My College Basketball Method

I know people are anxious for college hoops picks, but I need a few weeks worth of games to be in the books. I solely use stats from the current year and need each team to have several games under their belts. That should be around Dec. 14 but who knows with COVID?

Here is the college basketball betting chapter from the second edition of Becoming a Winning Gambler. It explains my method in detail and will get you going if you want to start a bit earlier.

College basketball has been fairly good to me over the past couple of years and my handicapping method is one that I presented in the first edition, so we’ll concentrate on it in this chapter and using it yielded a 154-123-6 (55.6%) record against the spread before the 201920 season came to an abrupt halt in March during the conference tournaments.

There were quite a few plays, but it’s important to note that in the daily college basketball write-ups I did to get that record, typically involved just the late games on the schedule. Due to time constraints of getting the articles out in a timely manner, I might have just looked at 12 games on a 40-game schedule. Somebody with more time could have easily had many more plays with a similar winning percentage, so once again, it comes down to time and how much you can put into your handicapping.

This method is a bit time consuming, as I mentioned earlier, but it does allow you to predict both sides and totals and is a valuable tool during the season. For this method you need the offensive and defensive averages for both teams playing, as well as both team’s average opponent power rating (AOPR), which is also referred to as strength of schedule. You also need the average number of points scored by college basketball teams, both for the full game, as well as in the first half.

Obtaining the Numbers

To get the average number of points scored by college basketball teams, I’d again use SDQL, as it will just take several seconds.

Getting an AOPR figure is a little bit trickier. There are several websites that will have them posted, with USA Today’s Jeff Sagarin probably the most widely known, but I prefer to use Ken Pomeroy’s website at kenpom.com, but you do have to make an adjustment to his numbers.

According to Pomeroy, the Michigan Wolverines had the toughest schedule in college basketball for the 2019-20 season, a shade more difficult than the schedule of Kansas, who was listed as having the toughest schedule for the majority of the season.

Michigan’s strength of schedule rating for the season was a +12.79. To get an AOPR that can be used with this method, subtract Pomeroy’s highest strength of schedule rating from 100. In this case, the number is 87.21, as 100 – Michigan’s rating of 12.79 is 87.21. Add 87.21 to Pomeroy’s strength of schedule rating for every team to get your numbers.

For scoring averages you can use SDQL, but I found it’s easiest to use the game matchups found at StatFox, which not only gives you overall scoring averages, but also home or away averages depending on the location of the game, as well as first-half scoring averages, which makes doing our first-half calculations much quicker.

Using the System

Once you have the average number of all teams, you will compare it to the numbers of both teams playing in the game that you are handicapping.
For an example, we will use an average number of 68.1 and look at a game between Team A and Team B. Team A is the road team and is averaging 70.1 points per game and allowing 64.6 points a game with an AOPR of 71.7. Team B is averaging 68.8 points per game and allowing 69.4 points per game with an AOPR of 76.4.

The first step is to compare both teams’ averages against the average and come up with offensive and defensive percentages. Team A is scoring 70.1 points per game, which divided by 68.1 gives an offensive percentage of 1.03, while Team B is scoring 68.8 points per game, which divided by 68.1 gives an offensive rating of 1.01.

For defense, Team A is allowing 64.6 points per game, which divided by 68.1 yields a defensive rating of .95. Team B is allowing 69.4 points per game, which divided by 68.1 yields a defensive rating of 1.02.

The next step is to add Team A’s offensive rating to Team B’s defensive rating and then subtract one, giving a figure of 1.05. Do the same for Team B, using its offensive rating of 1.01 and Team A’s defensive rating of .95 to get a figure of 1.96, which becomes .96 after you subtract one.

Next, multiply Team A’s figure of 1.05 by the median figure of 68.1 to get a predicted score of 71.5 points. Multiplying Team B’s figure of .96 by 68.1 yields a predicted score of 65.4 points.

The next step is to factor in AOPR and you will divide the higher AOPR by the lowest to get a percentage. Team B’s AOPR of 76.4 divided by Team A’s AOPR of 71.7 gives a figure of 1.07, meaning Team B has played a schedule that is 7% more difficult than Team A.

