Topic: About home field advantage...

[ought-three]

I didn't include home field advantage in my analysis on this season.I looked around and found the following statement at: http://predictionmachine.com/college-football-homefield-advantage:

"The truth is that some stadiums could actually mean up to nine points, while some do not help much at all." We ranked 16th on their list of best home field advantage.

To try this out, I pulled 10 teams at random and their home-away records over the last four years (to allow a few in-conference cycles).

I found that these teams averaged a record of 18-8 at home and 10-12 away over the four-year period. I tested this for statistical significance (one number being higher than the other is not enough) and found that this difference is not considered to be statistically significant (Fisher's Exact test). It creates an effect size of 0.22, which can be considered a moderate effect.

According to the website, and others I've read, the average home field advantage is around 3-4 points, or a possible 6-8 point swing from one venue to the other. I looked at these teams schedules from last year to see how many games were decided by 8 points or less in favor of the home team:

Auburn - 2/12
BYU - 5/12
Ga. Tech - 1/12
SD St. - 1/12
Colorado - 0/12
Cincy - 3/12
Fresno St. - 1/12
Mich. St. - 2/12
Washington - 3/12
Tex Tech - 1/12
TOTAL: 19/120 or about 16% (1 in 6 games)

   Clearly, playing at home has a moderate effect on winning and losing. The fact is, the grand majority of games are decided by more than 8 points and, of those, not all go in favor of the home team.

I found this article from 2002 (http://www.nytimes.com/2002/10/03/sports/college-football-in-real-life-victory-margin-matters.html) that shows the average winning margin to be between 17-20 points, so 8 points isn't going to be a big deal around 80% of the time.

 Is home field advantage a factor in winning? Certainly. Would I consider it among the biggest factors? NO. What I do know is that my system has been below 50% in picking close games and the difference can be attributed to home field advantage swinging the prediction to the team that ended up losing. Therefore, it is out of this fella's spreadsheet.

[Scottyo614]

I think the only way you could make home field advantage work is if you could factor in weather as well. So your southern teams coming north would have issues based off their past. In a data spreadsheet that can't be possible unless you pick from the average temp, record below the temp and about six other numbers lol. That would take forever to do ill bet

[Greg H]

I think the ECU/Marshall series is very strong evidence that home field advantage matters, a lot.

[BHFIOHIO]

Where teams are closely matched the home team will win more of them.

[ought-three]

I think the ECU/Marshall series is very strong evidence that home field advantage matters, a lot.

That implies, as BHF mentioned, that Marshall and ECU have typically been closely matched:
Marshall vs. ECU since 2000 (via the Stassen site)
Home Team (Win/Lose)
2012 ECU (W)
2011 Marshall (W)
2010 ECU(W)
2009 Marshall(L)
2008 ECU (W)
2007 Marshall (W)
2006 ECU (W)
2005 Marshall (L)
Home team record in last 8 games: 6-2 (75%)

But you can't draw a conclusion from one particular instance. I could pick out another pair that is evidence that home field advantage has no bearing whatsoever:

Ohio State vs. Indiana since 2000 (via the Stassen site)
Home Team (Win/Lose)
2012 Indiana (L)
2011 Ohio State (W)
2010 Ohio State (W)
2009 Indiana (L)
2006 Ohio State (W)
2005 Indiana (L)
2004 Ohio State (W)
2003 Indiana (L)
2002 Ohio State (W)
2001 Indiana (L)
Home team record in last 10 games 5-5 (50%)

I'm not saying home field doesn't matter, but just using ECU-Marshall as your evidence doesn't work outside the ECU-Marshall series. You have to analyze a larger and broader set of data to draw a general conclusion. I would encourage you to do more research on the topic.

Again, ECU and MU have been fairly closely matched over the last decade. Ohio State and Indiana have not been closely matched. That suggests there are more important factors out there and that home field only matters in closely-matched contests.