Betting College Football

Introduction

The purpose of this page is to explain the various ways to bet on college football. I've provided my usual statistics and advice.
 

Averages

The following table may not be of any help in the sports books, but I thought it was an interesting way to compare college football to the NFL.
 
 

Margin of Victory

The following table shows the count and frequency of each margin of victory.
 

The next table shows the top ten margin of victories, in order.
 
 

Point Spread

The primary way to bet any football game is against the spread. The following table shows the probability of a win, loss, and tie, as well as the expected value according to all combinations of betting on a home/away and underdog/favorite. This table assumes the bettor is laying 11 to win 10, which is almost always the case.

 

Over-Under

The next table shows the possible outcomes and expected value of betting against the over/under line.
 
 

Money Line

The following table shows the number of games in my data at each point spread on the underdog, and the number of times the underdog won. It also shows my estimated probability of winning after smoothing out all the ups and downs. The right column shows the fair line based on the estimated probability of winning.
 

To get the probability of a favorite winning, just take the difference from 100% of the underdog winning in the same game. For example, a 7-point underdog has a 30.6% estimated chance of winning. Thus, a 7-point favorite would have a 100%-30.6% = 69.4% chance of winning.

It should be mentioned that the estimated probability does not take into consideration key numbers. For example, the difference in probability of winning between a 2.5- and 3.5-point underdog would likely be more than the 42.7% - 39.9% = 2.8% suggested in the estimated probability column.

I also looked at this curve by whether it was a home or away underdog but did not find it statistically different enough to pay special treatment to.

The following graph illustrates the actual and estimated probabilities of winning.


 

Buying Extra Points

The following table shows the probability of winning, not counting ties, after getting an extra 0.5 to 2.5 points according to the fair spread. For example, you consider the fair spread on a team to be +3, but somebody will give you +3.5, then your probability of winning would be 53.42%.
 

A practical use for the table above is in buying an extra half point. Most casinos will give you an extra half point if you lay 120, instead of the usual 110. The following table shows the probability of winning and expected value off of the five best spreads to buy a half point off of. Assuming a fair spread, the house edge laying 110 is 4.76%, this is worth the price to buy the extra half point off of spread of 3, 7, 10, and 14 only.
 
 

Teasers

Teasers are basically a sucker bet. Under the exact right circumstances, with generous odds, they can be a good bet. However, for recreational bettors picking teasers arbitrarily, they are an awful value.

For those who don't know, a teaser is like a parlay, except in the case of college football the player gets 6 to 7.5 extra points per leg. For these extra points, the wins are significantly less than a parlay.

To begin my analysis, the following table shows the probability of winning each leg with 6 to 7.5 extra points and according to whether the player is teasing a side, under, over, or doing a "Wong" teaser, which I'll explain momentarily.
 

Before going further, let me explain what a Wong Teaser is. As far as I know, the first time the idea was published was in Stanford Wong's book Sharp Sports Betting. Wong correctly noted that 3 and 7 were the two most common margins of victory in the NFL. He advised teasing over these two key numbers. It works only fair spreads of +1.5 to +2.5 and -7.5 to -8.5. For example, teasing a +2 underdog to +8, allowing the team to lose by a field goal or touchdown and still cover the spread.

While this could be done with any number of points in the teaser, it is only worthwhile with six-point teasers, due to the depressed payouts getting more than six points.

That said, the next table shows the expected value of six-point teasers by the various number of picks and pays I have seen and across the type of bet being teased.
 

As you can see, there are a lot of very negative expected values in the table above. The few positive ones are for Wong teasers only and only under the most generous odds, which can be hard to find. The liberal pays with the positive expected value for Wong teasers can be found at Coast/Boyd casinos.

I should also state that there were only 1,780 games in my data that had a spread in the Wong teaser range. The 71.57% chance I show of a Wong teaser winning should be considered rather rough, with a standard deviation of 1.70%. Other research leads me to believe the probability of a Wong teaser winning any given leg is closer to 68.9%.

The next table shows the expected value for 6.5-point teasers. As mentioned before, Wong teasers are only good when buying six points, so that column is omitted from the 6.5, 7, and 7.5-point tables.
 

The next table shows the expected values for 7-point college football teasers.
 

The next table shows the expected values for 7.5-point college football teasers.
 

To conclude the topic of college football teasers, stay away from them with a 10-foot pole unless you are crossing both the 3 and 7 point margins of victory with a six-point teaser on every leg as well as getting the liberal Coast/Boyd casino odds.
 

Data

The data used in this analysis is from 12,231 college football games from 1996 to 2013. The reason I started in 1996 is that is when overtime rules were introduced to eliminate ties. The data was rather dirty, so I had to cut out a lot of games that had ridiculous information, like a point spread of -100, a negative over/under line, the sum of the points scored by quarter was less than the total points in the game, or the game ended in a tie. I'm sure there is still some bad data in there, but I believe the vast majority to be okay.
 

Written by: Michael Shackleford