College Football

Total Turnover Margin - corrected for opposition

I was looking at turnover margin as a statistic and thought I should make a version of that statistic that was corrected for a team's opposition. That way for example, playing teams that give up a lot of interceptions wouldn't skew a team's turnover numbers, the same way that scoring a lot, or gaining a lot of yards against weak defenses, makes some teams look better than they are.

So I made a data file comparing turnovers the same way scores are compared for the Strength Power Rating. The numbers are compared and the algorithm runs until the numbers converge to determine which teams have the best Turnover Margin, Turnovers Gained, and Turnovers Lost.

Here's a look at the first run. The full list is on this page.

Rank Team                        MRec    Margin   Gained[rnk]   Lost [rnk]
  1. East Carolina                7-3     1.53    2.78 [  1]    1.25 [ 20]
  2. Rutgers                      8-2     1.43    2.59 [  5]    1.16 [ 17]
  3. Air Force                    6-1     1.23    2.28 [ 15]    1.04 [  9]
  4. Virginia                     5-4     1.18    1.98 [ 39]    0.80 [  3]
  5. Ohio State                   9-1     1.17    2.61 [  4]    1.43 [ 35]
  6. Virginia Tech                6-3     1.17    2.13 [ 24]    0.96 [  4]
  7. Central Florida              7-5     1.10    2.45 [  9]    1.35 [ 29]
  8. Georgia Tech                 4-4     1.06    2.06 [ 30]    0.99 [  7]
  9. Texas                        7-1     1.03    2.71 [  2]    1.67 [ 60]
 10. Southern Miss                8-2     1.02    2.37 [ 11]    1.35 [ 28]

The record, or "MRec" in this case, shows how many games in which a team "won" the turnover battle, and how many times they "lost", simply based on whether they gained or committed more turnovers. You can see there are a lot of ties, which I didn't include in the record.

The "raw" top ten in turnover margin looks like this:

1. Rutgers (+1.67/game)
2. Boise State (+1.46)
3. Air Force (+1.42)
4. Ohio State (+1.33)
5. Alabama (+1.23)
6. Arkansas (+1.08)
7. Ohio (+1.08)
8. East Carolina (+1.00)
9. Texas (+0.92)
10. Middle Tennessee (+0.92)

We see that five of the ten are common to both lists. So while there is some overlap (turnovers are rare events, so the raw and corrected numbers should align quite a bit) some teams are getting too much credit. Rutgers played two Division I-AA teams and was +6 against these teams, which inflates their non-corrected stats. East Carolina played more turnover-saavy teams and their "raw" rating jumped from 1.00 to 1.53 when corrected for this.

The page also lists each team's rank in terms of causing and allowing turnovers. Here are the top five at each:

Rank Team                        MRec     Gained[rnk]   Lost [rnk]
  1. East Carolina                7-3     2.78 [  1]    1.25 [ 20]
  2. Texas                        7-1     2.71 [  2]    1.67 [ 60]
  3. Iowa St                      6-3     2.62 [  3]    1.82 [ 70]
  4. Ohio State                   9-1     2.61 [  4]    1.43 [ 35]
  5. Rutgers                      8-2     2.59 [  5]    1.16 [ 17]

Rank Team                        MRec     Gained[rnk]   Lost [rnk]
  1. Oregon St                    6-1     1.32 [101]    0.59 [  1]
  2. Navy                         7-3     1.60 [ 73]    0.74 [  2]
  3. Virginia                     5-4     1.98 [ 39]    0.79 [  3]
  4. Virginia Tech                6-3     2.13 [ 24]    0.96 [  4]
  5. LSU                          4-2     1.51 [ 78]    0.99 [  5]

So East Carolina's top ranking is mostly from creating turnovers, but they're also good at keeping the ball. Iowa State scores well at causing turnovers (largely from their 8 turnover performance against Nebraska), but they give the ball up a bit as well.

Oregon State is fantastic at preventing turnovers, but they don't steal the ball that often. When they play, look for a game of very few turnovers! In fact the most total turnovers in any game of theirs all season is 3, which is below the average of 3.5 for an average game.

Using these rankings it's possible to project the number of expected turnovers for each team in a game the same way the Strength Power Rating projects scores. A team that forces a lot of turnovers, paired with a team that gives the ball up a lot = lots of turnovers gained for Team A. And since these numbers are corrected for opposition they're better for this purpose than the raw figures.

One things that's interesting to note is there is no "home field advantage" for turnovers. Only score and yardage numbers, the Home Field tracker results in -0.01, which indicates a tiny advantage for the opposing team in turnovers. Of course this number is negligible and is a clearly a random artifact of the data. It's encouraging to note and makes the method appear legitimate, since it seems obvious that turnovers are unaffected by home field. If there had been a home field advantage in turnovers I would have been surprised.

The rankings could be improved in an obvious way. Right now all turnovers are lumped together. This is similar to lumping all yards together when looking at offense and defense. But clearly separating out rushing and passing yards allows for a better comparison: you have rushing offense rankings, rushing defense rankings, passing offense rankings, and passing defense rankings, which is much more helpful than simple offense vs. defense.

In this same vein, it isn't always right to say Team A is going to force a lot of turnovers on Team B in the case where Team A has dozens of interceptions, and Team B fumbles the ball but throws no interceptions. In most cases there isn't much difference, since a fumbling team will still fumble, and a team that intercepts a lot may get some (but fewer) interceptions against a good passer. But it would be better to compare apples to apples, oranges to oranges, and then add the results, than lump them all together.

This of course would require two data files (and twice as much work!) and the resultant numbers would all be very small (0 to 2 in most cases) so the significance might be lost. I might try this next year, but for now the corrected Total turnovers chart will have to do.

Week 3 Power Rating: Miami #1, FSU #2

Is this 2009? or 1989?

With Miami #1 and Florida State #2, it looks like a throwback to the late 80s and early 90s when these teams dominated the college football scene for a solid decade. Miami's solid win over Georgia Tech combined with FSU's 54-28 upset of BYU pushed these teams to the top of the ratings, with the Hurricanes' win over the Seminoles giving them the nod at #1. Still very early, this is just the first look at an organized pecking order for the 2009 season, but it's an interesting one.

I'm still referring to the Cross-division power rating which uses all teams, since the I-A vs. I-AA games are crucial for having enough data to go by. Here are the top 25 teams as of Saturday:

