when favorites are covering, the $10-$100 bettors can really make the MNF-game-favorite liability explode, as the public loves to play parlays with the MNF favorite on the last leg. The exposure can really ramp up on those weeks.
Incidentally, while the basic strategy set forth above is a decent one (get flat and just earn as a book), in no way is it close to an optimal strategy. Most Nevada casinos don't realize just how much they leave on the table by not finding competent and savvy sports-book managers who can sniff out the "right" sides and manage their portfolios in such a way that they shade numbers to get more action on the "wrong" sides. Granted, that's easier said than done. Certainly, it can be disastrous to deal a rogue number and take several big bets, then later find out that your QB is hurt, the weather report is terrible, or any number of other random but negative developments that diminish your $50K position with the wrong side. The way to protect against this is to get to know the players. If a top winning player is firing on something, it's almost always correct to move the number ASAP.
One reason so many casinos balk at employing "wise guys" to run their books is that most of them have friends on the other side of the counter. It would just be too easy for them to deal soft opening numbers that their friends could take advantage with no one knowing it was going on. In fact, many former sports-book managers in Vegas (some of whom are well-respected) have checkered pasts involving this sort of thing.
In the final analysis, who can really blame the casinos for simply hiring a good business manager who has no handicapping ability, but 1) copies the current market numbers, and 2) knows to move his numbers when his exposure gets too high on any one side? I guess I could, but I'm not.
