What Home Field Advantage?

OK, some places really are tougher to play at than others, but in general if there's a home field advantage in college football it doesn't show up in the stats.

Testing the hypothesis that the home field is worth some number of points is not easy because the Division 1A field varies so much in team quality. A significant fraction of the games in which the home team wins it is just because the home team is much, much better than a visiting non-conference foe.

To avoid the problem and the associated one caused by non-division 1A opponents, I decided to look at scoring in conference games only, and by conference. Since the total wins and losses are the same for each conference-dependent subset of the games played to date, comparing

should do the trick.

Road vs Home Scoring Summary
10 Nov 2005 11:28:35 (US Central)

V wins WP LP MOV   H wins WP LP MOV V pct   H_M-V_M .5-V Pct
WAC 10 34.3 19.0 15.3   14 42.1 17.0 25.1 0.417   9.8 0.083
SBC 9 22.9 15.4 7.4   9 32.0 16.9 15.1 0.500   7.7 0.000
P10 17 34.9 21.7 13.2   13 42.3 22.2 20.2 0.567   6.9 -0.067
MAC 19 32.3 20.5 11.8   17 37.2 18.8 18.4 0.528   6.6 -0.028
B12 14 34.6 16.1 18.5   21 38.6 16.6 22.0 0.400   3.5 0.100
MW 12 36.7 23.7 13.0   16 35.4 20.2 15.2 0.429   2.2 0.071
BigE 8 31.3 15.4 15.9   10 37.2 19.8 17.4 0.444   1.5 0.056
SEC 15 29.5 15.1 14.4   20 26.1 12.9 13.3 0.429   -1.2 0.071
ACC 16 30.2 14.4 15.8   20 31.1 16.9 14.2 0.444   -1.6 0.056
B10 15 40.0 20.5 19.5   19 34.8 18.9 15.9 0.441   -3.6 0.059
CUSA 16 33.3 17.8 15.4   17 32.6 22.2 10.4 0.485   -5.0 0.015
                         
All 151 33.1 18.7 14.4   176 35.0 18.2 16.9 0.462   2.5 0.038
All w/
MOV<3
14 28.9 27.5 1.4   14 23.1 21.6 1.5 0.500   0.1 0.000

The only places where the two values have the same sign and the magnitude of the MOV difference is enough to matter are the WAC and the Big 12. The WAC is so bad this year I am tempted to ignore that, and the Big 12 numbers may be skewed by the Texas and Texas Tech home winning margins.

  1. In six of the 11 conferences, either the visitor won at least half the games or the average margin of victory for the home winners was lower than that of the road winners.
  2. In the 28 games decided by less than three points, the visitors were .500 and in games won by the home team the MOV average was only a tenth of a point higher than when the visitor won.
There were 64 more games decided by three points, but in those the visitors won 56.25 percent of the time, going 36-28.