  #  Team                         rec   Rating  Offense[rnk]  Defense[rnk]
  1. Miami FL                     2-0    91.69   68.94 [  8]  -22.74 [  1]
  2. Florida St                   2-1    84.92   75.79 [  1]   -9.12 [ 22]
  3. Cincinnati                   3-0    82.04   74.18 [  3]   -7.86 [ 27]
  4. Alabama                      3-0    81.33   68.48 [  9]  -12.85 [ 10]
  5. Mississippi                  2-0    80.98   74.72 [  2]   -6.26 [ 36]
  6. Georgia Tech                 2-1    78.47   71.96 [  5]   -6.50 [ 33]
  7. Virginia Tech                2-1    75.28   65.31 [ 12]   -9.97 [ 20]
  8. Northern Iowa                2-1    74.33   60.36 [ 18]  -13.96 [  7]
  9. Nebraska                     2-1    71.50   51.04 [ 53]  -20.46 [  2]
 10. Florida                      3-0    71.39   61.30 [ 16]  -10.09 [ 18]
 11. California                   3-0    70.53   71.24 [  6]    0.70 [ 68]
 12. Clemson                      2-1    70.49   54.91 [ 31]  -15.58 [  5]
 13. Oklahoma                     2-1    70.36   52.94 [ 39]  -17.42 [  4]
 14. Iowa                         3-0    69.47   58.43 [ 20]  -11.04 [ 13]
 15. Jacksonville St              1-2    68.51   50.70 [ 56]  -17.81 [  3]
 16. Southern Cal                 2-1    67.94   54.48 [ 35]  -13.46 [  8]
 17. Texas                        3-0    66.88   66.44 [ 10]   -0.43 [ 61]
 18. LSU                          3-0    66.05   55.78 [ 29]  -10.26 [ 16]
 19. Tennessee                    1-2    65.13   55.62 [ 30]   -9.51 [ 21]
 20. Brigham Young                2-1    64.85   60.77 [ 17]   -4.08 [ 41]
 21. Kansas                       3-0    64.71   62.13 [ 15]   -2.57 [ 49]
 22. Ohio State                   2-1    63.03   49.93 [ 61]  -13.10 [  9]
 23. UCLA                         3-0    62.83   51.22 [ 50]  -11.60 [ 11]
 24. Boise St                     3-0    61.67   46.58 [100]  -15.09 [  6]
 25. Auburn                       3-0    61.54   65.24 [ 13]    3.69 [ 89]
 26. Washington                   2-1    61.36   59.19 [ 19]   -2.17 [ 52]

Cincinnati, Alabama, and Mississippi round out the early top five due to their high victory margins; #7 Virginia Tech lost only to #4 Alabama, and beat #9 Nebraska. Florida, the consensus pre-season and current #1, ranks only #10 after an unimpressive performance against Tennessee, but it was enough to push the 1-2 Volunteers up to #19 (even ahead of UCLA, who beat them).

BYU's big loss dropped them below Oklahoma, who they beat in week one. Another big upset, Washington's win over Southern Cal, has the Trojans at #16 and the Huskies up to #26, with Ohio State (who lost to USC) in between.

A couple of FCS (I-AA) teams are currently in the top 25.First Northern Iowa sits at #8; they lost to Iowa in Iowa City by a single point, then beat two I-AA foes by a combined 96-7. The Panthers could very well be the best team in the FCS this year, as their defense is likely a true top 25 defense in any division based on these early games. Jacksonville State is at #15 despite being 1-2 (remember, winning and losing isn't considered in this score-margin-based power rating). The Gamecocks lost at Florida State just 19-9 a week after falling to Georgia Tech 37-17. Then on Saturday they won 45-13 against 2-1 Alabama A&M. Again it's their defense that looks to be their strong suit, and the offense is led by ex-LSU QB Ryan Perrilloux. I-AA teams almost never finish in the top 25 in power ratings and with more games and information both these teams will slide, but they're interesting cases and both teams should be in the title hunt come playoff time.

Just for comparison's sake, here are the top ten teams from (roughly) the same time in 2008:

  1. Monmouth IL                  2-0    82.72   56.06 [ 25]  -26.66 [  1]
  2. Kansas St                    2-0    70.50   62.88 [ 10]   -7.62 [ 15]
  3. Loras                        1-1    69.41   65.18 [  7]   -4.23 [ 26]
  4. Oklahoma                     3-0    69.09   76.25 [  1]    7.16 [ 97]
  5. Southern Cal                 2-0    68.87   62.31 [ 11]   -6.55 [ 18]
  6. Penn State                   3-0    66.86   67.38 [  4]    0.51 [ 44]
  7. Florida                      2-0    63.40   63.94 [  9]    0.53 [ 45]
  8. Florida St                   2-0    63.00   69.35 [  2]    6.34 [ 86]
  9. LSU                          2-0    62.77   49.17 [ 52]  -13.59 [  6]
 10. TCU                          3-0    62.08   48.91 [ 53]  -13.17 [  7]

Obviously the top teams were flukes, but the 2008 season wasn't as far along in week three last year (it was September 13th, 2008). This year the non-division I-A schedule started earlier and almost all I-A teams have played three games. But even at this point in 2008, five of the teams in the top ten in week three finished in the top ten at the end of the season. By week four, seven of the ten were present in the top ten. So look at the 2009 list (substitute Cal for UNI)—which of the top ten will still be there come January? And which teams outside the top ten will move in?

Week 2 Power Rating: FBS/FCS contests muddle results

After two weeks of action the power ratings aren't as random as they are (of course) after one game; already some reasonable ordering is starting to take place. But without pre-conceived notions about the teams or use of information left over from last year, the numbers guide the algorithm in some interesting directions.

I've only updated the cross-division power rating so far; most I-A (FBS) teams have played a I-AA (FBS) team so the I-A only rating isn't nearly as interesting. Here are the top 25 teams from the all-division power rating (click here for the full list).

  #  Team                         rec   Rating  Offense[rnk]  Defense[rnk]
  1. Texas                        2-0   114.49  105.07 [  1]   -9.41 [ 34]
  2. Boise St                     2-0   104.01   68.31 [ 25]  -35.69 [  4]
  3. California                   2-0   101.97   75.84 [ 12]  -26.12 [  9]
  4. Kentucky                     1-0   100.89   63.07 [ 41]  -37.81 [  3]
  5. Purdue                       1-1    96.80   88.21 [  6]   -8.58 [ 38]
  6. Oregon                       1-1    95.86   72.32 [ 16]  -23.54 [ 13]
  7. Northern Iowa                1-1    83.71   69.85 [ 20]  -13.85 [ 27]
  8. Iowa                         2-0    81.35   57.38 [ 64]  -23.96 [ 11]
  9. Wyoming                      1-1    80.53   45.14 [122]  -35.39 [  5]
 10. Cincinnati                   2-0    80.47   62.14 [ 46]  -18.33 [ 21]
 11. Toledo                       1-1    78.67   68.22 [ 26]  -10.44 [ 33]
 12. LSU                          2-0    78.50   28.91 [292]  -49.58 [  2]
 13. Louisiana-Monroe             1-1    78.37   58.06 [ 61]  -20.31 [ 20]
 14. Pittsburgh                   2-0    77.82   65.47 [ 33]  -12.35 [ 30]
 15. Weber St                     0-2    76.43   86.04 [  8]    9.60 [128]
 16. Wisconsin                    2-0    75.85   52.18 [ 87]  -23.67 [ 12]
 17. UCLA                         2-0    75.62   53.22 [ 79]  -22.39 [ 15]
 18. Fresno St                    1-1    75.30   82.08 [  9]    6.78 [109]
 19. Arizona                      2-0    74.85   53.74 [ 74]  -21.10 [ 18]
 20. Colorado St                  2-0    74.48   40.12 [165]  -34.36 [  6]
 21. Houston                      2-0    74.33   90.95 [  4]   16.62 [198]
 22. Northern Illinois            1-1    71.20   73.55 [ 13]    2.34 [ 77]
 23. Tennessee                    1-1    68.28   62.91 [ 43]   -5.36 [ 46]
 24. Washington                   1-1    67.59   98.32 [  3]   30.73 [428]
 25. Vanderbilt                   1-1    67.45   87.26 [  7]   19.81 [245]

Texas at #1 is not so unreasonable; they're at the top because they won by a lot of points (70 points over 2 games) and more importantly, the teams they beat are 1-1. If Wyoming and ULM had lost their other game, the Longhorns might be #100 rather than #1. In fact this is what happened to Florida, who sits at #75 because their wins over Charleston Southern and Troy, both 0-2, don't push them up. But since Wyoming beat Weber State and ULM beat Texas Southern 58-0, all three teams are in the top 25. Even Weber State at 0-2 is #15. Why? Because they only lost to Wyoming 28-22, of course! Their other game is a 24-23 road loss to #20 Colorado State. And why is Colorado State #20? Mostly for beating Weber State. With just two games to go by every bit of circular logic applies, sending teams up or down.

