NFL - Is home team advantage dropping?

psuasskicker · #1 10-30-2007, 09:57 AM

A friend of mine put this together. Apologize if formatting is off...

...... W - L

2007 64 - 51

2006 136 - 120

2005 151 - 105

2004 145 - 111

2003 157 - 99

2002 148 - 107

2001 136 - 112

2000 138 - 110

1999 146 - 102

1998 151 - 89

1997 145 - 95

1996 149 - 91

1995 144 - 96

Totals 1810 - 1288, 58.4%

Last 5 years 653 - 486, 57.3%

Last 3 years 351 - 276, 56.0%

May not be a huge shift, but seems like there is one. Two questions:
- Any adjustments to be made from this over the coming years?
- What could be causing this? The only thing we could come up with is cushier stadiums and better conditions for visitting teams, but that doesn't seem like it'd cover that much...

- C -

Nostrawho? · #2 10-30-2007, 10:19 AM

Competitiveness of airline industry = lower fares?

Past that I'm just not sure.

Central Limit · #3 10-30-2007, 12:04 PM

Have your friend calculate the following: average margin by which home teams are favored year by year. If that is also dropping, then there is no easy way to take advantage.

kdog · #4 10-30-2007, 04:23 PM

[ QUOTE ]
Totals 1810 - 1288, 58.4%

Last 5 years 653 - 486, 57.3%

Last 3 years 351 - 276, 56.0%

May not be a huge shift, but seems like there is one. Two questions:
- Any adjustments to be made from this over the coming years?
- What could be causing this? The only thing we could come up with is cushier stadiums and better conditions for visitting teams, but that doesn't seem like it'd cover that much...

[/ QUOTE ]

A 2.4% drop is like 5 games more won by visiting teams over the course of the whole season. Or less than one every three weeks. I see no way to create an edge here. It's variance, pure and simple.

knicknut · #5 10-30-2007, 04:28 PM

can you do a t test on this? a binomial testing the last 5 years with a mu of time before that or something?

Central Limit · #6 10-30-2007, 05:10 PM

[ QUOTE ]
can you do a t test on this? a binomial testing the last 5 years with a mu of time before that or something?

[/ QUOTE ]

Yeah, I just did that - comparing 2001 and after to 2000 and before. 2000 and before is lower with p < 0.014. That's a significant result. I think most people want a p-value much lower than this before they're willing to bet large sums of money. I think Stanford Wong wrote a chapter on why this is so, but I can't recall off the top of my head.

I also did a regression through "home winning percentage" :

y = -0.0048x + 0.6174
R2 = 0.393

So, pretty clearly a negative slope.

MacGuyV · #7 10-30-2007, 06:25 PM

I think there's something to the new stadiums angle. Denver hasn't been the same since they moved (I think in 2000?) Of course that was also the beginning of the post-Elway era.

Jazzy3113 · #8 10-30-2007, 06:26 PM

I know you can get totals and spreads for the history of the nfl so why limit your sample size?

Also, the best way to do it would just be to set up a test hypothesis with whatever interval you feel comfortable with, but who is to say what percentage drop is significant?

Personally, I think there are several variables that could throw the findings off. For those errrant games, was a huge away favorite playing, weather, etc? Maybe since there are more teams in the nfl there is not as much parity and the talent pool has become diluted. What I mean is that some teams are finding it hard to compete at an elite level adn this is causing a skew in the home field results. This happens as sports expands to more teams.

With all these types on confounding variables and not using more years to create a larger sample size, I think home field advantage is not losing its luster.

TooTallT · #9 10-30-2007, 06:46 PM

[ QUOTE ]
I know you can get totals and spreads for the history of the nfl so why limit your sample size?

[/ QUOTE ]

1994 was the advent of the 2pt conversion in the NFL, so most of the time stats are only analyzed until then. I guess HFA wouldn't be affected too much by that, but I was too lazy to pull prior to 1994 (I'm the infamous "friend").

J

StuckaRack · #10 10-30-2007, 08:38 PM

How about the AFCs domination of the NFC over the last 3 or 4 years ?

In 2006.......AFC Home Teams were 77-51 60.16%

NFC Home Teams were 59-69 46.09%

In 2006, NFC teams at HOME vs AFC teams were 12-20 37.5%

I have to think this through a little more, but this seems to be the reason why.