Is there a home court advantage in the Patriot League?
We know that the home court is worth about 4 points in the average NCAA game. (See Jeff Sagarin's ratings for example.) This 4 point advantage creates a bigger differential than some might assume since it is either added to a team's expected margin if they are at home or subtracted if they are on the road. For example, if Holy Cross is considered to be three points better than Siena, they would either be a seven point favorite (in theory) at home or a 1-point underdog on the road.
But does this four point home court advantage (which creates an 8-point home/road differential) exist in a league with the type of crowds we see in the Patriot League? Indeed is there any differential at all??
To try to answer this question, I took a look at all Patriot League games for the past three years and looked at each team's home and away margins of victory/defeat and won-loss record. For the three years of games (168 games in all), the short answer is that the home court in the average game was worth 3.7 points (or a differential of 7.4 points between home and road). And the home teams compiled a 102-66 record (.607), as opposed to the .500 record that would be expected if there were no home court bias. That's 18 more wins and 18 less losses than expected. I'm not sure whether this difference is within a likely statistical deviation but it looks pretty meaningful to me.
It gets more interesting when you look at each team's home and road records:
Team Home W-L / Road W-L / Difference
American.....14 - 7 / 11 - 10 / +3
Army.......... 2 - 19 / 3 - 18 / -1
Bucknell.... 21 - 0 / 12 - 9 / +9
Colgate...... 12 - 9 / 5 - 16 / +7
Holy Cross.. 18 - 3 / 13 - 8 / +5
Lafayette.. 12 - 9 / 7 - 14 / +5
Lehigh..... 17 - 4 / 11 - 10 / +6
Navy....... 6 - 15 / 4 - 17 / +2
Total 102 - 66 / 66 - 102 / +36
Thus five PL teams won a lot more at home during this period, led by Bucknell, who has gone an incredible 21-0 at home over this period but nine games worse on the road. The other four with a big edge, in order, are Colgate, Lehigh, Holy Cross, and Lafayette. American and Navy have small edges in the W-L category and Army actually has a worse record, although that is meaningless given their futility in all venues over the past three years.
What may be more interesting is the point differential shown by the various teams home versus road:
------Point Differential at Home------
Team ....... 2006/2005/2004/Average
Bucknell …. 4.71 / 5.21 / 9.07 / 6.33
Holy Cross …. 4.36 / 5.29 / 6.86 / 5.50
Lafayette …. 3.93 / 5.00 / 4.14 / 4.36
Colgate …. 1.93 / 4.43 / 4.93 / 3.76
Navy …. 2.57 / 7.93 / 0.64 / 3.71
American …. 2.21 / 4.86 / 2.57 / 3.21
Lehigh … . -1.57 / 3.07 / 3.14 / 1.55
Army …. -1.86 / 1.79 / 3.64 / 1.19
Total …. 2.04 / 4.70 / 4.38 / 3.70
There are two interesting things that show up here. First, when it comes to the actual point value of the home court, three teams jump to the forefront: Bucknell, followed by Holy Cross and Lafayette. The first two have an edge that is significantly larger than the NCAA average. This seems to make sense in that the Bison in recent years have had the largest, loudest crowds in the PL, while the Crusaders likely are second. LC also has had a tradition of good crowd support. Two teams, Lehigh and Army, are way below average in the point value of the home court. The Army situation is understandable. Lehigh's low number is less understandable, but there are typically a lot of empty seats in Stabler. The perplexing part of the Lehigh record is that they have enjoyed a much better winning percentage at home even though they have not had much of an edge in point differential.
Also of interest, although it may well just be a statistical fluke, is that the value of the PL home court fell significantly in 2006 to 2.04 after it averaged 4.54 points in the two previous years. Indeed, Lehigh and Army actually played better on the road in 2006, at least according to their point differentials.
To add some perspective to these numbers, there were 24 seasons in the study (8 teams x 3). Of those 24 seasons, there was a home court advantage in point differential in 22 of them. And teams had a better record at home 18 times, the same record 3 times, and a worse record only 3 times (twice by Army).
The next question is why is there a significant home court advantage. Some of it might be due to the home crowds but that certainly can't be the only reason. Colgate, for example, had an almost average home court edge and yet plays before mostly empty seats. (Their LARGEST home crowd of last year other than for the Bucknell game was 535 fans. And I suspect that is a made-up, inflated number since every CU home game, including the games during semester break, had virtually the same attendance.) I imagine that a reasonable amount of the edge is the travel effect plus the home team's comfort level with their own court and shooting background. If I get ambitious I may take a look at whether there is any difference when schools are on break or in session.