Boise State is #2, with wins over #6 Oregon and a 48-0 win over Miami of Ohio. #4 Kentucky played one game, beating the same Miami team 42-0, so they rank just a bit behind the Broncos. Purdue is at #5 despite losing to the Ducks; as it was a 2 point road loss they're covered by home field advantage. The Boilermakers also beat Toledo (#11), who remains high due to their win over Colorado (#29 despite their 0-2 record).

#7 and #8 are interesting. In week one Iowa escaped with a 17-16 home win over Northern Iowa. The next week both teams dominated, the Hawkeyes beating Iowa State on the road 35-3 and UNI pasting South Dakota 66-7. Add to that the fact that both of those victims are 1-1, with the Cyclones having dominated North Dakota State 34-17 and South Dakota having destroyed William Penn 52-0, and it's easier to see why the algorithm is high on these teams, for now. The correction for home field advantage puts UNI on top.

California and Cincinnati are ranked high almost solely for margin of victory. The Bears have defeated their two opponents 111-20 and the Bearcats have won 117-18. It doesn't hurt that all four of their victims are 1-1. Cincinnati's foes did particularly well, with Rutgers beating Howard 45-7 and SE Missouri beating Quincy 72-3. Yet after the algorithm churns away the Bears are #3 and the Bearcats #10.

It's interesting that even this early I-A teams are dominating the top 25; this isn't the case in week one. But with I-A teams winning the overwhelming majority of cross-division matchups it makes sense. Over the next few weeks new information will correct the obvious deficiencies in the ratings. At this time teams that have played a tough schedule generally sit in the middle of the ranks, as they may have only won by 10 points rather than 50. But as their opponents grow in stature so too will those wins.

Some examples: USC is #144 and Ohio State #188; Alabama is #154 and Virginia Tech is #216. Michigan is #223 and Notre Dame is #222.  Even by next week all of those teams should be in the top 100 if not the top 25 as more I-A teams play each other and strength of schedule naturally becomes dominant over margin of victory.

Best defense of 2008: Where does USC rank all-time?

The 2008 USC Trojans were dominant on defense: they shut out three of their opponents, held three more to a field goal, two more to a touchdown, and two more to just 10 points. One of those held to a field goal was Ohio State; Oregon, which finished with the #3 scoring offense in 2008, was held to 10 points.

So where does USC rank all-time? Before the Rose Bowl the commentators remarked that USC was, at that point, the all-time best when their average points allowed was subtracted from the average game score for the year. I'll look at that measure, adjusted for competition of course, in a bit.

First, here are the top ten defenses of 2008:

Top defenses, 2008 football season
#   Team                       record   adj. ave      
 1. Southern Cal                12-1     6.71 [  1]
 2. Florida                     13-1     8.33 [  2]
 3. TCU                         11-2     9.87 [  3]
 4. Ohio State                  10-3    11.88 [  4]
 5. Texas                       12-1    11.89 [  5]
 6. Penn State                  11-2    13.29 [  6]
 7. Iowa                         9-4    13.42 [  7]
 8. Alabama                     12-2    14.70 [  8]
 9. Boise St                    12-1    14.80 [  9]
10. Wake Forest                  8-5    16.37 [ 10]

USC allowed just 6.7 points per game, adjusted for their competition. Their raw average was exactly 9 points per game, up from 7.75 before the bowl game. But they played a tougher-than-average schedule, as does almost every BCS conference team, especially one that goes to a BCS bowl game. Their adjusted average stayed almost exactly the same after the Rose Bowl as holding Penn State to 24 isn't bad at all.

Florida's strong showing against top offense Oklahoma boosted them to #2. #3 TCU held Boise State to 16 points and #4 Ohio State held Texas to 24 points, their lowest total of the year.

#5 Texas and #6 Penn State had solid defenses to go with their offenses; the Nittany Lions' performance against USC didn't affect them much as USC has a top ten offense. #7 Iowa and #10 Wake Forest built their winning seasons on defense as they didn't have nearly as much to offer on offense, while Alabama and Boise State had more but were still defense-oriented teams.

In terms of yardage, TCU ranked slightly better than USC, mainly due to their extremely stingy rushing defense.

Rank Team                         Rec   Total[rank] Rush[rank]  Pass[rank]
  1. TCU                         11-2   201.9[  1]   43.6[  1]  158.2[  5]
  2. Southern Cal                12-1   224.9[  2]   85.8[  6]  139.0[  1]
  3. Florida                     13-1   267.1[  3]   99.2[ 15]  167.9[  8]
  4. Alabama                     12-2   270.5[  4]   71.5[  2]  198.9[ 35]
  5. South Carolina               7-6   284.8[  5]  109.9[ 20]  174.8[ 10]
  6. Tennessee                    5-7   285.4[  6]   99.0[ 14]  186.4[ 21]
  7. Connecticut                  8-5   288.3[  7]  112.3[ 23]  176.0[ 12]
  8. Utah                        13-0   288.7[  8]   91.1[ 11]  197.6[ 34]
  9. Texas                       12-1   292.9[  9]   82.9[  5]  210.0[ 51]
 10. Ohio State                  10-3   293.0[ 10]  112.5[ 25]  180.4[ 14]

The Trojans were second and Florida third, with many of the teams from the scoring defense list joined by South Carolina, Tennessee, and Connecticut.

Back to scoring defense, let's look at how USC ranks for the decade:

Top defenses 2000-2008 seasons (uncorrected for year played)
#   Year Team                  record   adj. ave      
 1. 2000 Florida State           11-2    6.54 [  1]
 2. 2008 Southern Cal            12-1    6.71 [  1]
 3. 2005 Ohio State              10-2    7.26 [  1]
 4. 2001 Miami (Florida)         12-0    7.54 [  1]
 5. 2003 Louisiana State         13-1    7.74 [  1]
 6. 2008 Florida                 13-1    8.33 [  2]
 7. 2004 Southern Cal            13-0    8.80 [  1]
 8. 2002 Kansas State            11-2    9.27 [  1]
 9. 2000 Oklahoma                13-0    9.56 [  2]
10. 2002 Ohio State              14-0    9.61 [  2]

USC ranks #2 among teams from this decade, with Florida coming in at #6. Florida State's 2000 team that lost to Oklahoma in the national championship game. Their raw average was just over 10 points, but they held 11 of their 13 opponents to 14 points or fewer, and faced many of the year's best offenses. Note that two of USC's shutouts were Washington and Washington State and that's what probably gives 2000 FSU the edge.