Next, divide the percentage in two, which will give you 3.5%, meaning you will decrease the score of the team with the lower AOPR by 3.5% and increase the score of the team with the higher AOPR by 3.5%. The new projections would now be for Team A to score 69.1 points, which is 71.5 divided by 1.035, and Team B to score 67.7 points, which is 65.4 multiplied by 1.035.

The final step is to factor in four points for home court advantage. Since you are using the numbers to predict a final score you can’t simply add four points to the home team, as that will throw off your predicted total by four points. Instead subtract two points from the visitor’s score and add two points to the home team’s score, which in this case will give a predicted score of Team B 69.7, Team A 67.1.

Let’s look at one more example, using the same average of 68.1, between Team C and Team D. Team C is the road team and is averaging 75.8 points and allowing 79.6 with an AOPR of 77.2. Team D is averaging 72 points a game, allowing 68.7 and has an AOPR of 73.6.

Team C’s offensive rating is calculated by dividing 75.8 by 68.1 to get a figure of 1.11, while Team C’s defensive rating is calculated by dividing 79.6 by 68.1 to get a figure of 1.17. For Team D, its offensive rating is figured by dividing 72 by 68.1 to get a number of 1.06, while Team D’s defensive rating is 68.7 divided by 68.1 to get 1.01.

Team C’s offensive rating of 1.11 plus Team D’s defensive rating of 1.01 equals 2.12. After subtracting one you get a figure of 1.12. Team D’s offensive rating of 1.06 plus Team C’s defensive rating of 1.17 equals 2.23, which becomes 1.23 after one is subtracted.

Team C’s rating of 1.12 is multiplied by the median of 68.1 to get a predicted score of 76.3 points, while Team D’s rating of 1.23 is multiplied by the median to get a figure of 83.8 points.

Team C has the higher AOPR of 77.2, which divided by Team D’s AOPR of 73.6, is 1.05, meaning Team C’s schedule has been 5% more difficult than that played by Team D. Dividing 1.05 in half gives 1.025, meaning that Team C will see an increase of 2.5% in its predicted score, while Team D receives a decrease of 2.5% for its predicted score.

Team C’s predicted score of 76.3 multiplied by 1.025 becomes 78.2, while Team D’s predicted score of 83.8 points divided by 1.025 becomes 81.8. When you subtract two points from Team C and add two points to Team D for home court advantage, the final predicted score becomes Team D 83.8-76.2.

This method is loosely based on one presented in Sports Betting: A Winner’s Handbook by Jerry L. Patterson and Jack Painter and can be used during the course of the season as a solid mathematical foundation for your handicapping.

Despite my limited Excel abilities, I was able to create a spreadsheet that I use to do the calculations, so it’s just a matter of plugging in the numbers.

First-half handicapping

I’ve had a fair amount of success the past few years betting first halves and the calculation method is exactly the same. The only difference is in the numbers we use, which are first half scoring figures for both the teams and the league-wide scoring average.

One thing I like about betting first halves is that sportsbooks tend to follow a formula for creating first-half lines and totals. The spread is typically half of the full-game line, but that will climb for larger favorites. A team that is favored by 4 points for the game will be favored by 2 points in the first half. But an 18-point favorite for the game may be favored by 10.5 points in the first half.

For first-half totals, the ballpark figure is half of the full game total and then subtract 4.5. A contest with a full-game total of 140, will see a first half total of roughly 65.5, while a game with a total of 150 will see close to a 70.5 for the first half.

Some teams play better in the first half than they do in the second half, especially certain weaker teams, who may not be very deep and are forced to rely on their starters too much. Fatigue starts to set in during the second half of the game and things begin to come unraveled. But these teams can be decent wagers in the first half and some poor teams had solid scoring averages for the first half.

For first half wagers, I liked to have a difference of at least three points between the projection and the line. I also look at the home and away tendencies of both teams and ideally, they will agree with the play. If the numbers like the under, the road team will see fewer points away from home than their overall average, while the home team will have lower-scoring first halves in front of the home fans.

If the overall numbers like one side, but a team tends to play the opposite in their location, I will frequently pass the game. If the numbers are calling for 62 points in the first half of a game between Iona at Marist, but Marist sees 6.4 more points in the first half of their home games compared to their overall numbers, I would most likely sit the game out.

Leave a Reply

Your email address will not be published. Required fields are marked *