We know that the home court is worth about 4 points in the average NCAA game. (See Jeff Sagarin's ratings for example.) This 4 point advantage creates a bigger differential than some might assume since it is either added to a team's expected margin if they are at home or subtracted if they are on the road. For example, if Holy Cross is considered to be three points better than Siena, they would either be a seven point favorite (in theory) at home or a 1-point underdog on the road.
But does this four point home court advantage (which creates an 8-point home/road differential) exist in a league with the type of crowds we see in the Patriot League? Indeed is there any differential at all??
To try to answer this question, I took a look at all Patriot League games for the past three years and looked at each team's home and away margins of victory/defeat and won-loss record. For the three years of games (168 games in all), the short answer is that the home court in the average game was worth 3.7 points (or a differential of 7.4 points between home and road). And the home teams compiled a 102-66 record (.607), as opposed to the .500 record that would be expected if there were no home court bias. That's 18 more wins and 18 less losses than expected. I'm not sure whether this difference is within a likely statistical deviation but it looks pretty meaningful to me.
It gets more interesting when you look at each team's home and road records:
Team Home W-L / Road W-L / Difference
American.....14 - 7 / 11 - 10 / +3
Army.......... 2 - 19 / 3 - 18 / -1
Bucknell.... 21 - 0 / 12 - 9 / +9
Colgate...... 12 - 9 / 5 - 16 / +7
Holy Cross.. 18 - 3 / 13 - 8 / +5
Lafayette.. 12 - 9 / 7 - 14 / +5
Lehigh..... 17 - 4 / 11 - 10 / +6
Navy....... 6 - 15 / 4 - 17 / +2
Total 102 - 66 / 66 - 102 / +36
Thus five PL teams won a lot more at home during this period, led by Bucknell, who has gone an incredible 21-0 at home over this period but nine games worse on the road. The other four with a big edge, in order, are Colgate, Lehigh, Holy Cross, and Lafayette. American and Navy have small edges in the W-L category and Army actually has a worse record, although that is meaningless given their futility in all venues over the past three years.
What may be more interesting is the point differential shown by the various teams home versus road:
------Point Differential at Home------
Team ....... 2006/2005/2004/Average
Bucknell …. 4.71 / 5.21 / 9.07 / 6.33
Holy Cross …. 4.36 / 5.29 / 6.86 / 5.50
Lafayette …. 3.93 / 5.00 / 4.14 / 4.36
Colgate …. 1.93 / 4.43 / 4.93 / 3.76
Navy …. 2.57 / 7.93 / 0.64 / 3.71
American …. 2.21 / 4.86 / 2.57 / 3.21
Lehigh … . -1.57 / 3.07 / 3.14 / 1.55
Army …. -1.86 / 1.79 / 3.64 / 1.19
Total …. 2.04 / 4.70 / 4.38 / 3.70
There are two interesting things that show up here. First, when it comes to the actual point value of the home court, three teams jump to the forefront: Bucknell, followed by Holy Cross and Lafayette. The first two have an edge that is significantly larger than the NCAA average. This seems to make sense in that the Bison in recent years have had the largest, loudest crowds in the PL, while the Crusaders likely are second. LC also has had a tradition of good crowd support. Two teams, Lehigh and Army, are way below average in the point value of the home court. The Army situation is understandable. Lehigh's low number is less understandable, but there are typically a lot of empty seats in Stabler. The perplexing part of the Lehigh record is that they have enjoyed a much better winning percentage at home even though they have not had much of an edge in point differential.
Also of interest, although it may well just be a statistical fluke, is that the value of the PL home court fell significantly in 2006 to 2.04 after it averaged 4.54 points in the two previous years. Indeed, Lehigh and Army actually played better on the road in 2006, at least according to their point differentials.
To add some perspective to these numbers, there were 24 seasons in the study (8 teams x 3). Of those 24 seasons, there was a home court advantage in point differential in 22 of them. And teams had a better record at home 18 times, the same record 3 times, and a worse record only 3 times (twice by Army).
The next question is why is there a significant home court advantage. Some of it might be due to the home crowds but that certainly can't be the only reason. Colgate, for example, had an almost average home court edge and yet plays before mostly empty seats. (Their LARGEST home crowd of last year other than for the Bucknell game was 535 fans. And I suspect that is a made-up, inflated number since every CU home game, including the games during semester break, had virtually the same attendance.) I imagine that a reasonable amount of the edge is the travel effect plus the home team's comfort level with their own court and shooting background. If I get ambitious I may take a look at whether there is any difference when schools are on break or in session.
posted by bison137 | 7:57 PM
|
1 comments