Let's see what happens if we include all the teams from the 80s and 90s as well:

Top defenses 1980-2008 seasons (uncorrected for year played)
#   Year Team                   record   adj. ave     
 1. 1985 Michigan                10-1-1  1.87 [  1]
 2. 1980 Florida State           10-2    3.13 [  1]
 3. 1991 Miami FL                12-0    3.87 [  1]
 4. 1980 Alabama                 10-2    4.16 [  2]
 5. 1980 Nebraska                10-2    4.17 [  3]
 6. 1991 Washington              12-0    4.56 [  2]
 7. 1983 Miami FL                11-1    4.67 [  1]
 8. 1980 Southern California      8-2-1  4.80 [  4]
 9. 1986 Oklahoma                11-1    4.81 [  1]
10. 1988 Miami FL                11-1    4.84 [  1]

...and the 2008 Trojans have disappeared from the list ; as have the Seminoles and everyone else from this decade. The 1980 Trojans are on the list, though; apparently, 1980 was a really good year for defense, with four teams on the list. Or could that be a result of lower scoring in those days? To correct for rule changes and other things that can boost or quell scoring, we subtract the year's average game score from the team's adjusted average defensive score.

Top offenses 1980-2008 seasons (standardized for year played)
#   year Team              rec    adj. ave    year's ave   difference
1.  1985 Michigan          10-1-1  1.87 [  1]   22.26        20.39
2.  2008 Southern Cal      12-1    6.71 [  1]   26.34        19.63
3.  2000 Florida State     11-2    6.54 [  1]   25.74        19.20
4.  1991 Miami FL          12-0    3.87 [  1]   23.04        19.17
5t  2001 Miami FL          12-0    7.54 [  1]   26.56        19.02
5t  2005 Ohio State        10-2    7.26 [  1]   26.28        19.02
7.  1996 Ohio State        11-1    6.60 [  1]   25.46        18.86
8.  1988 Miami FL          11-1    4.84 [  1]   23.66        18.82
9.  2003 Louisiana State   13-1    7.74 [  1]   26.24        18.50
10. 1991 Washington        12-0    4.56 [  2]   23.04        18.48

This gives a much more balanced view of the relative performance across decades. Michigan's 1985 team still has the best season. In a year when the average score was over 22 points, the Wolverines allowed an incredible 1.87 points when adjusted for competition. They allowed 8.2 points per game in raw numbers, but that included playing Nebraska, Ohio State, and the year's number one offense, Iowa, whom they held to 12 points. Beyond that they had three shutouts and seven opponents held to single digits.

This year's USC places second, a bit ahead of 2000 FSU; the average score this year was about 0.6 points higher than in 2000, pushing USC above the Seminoles as the best defense of the decade. Two undefeated Miami teams, from 1991 and 2001, follow at 4th and 5th, with the latter tied with Ohio State's 2005 squad. Another Buckeye team follows, and then a third Hurricane defense at #8. LSU's BCS title team from 2003 and Washington—the second undefeated team from 1991—rounds out the top ten.

So USC can fairly argue they had the best defense in college football this decade, and over the last 20+ years. If you look from the 1940s on, or even more severely, the 1870s on, the defensive numbers almost become meaningless, with negative averages due to the best teams racking up very high score and garnering 10 or more shutouts, so we'll have to stick the the last few decades.

Despite loss, Oklahoma's offense still #1 in 2008 in scoring and yardage

Despite scoring just 14 points against Florida in a loss to the Gators, the 2008 Oklahoma Sooners finished the season with the #1 scoring offense of the year. The performance did lower their averages—both raw and adjusted—quite a bit, but not enough to drop them below #2 Florida, who had 24 points which was below their average as well.

Oklahoma's relatively poor performance also lowered their standing in the all-time offensive rankings which I posted just before the game. They also remained #1 in adjusted total yardage on offense.

Here are the final top ten offenses of 2008:

Top scoring offenses, 2008 football season
 #  Team                       record   adj. ave      
 1. Oklahoma                    12-2    54.47  
 2. Florida                     13-1    46.38
 3. Oregon                      10-3    44.66
 4. Texas                       12-1    42.91
 5. Texas Tech                  11-2    42.91
 6. Missouri                    10-4    42.71
 7. Tulsa                       11-3    41.58
 8. Penn State                  11-2    40.79
 9. Southern Cal                12-1    40.65
10. Oklahoma St                  9-4    39.45

Before the BCS title game, Oklahoma's adjusted average scoring output was 56.56, so they lost over two points after the loss. But they're still #1 for the year by over 8 points per game when opposition is taken into account. Florida finished second on offense, losing over a full point after facing the Sooners. The expected score of that game, using the adjusted averages for offense and defense, was about 41-40, so both teams fell short of expectations. (Or their defenses exceeded expectations.)

Oregon finished the year #3, followed by three high-power Big Twelve passing offenses: Texas, Texas Tech, and Missouri. Tulsa's strong passing and rushing attack put them at #7, then Penn State, USC, and Oklahoma State fill out the top ten. Three of the top ten offenses in the country were from Oklahoma teams.

In terms of total yardage generated on offense, Oklahoma is #1 as well.

Rank Team                         Rec   Total(rank) Rush(rank)  Pass(rank)
  1. Oklahoma                    12-2   564.0[  1]  210.3[ 16]  353.7[  3]
  2. Tulsa                       11-3   538.8[  2]  240.2[  8]  298.6[  9]
  3. Houston                      8-5   538.6[  3]  144.3[ 64]  394.2[  2]
  4. Texas Tech                  11-2   517.0[  4]  120.0[ 92]  397.0[  1]
  5. Oregon                      10-3   497.5[  5]  267.6[  3]  229.8[ 48]
  6. Nevada                       7-6   479.2[  6]  241.4[  6]  237.8[ 39]
  7. Missouri                    10-4   471.6[  7]  152.2[ 57]  319.4[  5]
  8. Oklahoma St                  9-4   471.3[  8]  225.0[ 12]  246.3[ 34]
  9. Southern Cal                12-1   470.4[  9]  196.2[ 25]  274.2[ 18]
 10. Texas                       12-1   467.0[ 10]  162.8[ 44]  304.1[  7]

The Sooners were 16th in rushing yards and 3rd in passing yards (all figured adjusted for competition) and overall were #1 by a wide margin. Tulsa and Houston were neck and neck for #2, with the Golden Hurricane finishing in the top 10 in both categories. Texas Tech's pass-oriented game puts them at #4.

Oregon and Nevada were both top ten in rushing with an above-average passing game that puts them at #5 and #6; Missouri is #7, mostly via the air, while Oklahoma State at #8 and USC at #10 are more balanced. #10 Texas is as balanced as Oregon and Nevada, though in the other direction.

By yardage, three of the top ten teams are from Oklahoma, and three more were from Texas.

Back to scoring: how did Oklahoma finish among the all-time comparisons? Let's go through them and see how much they dropped:

Top offenses 2000-2008 seasons (uncorrect for year played)
#   Year Team                  record   adj. ave      
1.  2008 Oklahoma               12-1    54.47 [1]  
2.  2005 Texas                  13-0    53.70 [1]
3.  2005 Southern California    12-1    52.48 [2]
4.  2001 Florida                10-2    47.91 [1] 
9.  2007 Florida                 9-4    47.36 [1] 
6.  2004 Utah                   12-0    46.91 [1] 
7.  2004 Louisville             11-1    46.81 [2] 
8.  2001 Miami (Florida)        12-0    46.80 [2] 
5.  2008 Florida                12-1    46.48 [2] 
10. 2000 Nebraska               10-2    45.26 [1] 

Oklahoma still comes out on top but not by a wide margin any more over 2005 Texas. This year's Florida squad dropped from #5 to #9, placing them lower than last year's team.

