So if you've had a chance to check out my College Basketball Q Ratings, you might be wondering how you can use this to make money betting on basketball games. Well, I can't guarantee you'll make any money, but I can say that my rankings correctly predicted the winner in more than 77% of the games last year.
If you just take the moneyline on whichever team is ranked higher in the Q Ratings, you'll do decently well. But for more accurate predictions against the spread, you'll probably want to understand how you can use the different components of the Q Ratings.
Let's look at two teams in the Q Ratings for an example of how to use it.
The Q Ratings are normalized so that you can compare two school's ratings for the predicted margin in a neutral court game played at a D1 average Pace. Most games, however, are not played at a D1 average pace. As you can see here, Villanova had a pace of 156.75 and Michigan had a pace of 136.23. This means that, on average, there a little over 20 more points per game scored in the games Villanova played in than Michigan.
In order to make an accurate prediction about a game played between Villanova and Michigan, first we need to figure out what the projected pace for the game would be. The formula I use to make this estimated Projected Pace or PP:
PP = (Pace * OppPace) / D1AvgPace
Where OppPace is Opponent's Pace and D1AvgPace is the D1 Average Pace (145.99 for 2018).
In this case a game played between Villanova and Michigan would be projected to have a pace of:
PP = (156.75 * 136.23) / 145.99 = 146.27. This means we estimate that 146.27 points will be scored in a game between the two teams.
Next we look at the Difference in QRat. This will give us the normalized Projected Margin or nPM:
nPM = QRat - OppQRat
Where OppQRat is Opponent's Q Rating.
In a game played between Villanova and Michigan, Villanova would have a nPM of:
nPM = 23.64 - 19.01 = 4.63
In order to find a Pace Adjusted Margin for this game, or PM:
PM = nPM * (PP / D1AvgPace)
In a game between Villanova and Michigan, Villanova would have a PM of:
PM = 4.63 * (146.27 / 145.99) = 4.64
In this case, because Nova is a fast paced team and Michigan is a slow paced team, the expected pace is about the D1 average, so the PM and nPM are almost identical. In a game between two fast paced teams or two slow paced teams, you will see a bigger difference between nPM and PM.
PM gives you the projected margin if the game is played on a neutral court. But what if it is played on Michigan's Home Court? In this case you would apply the Home Court Advantage. My model looks at home court margins in conference games to find this adjustment number. For the 2018 season, the final Home Court Margin, or HCM was 3.44. This means if you are finding the projected margin for a team playing on the road, you will subtract 3.44 points from the PM.
So, for a home adjusted Projected Margin or haPM:
Away Team haPM = PM - HCM
Home Team hapM = PM + HCM
In this case Villanova is playing on the road at Michigan so their haPM is:
Villanova haPM = 4.64 - 3.44 = 1.20
Finally, we have a prediction. Villanova would be favored by 1.20 points in a road game against Michigan, according to my model.
If you just take the moneyline on whichever team is ranked higher in the Q Ratings, you'll do decently well. But for more accurate predictions against the spread, you'll probably want to understand how you can use the different components of the Q Ratings.
Let's look at two teams in the Q Ratings for an example of how to use it.
The Q Ratings are normalized so that you can compare two school's ratings for the predicted margin in a neutral court game played at a D1 average Pace. Most games, however, are not played at a D1 average pace. As you can see here, Villanova had a pace of 156.75 and Michigan had a pace of 136.23. This means that, on average, there a little over 20 more points per game scored in the games Villanova played in than Michigan.
In order to make an accurate prediction about a game played between Villanova and Michigan, first we need to figure out what the projected pace for the game would be. The formula I use to make this estimated Projected Pace or PP:
PP = (Pace * OppPace) / D1AvgPace
Where OppPace is Opponent's Pace and D1AvgPace is the D1 Average Pace (145.99 for 2018).
In this case a game played between Villanova and Michigan would be projected to have a pace of:
PP = (156.75 * 136.23) / 145.99 = 146.27. This means we estimate that 146.27 points will be scored in a game between the two teams.
Next we look at the Difference in QRat. This will give us the normalized Projected Margin or nPM:
nPM = QRat - OppQRat
Where OppQRat is Opponent's Q Rating.
In a game played between Villanova and Michigan, Villanova would have a nPM of:
nPM = 23.64 - 19.01 = 4.63
In order to find a Pace Adjusted Margin for this game, or PM:
PM = nPM * (PP / D1AvgPace)
In a game between Villanova and Michigan, Villanova would have a PM of:
PM = 4.63 * (146.27 / 145.99) = 4.64
In this case, because Nova is a fast paced team and Michigan is a slow paced team, the expected pace is about the D1 average, so the PM and nPM are almost identical. In a game between two fast paced teams or two slow paced teams, you will see a bigger difference between nPM and PM.
PM gives you the projected margin if the game is played on a neutral court. But what if it is played on Michigan's Home Court? In this case you would apply the Home Court Advantage. My model looks at home court margins in conference games to find this adjustment number. For the 2018 season, the final Home Court Margin, or HCM was 3.44. This means if you are finding the projected margin for a team playing on the road, you will subtract 3.44 points from the PM.
So, for a home adjusted Projected Margin or haPM:
Away Team haPM = PM - HCM
Home Team hapM = PM + HCM
In this case Villanova is playing on the road at Michigan so their haPM is:
Villanova haPM = 4.64 - 3.44 = 1.20
Finally, we have a prediction. Villanova would be favored by 1.20 points in a road game against Michigan, according to my model.
Comments
Post a Comment
Practice Respect!