Monday, October 24, 2011

Week 9: Top 25 -- RBA


Rank +/- Team WinPct SoS Adjusted
Off. Def. Pace
1 -- Alabama 1.000 0.548 17 26.3 4 4.9 1 157.2 120
2 +2 Stanford 0.992 0.541 27 28.9 1 9.2 5 167.1 52
3 -1 LSU 0.992 0.551 13 23.8 9 6.5 2 160.6 111
4 -1 Wisconsin 0.975 0.534 39 28.3 2 10.8 15 162.0 107
5 +1 Oregon 0.966 0.532 43 26.5 3 8.0 3 176.2 3
6 -1 Oklahoma 0.966 0.545 23 24.4 7 9.7 6 171.2 17
7 -- Boise St. 0.950 0.461 93 26.2 6 9.7 7 168.9 32
8 +1 Oklahoma St. 0.941 0.551 12 26.2 5 14.0 42 172.7 9
9 +1 Clemson 0.932 0.526 58 21.3 17 12.9 28 166.7 61
10 NA Southern Miss. 0.930 0.484 76 21.2 18 14.7 49 168.6 36
11 -- South Carolina 0.922 0.542 26 20.9 20 10.3 10 158.0 119
12 -- Arizona St. 0.911 0.530 48 19.5 30 10.6 12 174.0 6
13 -5 TCU 0.911 0.460 95 23.8 8 12.4 24 168.1 42
14 +6 Michigan St. 0.911 0.543 25 20.8 21 10.2 9 167.3 49
15 -- Texas A&M 0.903 0.552 10 20.7 23 12.0 21 171.9 12
16 -3 Arkansas 0.902 0.570 1 22.2 12 12.6 26 165.7 77
17 +2 Florida St. 0.891 0.551 11 20.2 24 11.7 19 166.3 68
18 -2 Georgia 0.885 0.551 14 19.9 26 11.2 17 159.6 114
19 -5 Florida 0.871 0.555 8 21.0 19 8.5 4 161.4 109
20 NA USC 0.867 0.535 38 18.9 37 12.2 22 166.7 62
21 -3 Ohio St. 0.860 0.530 46 18.4 43 9.8 8 159.3 117
22 -5 Notre Dame 0.836 0.566 2 18.7 39 11.9 20 165.0 83
23 -2 Iowa 0.814 0.521 60 19.1 34 13.6 37 164.1 92
24 +1 Nebraska 0.810 0.531 45 21.5 15 13.9 41 166.8 60
25 -3 West Virginia 0.809 0.537 36 19.5 29 13.8 38 165.8 74
Rankings through games of 2011-10-23


New entries: Southern Miss., USC.

Dropped out: Virginia Tech, Auburn.

Two major upsets shake the college football world, and... (4) Wisconsin and (6) Oklahoma drop one place in the RBA polls. Why? RBA looks forward at what will happen tomorrow, not what happened yesterday. Wisconsin and Oklahoma are still very good football teams with strong offenses and defenses. More than likely, they will win the rest of their football games.

Wisconsin's offensive efficiency shifted from 37.8 +/- 18.5 PPH to 38.2 +/- 19.7 PPH, reflecting a very slight reduction in performance against high-end competition. The Badgers hit their target point total of 31 points against the (14) Spartans, so we shouldn't expect their offense to shift very much. The Badgers' defensive efficiency changed from a flat efficiency of 6.3 +/- 6.6 PPH to a steeper 3.7 +/- 14.7 PPH. The outcome is that the Badgers are expected to surrender more points against strong opponents than before, but they didn't suddenly forget how to play defense. Wisconsin remains a strong team, and you should expect to see them in the Rose Bowl, even if they have to play Michigan State again.

Shifting the focus to Oklahoma, the Sooners still rang up 38 points against (39) Texas Tech, and RBA expected them to score 41. Clearly, this doesn't significantly change the offensive model. Last week, RBA expected Oklahoma to have an offensive efficiency of 34.5 +/- 19.5 PPH. This week, RBA expects Oklahoma to perform at 34.3 +/- 19.9 PPH, representing a statistically insignificant shift in performance. On the defensive side of the ball, the Sooners see a slight shift from 3.9 +/- 10.0 PPH to 4.3 +/- 10.8 PPH. The shift is small because Texas Tech represents only one data point. Which is more important? Surrendering 41 to Texas Tech or 13 to Florida State, 28 to Missouri, and 17 to Texas? Oklahoma had an ugly game, but they didn't completely spoil their defense for 2011.

The WTF ranking of the week comes from the addition of (10) Southern Miss. The Golden Eagles were just outside the RBA top 25 last week and debuted in the AP poll this week. The Eagles benefit from a very odd mathematical model for offensive efficiency: 14.8 -/+ 12.9 PPH. (Note the sign flip.) This means that Southern Miss plays up to their opponent on offense, rather than dropping off against top competition like nearly every other team. Unlike (76) Northwestern and (43) Miami-FL, who have slightly positive slopes, Southern Miss has a very positive slope. This statistical anomaly occurs because the Golden Eagles have played a collection of middle-low quality teams and stomped the crap out of them. Since there are so few data points with high strength, the least-squares approximation must assume that their performance against these middling teams extends out to the strong teams, as well. Southern Miss probably isn't #10 good, but until they play somebody strong or stumble offensively, we're left with them here. Remember that the only poll that matters is the last one.

Follow us on Twitter @TFGridiron