Now for the 80s, 90s, and 00s:

Top offenses 1980-2008 seasons (uncorrected for year played)
#   Year Team                       record   adj. ave     
1.  2008 Oklahoma                    12-1    54.47 [  1]   
2.  1995 Nebraska                    12-0    54.30 [  1]   
3.  1989 Houston                      9-2    54.28 [  1]   
4.  2005 Texas                       13-0    53.70 [  1]   
5.  2005 Southern California         12-1    52.48 [  2]  

That's really cutting it close. Oklahoma still ranks better than 1995 Nebraska but just by a hair. Now we look at the "corrected" figures that account for the increase in overall scoring average.

Top offenses 1980-2008 seasons (standardized for year played)
#   year Team              rec    adj. ave    year's ave     difference
1.  1989 Houston            9-2   54.28 [1]  -  24.13     =    30.15
2.  1983 Nebraska          12-1   51.47 [1]  -  21.93     =    29.54
3.  1995 Nebraska          12-0   54.30 [1]  -  25.20     =    29.10
4.  2008 Oklahoma          12-1   54.47 [1]  -  26.34     =    28.13
5.  2005 Texas             13-0   53.70 [1]  -  26.28     =    27.42

Oklahoma can no longer claim to have the best offense in the last 30 years, let along the last 50 or 60. The Sooners had finished first on this list just last week, and even #6 from 1940 on. But for the last three decades, they fall behind 1989 Houston and two powerhouse Nebraska teams, including the one from 1995. So the "best offense in the last dozen or so years" is the best that can be reasonably argued. Going back to 1940, 2008 Oklahoma no longer appears in the top ten.

So being the best offense of the 2008 season is going to have to be enough.

Are the 2008 Florida Gators really one of the best teams in college football history?

Urban Meyer believes—and columnist Tom Dienhart agrees—that the 2008 Florida Gators should go down as one of the best teams in college football history.

Can a team with a loss be considered one of the best teams of all time? Dienhart's article is scant on supportive arguments, the basic idea put forth is that Florida played 11 bowl team this season. And they did beat 10 of them, largely because nearly the entire SEC went to a bowl. Here are their bowl team wins, in order of most to least impressive:

• = #5 Oklahoma (12-2) 24-14
• = #6 Alabama (12-2) 31-20
• = #14 Georgia (10-3) 49-10
  
• @ Florida State (9-4) 45-15
• LSU (8-5) 51-21
• Kentucky (7-6) 63-5
• @ Vanderbilt (7-6) 42-14 
• Miami FL (7-6) 26-3
• South Carolina (7-6) 56-6
• Hawaii (7-7) 56-10

Though Florida's schedule was solid, 68 out of 120 teams going to bowls so it isn't as impressive as it sounds. They only beat three top-notch teams, only two of the ten wins were strictly road games, and five of the ten were 7-6 or worse. Does that offset losing a game at home to 9-4 Mississippi?

In terms of magnitude, a majority of their wins were impressive. They beat eight bowl teams by 23 points or more, four of them by 45 or more. This was part of Meyer's argument: "Look at their scores, look at the offense, look at the balance, look at the defense, and it's all relative. That offense … nine of our opponents finished in the top 25 in defense. A lot of people overlook that. Same thing defensively. Our defense finished in the top five in America playing the schedule we play. … They've got my vote as one of the best teams in history."

It's true, 2008 Florida was a balanced team, finishing #2 in adjusted scoring offense and #2 in adjusted scoring defense. They indeed finish #1 in score-margin based power ratings, but some consideration must be given for winning, as they finish #2 to unbeaten Utah in most power ratings where winning vs. losing is paramount.

We can probably assume that any modern team that wins the national championship is one of the best in history in real terms; with bigger, stronger players, you can argue that a modern champ could beat any team at least through the 60s, and maybe through the 80s or even the 90s. But looked at in relative terms, the argument gets tougher. Setting aside that question, I looked at my power ratings to compare 2008 Florida just to the teams of this decade: how do they rank among teams from 2000-2008?

Here are the top 25 teams from 2000-2008 in my Overall power rating, which combines the Strength rating (score margin-based) and Success rating (win-loss based) to output the teams that have been the most dominant during individual games as well as the entire season. The numbers in brackets are the team's rank within their own year.

rnk   Yr Team                   record    Overall    Strength    Success
  1. '04 Southern California     13-0    68.22 [ 1]  36.02 [ 1]  32.19 [ 1]
  2. '05 Texas                   13-0    68.21 [ 1]  40.99 [ 1]  27.22 [ 1]
  3. '01 Miami FL                12-0    60.63 [ 1]  39.26 [ 1]  21.36 [ 1]
  4. '08 Florida                 13-1    58.46 [ 1]  38.05 [ 1]  20.40 [ 2]
  5. '00 Oklahoma                13-0    57.15 [ 1]  29.91 [ 3]  27.23 [ 1]
  6. '00 Miami FL                11-1    49.55 [ 2]  33.69 [ 2]  15.85 [ 3]
  7. '02 Ohio State              14-0    49.24 [ 1]  19.70 [ 9]  29.54 [ 1]
  8. '04 Auburn                  13-0    48.33 [ 2]  22.94 [ 7]  25.38 [ 2]
  9. '05 Southern California     12-1    48.00 [ 2]  37.31 [ 2]  10.69 [ 3]
 10. '08 Oklahoma                12-2    47.28 [ 2]  36.15 [ 2]  11.13 [ 5]
 11. '01 Florida                 10-2    46.99 [ 2]  35.33 [ 2]  11.65 [ 3]
 12. '06 Florida                 13-1    46.12 [ 1]  22.49 [ 5]  23.62 [ 1]
 13. '08 Southern California     12-1    45.48 [ 3]  33.93 [ 3]  11.54 [ 4]
 14. '03 Louisiana State         13-1    44.87 [ 1]  29.10 [ 2]  15.76 [ 1]
 15. '08 Texas                   12-1    44.86 [ 4]  31.02 [ 4]  13.84 [ 3]
 16. '00 Florida State           11-2    44.23 [ 3]  35.36 [ 1]   8.87 [ 6]
 17. '02 Southern California     11-2    41.70 [ 2]  30.91 [ 2]  10.78 [ 5]
 18. '02 Kansas State            11-2    41.13 [ 3]  33.43 [ 1]   7.70 [ 7]
 19. '04 Oklahoma                12-1    39.93 [ 3]  25.64 [ 5]  14.29 [ 3]
 20. '08 Utah                    13-0    39.93 [ 5]  19.04 [10]  20.89 [ 1]
 21. '02 Oklahoma                12-2    39.79 [ 4]  27.59 [ 3]  12.19 [ 4]
 22. '00 Washington              11-1    39.65 [ 4]  19.41 [ 9]  20.24 [ 2]
 23. '02 Miami FL                12-1    39.20 [ 5]  26.02 [ 4]  13.18 [ 2]
 24. '04 Utah                    12-0    38.89 [ 4]  25.92 [ 4]  12.97 [ 4]
 25. '06 Southern California     11-2    38.74 [ 2]  25.22 [ 2]  13.52 [ 4]

This season's Florida squad really does rate pretty well, finishing fourth among the teams of this decade. If you consider the past nine years to essentially encapsulate the best teams of all time, then Meyer is right. The teams of the 70s, 80s, and 90s would argue that assumption, of course.

