In Las Vegas sports books, when the lines for upcoming games are released, a select few gamblers are allowed to place their bets before anyone else as the sports book tries to determine whether their line is good. Depending on which side these "sharp players" bet, how much they bet and whether they bet the line at all will determine if the line will move up, move down or stay where it is. Once the sports book is confident it has a good line, it is then released to the general public.
Once the line is released and the general public starts to bet against that line, the sports book will move the line (the point spread) and the price as needed in an effort to attract balanced action on the game. When it comes to football point spreads, you'll often see a line stay at -2.5, +3.5 or -6.5 for a long time and some odd prices like -125 or +110 associated with those lines. What the sports book is trying to avoid is moving the line onto or off of a "key number". Because of the way football is scored, there are certain margins of victory that occur more often than others. The margins are 3, 4, 6, 7, and 10. To a lesser extent in the parity-driven NFL, but more common in college football, 14, 17, 20 and 21 come into play as key numbers.
When a sports book has been taking heavy action on the favorite at -2.5 and moves the line to -3, then they risk getting "middled" on the game. What would be the bookmaker's worst nightmare is for the favorite to win by exactly 3 points, forcing him to pay the bets laid at -2.5 but pushing with everyone who played the game at -3. Even worse would be if the book decided later during the week to go from -3 to -3.5 with the favorite subsequently winning by 3 points. The book would then lose on the -2.5 bets, push on -3 and lose on those customers who were finally enticed to bet the underdog at +3.5.
A lot of times, especially in college football, it does not matter what the sports book does with the line and the price; they just cannot stop the money from coming in on the favorite. Just such an example happened during the afternoon in advance of the first major college football game between LSU and Mississippi State. The line opened at LSU -16 (-110) in early August and I immediately placed a bet at that number, believing the true number to be somewhere around LSU -24 to LSU -27. The spread then hung around -16 and -16.5 for a couple of weeks before finally making its way across the key number of -17 and more or less settling on LSU -18 (-110)
Then, on the day of the game, as the betting public started to get their action down, the line started to move. Evidently nobody wanted anything to do with Mississippi State as the line was -18.5 at 2:00 pm and had moved to -19.5 by 3:30 pm. It got as high as LSU -21 (-105) at Bodog.com, which then presented an interesting opportunity for the people like me, who had an open bet on LSU -16 (-110).
Just like the sports books try to avoid getting middled by having a stable line, going for the middle as a bettor is an excellent opportunity that not only offers attractive odds on your bet, but also dramatically lowers the variance in your bankroll.
From the bettor's point of view, trying to "Middle" a game is a form of hedging your bet. In this case, I had an open bet on LSU -16 (-110) for 1.1 Units and was presented with the opportunity to place a second bet on Mississippi State +21 (-115), which I would make for 1.15 Units. If I opted to take this opportunity, then I would essentially be converting my bet from "LSU to win by more than 16 points" to "LSU to win by 16, 17, 18, 19, 20 or 21 points". The downside is that if LSU ends up blowing out MSU, then I win my original bet (+1 Unit) but lose the new bet (-1.15 Units), for a net loss of .15 Units. The upside is the potential payoff (+2 Units) should LSU win by any of those scores. Since I've defined my risk (.15 Units) and reward (2 Units) on the bets, I'm receiving 13.33:1 odds (2 Units / .15 Units) to try for the middle. Getting these odds means I only need to win both bets 7% of the time to break even. Anything above 7% means profit.
Obviously, the next problem is to quantify how often a favorite in this situation wins by exactly 16, 17, 18, 19, 20 and 21 points. To answer this, I researched my database to come up with all games where a road team was favored by between 13 and 22 points. Below is a table containing the results from the last 20 years of college football. There were 1,153 games fitting this criteria, including the percentage of those games where the favorite won by the point margins listed above:
| # of Games | Margin of Victory | Probability |
| 20 | 16 | 0.0173 |
| 49 | 17 | 0.0425 |
| 24 | 18 | 0.0208 |
| 18 | 19 | 0.0156 |
| 24 | 20 | 0.0208 |
| 44 | 21 | 0.0382 |
The last quantification needed before we make the comparison between our two options is how often we expect LSU to cover 16 points. When anticipating the early lines for Week 1 of college football, I felt the line for this game should be in the neighborhood of LSU -24 to LSU -27 and was fully prepared to immediately bet anything less than LSU -21. In order to determine the edge on LSU -16, I researched the margins of victory where the road team was between a 24 and 27 point favorite and calculated the cumulative probability of these road favorites winning the game by 16 or more points. For 165 games where the road team was favored by 24 to 27 points, the results are below:
| # of Games | Margin of Victory | Cumulative Probability |
| 1 | 16 | .6848 |
| 3 | 17 | .6788 |
| 3 | 18 | .6606 |
| 1 | 19 | .6424 |
| 3 | 20 | .6364 |
| 8 | 21 | .6182 |
| 2 | 22 | .5697 |
| 0 | 23 | .5576 |
| 5 | 24 | .5576 |
From this data, we can conclude that a road team favored by 24 to 27 points will cover 16 points 68.5% of the time. Since we need to win 52.5% of our bets when laying -110, we should be able to count on a 16% edge on this particular bet.
Finally, from our research, we should be able to make an informed decision in response to our opportunity to make the second bet on Mississippi State +21 (-115).
From the first table, we sum the probabilities of the favorite winning by exactly 17, 18, 19 or 20 and come up with 0.0997 or roughly 10%. Since we have to win both bets 7% of the time and the probability of that happening is 10%, this is a bet with positive expectation. However, our edge is only 3% on both bets while our edge is 16% should we stay with our current position.
In the end, there is no reason outside of bankroll management considerations where we should give up 13% of our edge to make the second bet. When playing on a short bankroll, however, you may want to consider making the second bet to considerably reduce your risk.
This analysis highlights several important aspects involving betting on football:
1. Be prepared to bet when lines are released.
This opportunity, even though I passed, would not have presented itself had I not handicapped this game prior to the line being released
2. Continuously monitor line movement
Again, I would not have had this opportunity had I placed my bet and then ignored what was going on in the betting market.
3. Have a general idea of your edge
The final decision to pass on Mississippi State +21 (-115) would not have been as easy without the ability to approximate the edge on my open bet and the potential second bet. I was ultimately surprised that getting the "key numbers" of 17 and 20 and pushing on the last key number of 21 did not offer nearly the edge I expected. Without being able to do this kind of analysis, I may have made a misstep and cost myself a lot of expected value.
T
. 6 players limp in, including tight old lady on the button and crazy asian gambler one seat to her right. With $28 (7 small bets) in the pot, the flop comes K
and the small blind checks. 
. I'm not sure there could have been a better card in the deck. There is no way that this card could have beaten me. So, I led with another $8 bet. The old man called and CAG folded. I showed my Tens and he mucked his hand without showing, presumably a pair of nines or something.