While competing in the Stardust football tournament in the fall of 2000, I had the opportunity to ask Robert Walker, race and sportsbook director at MGM Mirage, whether his company's action is balanced on football games. I mentioned that it seemed to be me that on about half the games, most people who had an opinion ageed on the same team. I said it therefore appeared to me that sportsbooks might be taking lopsided action on about half of all NFL games. Mr. Walker answered that what I hypothesized was indeed the case. He said he does not mind lopsided action as long as his sportsbook makes money overall on those games.
Then I asked him why he did not make more line moves to try to balance the action. His answer was that to move the line to try to balance the action would result in less profit for the sportsbook.
Balanced action is seldom achieved. It's a myth that balancing the action is the purpose of the line, except in the case of the Super Bowl.
...
Sportsbook managers strive to achieve better than the theoretical hold percentage. They try to set lines so that the excess action is on the side that is less likely to win. It's easier to do this on games with a higher proprtion of money being bet by unsophisticated bettors.
Suppose that a sportsbook manager thinks there is a 50 percent chance that Baltimore will win Sunday's game by more than 7.5 points. And further suppose that the manager thinks that hanging a line of Baltimore -7.5 would attract an imbalance of action on Baltimore, but that a line of Baltimore -10 would split the action evenly between Baltimore and its opponent.
A line of Baltimore -7.5 would allow the sportsbook to achieve its theoretical hold percentage, as would a line of Baltimre -10.
At a line of Baltimore -7, most of the action would come in on Baltimore, and Baltimore would be the side that is most likely to cover, so the sportsbook would not have an edge.
At a line of Baltimore -10.5, most of the action would come in on Baltimore's opponent, and that would be the side most likely to cover, so the sportsbook would not have an edge.
Any line between Baltomire -8 and Baltimore -9.5 will attract more money on the side that is less likely to cover the spread, and give the sportsbook the expectation of achieveing better than the theoretical hold percentage.
The manager wants to make money for the business, and the more money the better. The point of maximum dollars of profit could come at Baltmore -7.5, or it could come at Baltimore -10, but more likely it would come at a number between -7.5 and -10. Within the range of -7.5 [sic -- I think he meant -8 -- MT] and -9.5, the sportsbook would knowingly be accepting an imbalance of action on Baltimore.
Sure the sportsbook takes extra risk by intentionally accepting an imbalance of action on Baltimore. But the sportsbook diversifies by also accepting bets on many other games, each of which is independent. Therefore the risk due to an imbalance in action on any one game is unimportant unless the action on that game is huge, as is the case with the Super Bowl.
In the real world, of course, the sportsbook manager does not know for certain the spread at which Baltimore has a 50 percent chance of covering. But between his expertise and the expertise he has access to, the manager comes up with a good estimate of that spread.
Likewise in the real world, the sportsbook manager does not know for certain the spread at which action would be split evenly between both teams. He does not need to know it.
Suppose the manager thinks Baltimore has around a 50 percent chance of covering a spread of eight points. Further suppose the manager thinks that an excess of money will be coming in on Baltimore if he hangs a line of Baltimore -8. So the line he hangs is Baltimore -8.5 or Baltimore -9. He's not trying to balance the action on the game, and he's not even trying to attract bets against Baltimore. What he is trying to do is gain a little extra advantage from all those squares who will bet Baltimore no matter what the spread. If he hangs Baltimore -8.5 instead of Baltimore -8, then one possible game outcome that would have been a push for Baltimore bettors becomes a win for the sportsbook.
The result will be a line that might attract sharp money to the unpopular team. The manager does not really want sharp action, but will put up with a little bit of it in order to get an extra-big edge over a larger amount of action on the popular team.
If the manager can set a line that will attract some but not a lot of action on the unpopular team, the resulting spread will win more money for the business than if the line split the action evenly.
...
Squares prefer to bet favorites more often than not. So the manager is more likely to hang a line that penalizes betting on favorites.