The only teams that rank ahead of Florida are USC from 2004, Texas from 2005, and 2001 Miami, all undefeated national champions; the 2008 Gators rank ahead of undefeated champs 2000 Oklahoma (#5) and 2002 Ohio State (#7), and well ahead of LSU's two champs from 2003 (#14) and 2007 (not in the top 25). They also rank ahead of the 2006 Gators that went a similar 13-1 and beat Ohio State in the final two years ago. The 2001 team that went 10-2 is also ranked ahead of the 2006 team, as they rate much better in terms of strength.

USC dominates the chart with five teams listed including the #1 and this season's team at #13. Oklahoma doesn't do bad for itself with four teams including the 2008 squad at #10. Miami has three teams listed, all from 2000 to 2002, a champion and two that finished second in the polls.

2008 was a good year for college football, with five teams making the list. Along with Florida, Oklahoma, and USC, Texas came in at #15 and Utah at #20. Both the Longhorns and the Utes had two teams in the top 25, Texas well below their 2005 team, and Utah just barely above the 2004 team, who ranked much better in Strength but whose schedule in their 12-0 season didn't give them nearly as many points in Success as the 2008 team garnered. No teams from 2007 made the list.

Now let's put aside Florida's loss to Mississippi, in a way. The Strength component of the power rating does not care if a team won or lost a game, it only cares about the score of games. So losing a game per sé does not matter at all, except that you won't have the most dominant score margins if you go around losing games. In any case, here are the best teams of the decade regardless of whether they lost a fluke game here or there:

rnk   Yr Team                   record    Overall    Strength    Success
  1. '05 Texas                   13-0    68.21 [ 1]  40.99 [ 1]  27.22 [ 1]
  2. '01 Miami FL                12-0    60.63 [ 1]  39.26 [ 1]  21.36 [ 1]
  3. '08 Florida                 13-1    58.46 [ 1]  38.05 [ 1]  20.40 [ 2]
  4. '05 Southern California     12-1    48.00 [ 2]  37.31 [ 2]  10.69 [ 3]
  5. '04 Southern California     13-0    68.22 [ 1]  36.02 [ 1]  32.19 [ 1]
  6. '08 Oklahoma                12-2    47.28 [ 2]  36.15 [ 2]  11.13 [ 5]
  7. '00 Florida State           11-2    44.23 [ 3]  35.36 [ 1]   8.87 [ 6]
  8. '01 Florida                 10-2    46.99 [ 2]  35.33 [ 2]  11.65 [ 3]
  9. '08 Southern California     12-1    45.48 [ 3]  33.93 [ 3]  11.54 [ 4]
 10. '00 Miami FL                11-1    49.55 [ 2]  33.69 [ 2]  15.85 [ 3]
 11. '02 Kansas State            11-2    41.13 [ 3]  33.43 [ 1]   7.70 [ 7]
 12. '02 Southern California     11-2    41.70 [ 2]  30.91 [ 2]  10.78 [ 5]
 13. '03 Oklahoma                12-2    38.41 [ 2]  31.59 [ 1]   6.81 [ 7]
 14. '08 Texas                   12-1    44.86 [ 4]  31.02 [ 4]  13.84 [ 3]
 15. '00 Oklahoma                13-0    57.15 [ 1]  29.91 [ 3]  27.23 [ 1]
 16. '00 Nebraska                10-2    35.93 [ 5]  29.81 [ 4]   6.12 [ 8]
 17. '04 California              10-2    34.76 [ 6]  29.62 [ 2]   5.14 [12]
 18. '03 Louisiana State         13-1    44.87 [ 1]  29.10 [ 2]  15.76 [ 1]
 19. '05 Ohio State              10-2    36.16 [ 4]  29.07 [ 3]   7.09 [ 7]
 20. '04 Louisville              11-1    37.15 [ 5]  28.55 [ 3]   8.60 [ 6]
 21. '02 Oklahoma                12-2    39.79 [ 4]  27.59 [ 3]  12.19 [ 4]
 22. '08 Penn State              11-2    33.03 [ 6]  27.49 [ 5]   5.54 [11]
 23. '07 West Virginia           11-2    34.81 [ 2]  26.84 [ 1]   7.97 [ 4]
 24. '07 Oklahoma                11-3    33.88 [ 4]  26.53 [ 2]   7.34 [ 6]
 25. '03 Southern California     12-1    37.28 [ 3]  26.26 [ 3]  11.01 [ 2]

Rid of that pesky loss getting in the way, the Gators rank third, behind only 2005 Texas and 2001 Miami, and just ahead of the Leinart/Bush USC teams. The 2008 Sooners are 6th and the 2008 Trojans 10th, giving 2008 three of the top ten teams, and still five in the top 25, with Penn State replacing Utah.

USC still has five teams in the top 25, with the 2003 squad replacing the '06 team at #25. Five Sooner squads make the list, too, with with 2003 and 2007 teams joining the 2000, 2002, and 2008, and the 2004 team not making it.

Now we get to the area that hurts Florida the most: Success. Though they were 13-1, they lost to a 9-4 team. To claim your team is one of the best ever is one thing, but with so many undefeated teams over the years many would scoff at a team with a loss claiming to have one of the best seasons ever. Best teams, maybe; best seasons? Probably not. But let's see how they rank among the most successful campaigns from 2000-2008

rnk   Yr Team                   record    Overall    Strength    Success
  1. '04 Southern California     13-0    68.22 [ 1]  36.02 [ 1]  32.19 [ 1]
  2. '02 Ohio State              14-0    49.24 [ 1]  19.70 [ 9]  29.54 [ 1]
  3. '00 Oklahoma                13-0    57.15 [ 1]  29.91 [ 3]  27.23 [ 1]
  4. '05 Texas                   13-0    68.21 [ 1]  40.99 [ 1]  27.22 [ 1]
  5. '04 Auburn                  13-0    48.33 [ 2]  22.94 [ 7]  25.38 [ 2]
  6. '06 Florida                 13-1    46.12 [ 1]  22.49 [ 5]  23.62 [ 1]
  7. '01 Miami (Florida)         12-0    60.63 [ 1]  39.26 [ 1]  21.36 [ 1]
  8. '08 Utah                    13-0    39.93 [ 5]  19.04 [10]  20.89 [ 1]
  9. '08 Florida                 13-1    58.46 [ 1]  38.05 [ 1]  20.40 [ 2]
 10. '00 Washington              11-1    39.65 [ 4]  19.41 [ 9]  20.24 [ 2]
 11. '00 Miami (Florida)         11-1    49.55 [ 2]  33.69 [ 2]  15.85 [ 3]
 12. '03 Louisiana State         13-1    44.87 [ 1]  29.10 [ 2]  15.76 [ 1]
 13. '06 Boise State             13-0    32.40 [ 6]  17.40 [12]  15.00 [ 2]
 14. '06 Auburn                  11-2    29.05 [ 7]  14.08 [20]  14.96 [ 3]
 15. '04 Oklahoma                12-1    39.93 [ 3]  25.64 [ 5]  14.29 [ 3]
 16. '08 Texas                   12-1    44.86 [ 4]  31.02 [ 4]  13.84 [ 3]
 17. '01 Tennessee               11-2    35.19 [ 3]  21.36 [ 5]  13.83 [ 2]
 18. '06 Southern California     11-2    38.74 [ 2]  25.22 [ 2]  13.52 [ 4]
 19. '02 Miami (Florida)         12-1    39.20 [ 5]  26.02 [ 4]  13.18 [ 2]
 20. '06 Louisville              12-1    37.05 [ 3]  23.86 [ 4]  13.18 [ 5]
 21. '04 Utah                    12-0    38.89 [ 4]  25.92 [ 4]  12.97 [ 4]
 22. '05 Penn State              11-1    38.28 [ 3]  25.46 [ 4]  12.82 [ 2]
 23. '02 Georgia                 13-1    34.79 [ 6]  22.01 [ 6]  12.78 [ 3]
 24. '00 Oregon State            11-1    35.31 [ 6]  22.92 [ 6]  12.39 [ 4]
 25. '02 Oklahoma                12-2    39.79 [ 4]  27.59 [ 3]  12.19 [ 4]

Now our Gators only rank #9 in the decade. 2004 USC is on top, followed by the 14-0 Buckeyes from 2002. Two more undefeated champs from Oklahoma and Texas follow, and then unbeaten Auburn from 2004. Interestingly, Florida from 2006—also 13-1—is #6, ahead of 2001 Miami. Then Utah from this year is slightly ahead of the 1-loss Gators.

2002 Ohio State, 2004 Auburn, and 2008 Utah are similar teams in a way; all three went undefeated but were clearly not the most dominant teams in that particular year. Ohio State needed last-second heroics or overtime to beat some very mediocre teams; Auburn's margins of victory were slight; and Utah had some very close games, too. But all three finished #1 or #2 in Success. 2008 Utah's Strength is almost exactly equal to 2002 Ohio State's, but the Buckeyes won 14 games, defeating unbeaten Miami, so their Success rating was far higher. As a contrast, Utah in 2004 had a stronger Strength rating, but few points from its light schedule in Success.

How does 2006 Florida rate better than 2008 Florida despite the 10 bowl team wins the latter can claim? Well, one reason is the 2006 team also beat 10 bowl teams! Pretty much any team that wins the SEC is going to face at least six bowl teams, maybe a couple in pre-conference games, one in the SEC championship game, and one in their own bowl. Last year LSU only had eight wins over bowl teams, since they lost two games, and South Carolina didn't go to a bowl at 6-6, otherwise they would have had 10, also.

The 2006 Florida team lost to 11-2 Auburn rather than 9-4 Mississippi, and they beat an undefeated Ohio State team to win the championship rather than a 1-loss Oklahoma team. So with stronger overall wins—when score margin doesn't matter—and a more excuseable loss, one can argue that the 2006 team achieved more than the 2008 squad.

In conclusion: I'd say that, based on how dominant Florida was after their loss, beating good teams by 30-50 points and winning the national championship, that they deserve to be remembered as one of the very best teams of this decade. Beyond that, it depends on how you compare teams from various decades as to where they finish all-time. But in my book, any team that doesn't go undefeated has shown a major imperfection. The best teams are both dominant in winning, and simply don't lose. By this measure, even if 2008 Florida might be one of the best teams of all time, I don't think can claim to have had one of the best seasons of all time.

Florida Gators #1 in final power rating; Utah tops Success ratings

Florida finished the year #1 in my Overall/Combined power rating. The Gators topped the Strength portion and were a close #2 in Success after undefeated Utah.

The final top ten for the 2008 season, Overall/Combined power rating:

  #  Team                         rec    Overall     Strength     Success
  1. Florida                     13-1    58.46 [ 1]  38.05 [ 1]  20.40 [ 2]
  2. Oklahoma                    12-2    47.28 [ 2]  36.15 [ 2]  11.13 [ 5]
  3. Southern Cal                12-1    45.48 [ 3]  33.93 [ 3]  11.54 [ 4]
  4. Texas                       12-1    44.86 [ 4]  31.02 [ 4]  13.84 [ 3]
  5. Utah                        13-0    39.93 [ 5]  19.04 [10]  20.89 [ 1]
  6. Penn State                  11-2    33.03 [ 6]  27.49 [ 5]   5.54 [11]
  7. Texas Tech                  11-2    30.40 [ 7]  21.69 [ 6]   8.70 [ 7]
  8. Alabama                     12-2    27.68 [ 8]  18.97 [11]   8.70 [ 6]
  9. TCU                         11-2    27.22 [ 9]  21.47 [ 7]   5.74 [10]
 10. Mississippi                  9-4    25.67 [10]  18.25 [13]   7.42 [ 8]

Oklahoma still finishes #2 despite their loss, with USC edging Texas for #3. The Longhorns' Success rating was hurt by Oklahoma's loss, as their win over the Sooners was giving them a lot of points.

Undefeated Utah comes in at 5th, their lower Strength rating sabotaging their top-ranked Success spot. Penn State, Texas Tech, Alabama, and TCU, all 2-loss teams, follow. Mississippi sneaks in at #10 after upsetting the Red Raiders.

The Strength subcomponent, which uses margin of victory, is a measure of how good teams are independent of their record. Florida and Oklahoma topped the list before their game last night and still do the next morning:

  #  Team                         rec    Overall     Strength     Success
  1. Florida                     13-1    58.46 [ 1]  38.05 [ 1]  20.40 [ 2]
  2. Oklahoma                    12-2    47.28 [ 2]  36.15 [ 2]  11.13 [ 5]
  3. Southern Cal                12-1    45.48 [ 3]  33.93 [ 3]  11.54 [ 4]
  4. Texas                       12-1    44.86 [ 4]  31.02 [ 4]  13.84 [ 3]
  5. Penn State                  11-2    33.03 [ 6]  27.49 [ 5]   5.54 [11]
  6. Texas Tech                  11-2    30.40 [ 7]  21.69 [ 6]   8.70 [ 7]
  7. TCU                         11-2    27.22 [ 9]  21.47 [ 7]   5.74 [10]
  8. Missouri                    10-4    22.42 [13]  20.08 [ 8]   2.33 [31]
  9. Ohio State                  10-3    22.47 [12]  19.41 [ 9]   3.05 [21]
 10. Utah                        13-0    39.93 [ 5]  19.04 [10]  20.89 [ 1]

Note that Utah is barely in the top ten. They often didn't win convincingly, but they always won. For comparison's sake, the undefeated Utah team from 2004 finished 4th in both Strength and Success, and #4 overall. That team had an easier schedule but won by a wider margin. You could say the 2004 team was better than the 2008 team but didn't have the necessary challenges on its schedule to rank higher. The 2004 team would rank #6 in Strength this year, and 6th overall.

Missouri and Ohio State rate quite a bit higher in Strength than Success, reflecting their lack of wins over tougher competition, but their ability to beat lesser foes convincingly, and/or stay close to tougher teams for the most part (the OSU-USC game and Missouri-Oklahoma notwithstanding).

In the Success component, undefeated Utah narrowly places ahead of 1-loss Florida:

  #  Team                         rec    Overall     Strength     Success
  1. Utah                        13-0    39.93 [ 5]  19.04 [10]  20.89 [ 1]
  2. Florida                     13-1    58.46 [ 1]  38.05 [ 1]  20.40 [ 2]
  3. Texas                       12-1    44.86 [ 4]  31.02 [ 4]  13.84 [ 3]
  4. Southern Cal                12-1    45.48 [ 3]  33.93 [ 3]  11.54 [ 4]
  5. Oklahoma                    12-2    47.28 [ 2]  36.15 [ 2]  11.13 [ 5]
  6. Alabama                     12-2    27.68 [ 8]  18.97 [11]   8.70 [ 6]
  7. Texas Tech                  11-2    30.40 [ 7]  21.69 [ 6]   8.70 [ 7]
  8. Mississippi                  9-4    25.67 [10]  18.25 [13]   7.42 [ 8]
  9. Boise St                    12-1    24.13 [11]  18.13 [14]   5.99 [ 9]
 10. TCU                         11-2    27.22 [ 9]  21.47 [ 7]   5.74 [10]

In this measure, which uses no score information other than to determine the winner of a game, Utah has enough big wins—#6 Alabama, #10 TCU, and #13 Oregon State—an no losses, which puts them just a touch above Florida. The Gators beat #5 Oklahoma, #6 Alabama, and #14 Georgia, and most of their other wins are over winning teams. When offset by their loss to #8 Mississippi, it almost gives them enough to place #1 despite the loss, but not quite.

The rest of the usual suspects follow, with notable inclusions of Mississippi—who has four losses but beat #2 Florida and #7 Texas Tech—Boise State (one loss to #10 TCU) and TCU, who beat #9 Boise State and only lost to #1 Utah and #5 Oklahoma.

#6 Alabama lost only to the top two teams, #2 Florida and #1 Utah, in their final two games.

Oklahoma tops BCS-style Success power rating

The change in my Success power rating (a BCS-style rating where score margin is not used) from last week to this week is telling. Here are last week's top ten teams:

Success Ratings (win/loss; score margin not considered) through games of November 22, 2008.

  #  Team                         rec   Success
  1. Utah                        12-0    15.04
  2. Texas                       10-1    14.99
  3. Alabama                     11-0    14.82
  4. Oklahoma                    10-1    14.47
  5. Texas Tech                  10-1    13.28
  6. Penn State                  11-1     9.56
  7. Florida                     10-1     9.15
  8. Southern Cal                 9-1     8.22
  9. Boise St                    11-0     7.94
 10. Ball St                     11-0     7.79

Note that undefeated Utah is ahead of Alabama, and Texas ahead of Oklahoma. Now here are this week's numbers (through games of November 29, 2008):

  #  Team                         rec   Success
  1. Oklahoma                    11-1    16.17
  2. Texas                       11-1    15.22
  3. Alabama                     12-0    14.49
  4. Utah                        12-0    14.21
  5. Texas Tech                  11-1    13.30
  6. Florida                     11-1    10.10
  7. Boise St                    12-0     9.57
  8. Southern Cal                10-1     8.92
  9. Penn State                  11-1     8.77
 10. Ball St                     12-0     8.52

Oklahoma has taken over the number one spot, while Utah has fallen to fourth. The Success ratings are "additive" in that a team receives points for each win and loses points for losing. The more wins, the more points a team has. Oklahoma gained a lot of points for beating the Cowboys while Texas gained only slightly by beating Texas A&M. Most of the time, a team that wins head-to-head is higher than a team it beat, but in this case Oklahoma's aggregate wins—over TCU, Cincinnati, Texas Tech, and Oklahoma State—are enough to rank them ahead of the Longhorns. I suspect most of the BCS computers will do the same; if not now, after the Big Twelve title game for sure (assuming a Sooner win).

Alabama earned very little for beating Auburn despite the 36-0 score, since win margin doesn't matter, but they lost points when Georgia was upset by Georgia Tech, as the Bulldogs are the best win on Alabama's slate right now. That would change if they beat Florida, naturally.

By itself, not playing doesn't hurt a team, but Utah actually loses a small amount of points. This is due to Oregon State's loss, which diminished their value as a win (idle Penn State also lost points with Oregon State's downgrade). So the Utes fall to fourth in a very tight race at the top.

In fact, the Oregon-Oregon State game affected a lot of teams in the top ten. With Oregon State losing, this not only hurt Utah and Penn State who beat them, but also USC who lost to them. And Oregon's win helped Boise State quite a bit; their win over Fresno State isn't why they jump to #7, it was their September win over Oregon. Likewise, while USC has a loss to Oregon State, they also have a win over Oregon, so the net result is mostly a wash and they gain a small amount for beating lowly Notre Dame.

Ball State is the lowest ranked of the undefeated teams, but is still in the top ten since the MAC is a lot better this year (the conference has several wins over Big Ten teams this season). Additionally, the 2-loss teams in general have few impressive wins, while the 3- and 4-loss teams are hurt too much by their multiple losses.

Power Ratings: week 7

Now at the mid-point of the season, here are the top 20 teams in the SportsRatings Overall (or Combined) Power Rating:

  #  Team                         rec    Overall     Strength     Success
  1. Texas                        6-0    44.33 [ 1]  34.84 [ 3]   9.49 [ 1]
  2. Penn State                   7-0    43.21 [ 2]  36.27 [ 1]   6.94 [ 5]
  3. Southern Cal                 4-1    40.43 [ 3]  35.54 [ 2]   4.88 [11]
  4. Oklahoma                     5-1    36.85 [ 4]  32.29 [ 4]   4.55 [12]
  5. Missouri                     5-1    35.19 [ 5]  32.03 [ 5]   3.16 [25]
  6. Alabama                      6-0    31.30 [ 6]  22.49 [ 7]   8.81 [ 2]
  7. Florida                      5-1    30.16 [ 7]  26.82 [ 6]   3.33 [23]
  8. Oklahoma St                  6-0    29.90 [ 8]  22.34 [ 8]   7.56 [ 3]
  9. Texas Tech                   6-0    27.80 [ 9]  22.19 [ 9]   5.61 [ 8]
 10. Georgia                      5-1    25.64 [10]  22.08 [10]   3.55 [19]
 11. Georgia Tech                 5-1    25.06 [11]  21.25 [11]   3.81 [18]
 12. California                   4-1    23.16 [12]  18.98 [12]   4.17 [15]
 13. North Carolina               5-1    22.44 [13]  18.13 [14]   4.31 [14]
 14. Ball St                      7-0    22.44 [14]  15.92 [19]   6.51 [ 6]
 15. Michigan St                  6-1    21.07 [15]  16.10 [18]   4.97 [10]
 16. TCU                          6-1    20.75 [16]  17.50 [15]   3.24 [24]
 17. Utah                         7-0    20.33 [17]  13.13 [27]   7.20 [ 4]
 18. Oregon St                    3-3    20.30 [18]  18.24 [13]   2.05 [34]
 19. Boise St                     5-0    19.30 [19]  14.11 [23]   5.19 [ 9]
 20. Brigham Young                6-0    19.04 [20]  15.21 [20]   3.82 [17]

It's interesting to note that the top 10 teams are the same ones, in a somewhat different order, as the teams in the AP top ten, except BYU in in AP's top ten in place of Missouri.

Georgia Tech is the highest ranked team here that isn't in the AP poll. They catch a break here because games with Division I-AA teams, such as their squeaker win over Gardner-Webb, don't count much (this is to reduce the effect of beating up on such teams). In the 727-team power rating, which includes all games in all divisions, the Yellow Jackets are only 31st.

The Success Ratings below show how teams have performed so far this year, without considering score margin:

  #  Team                         rec    Overall     Strength     Success
  1. Texas                        6-0    44.33 [ 1]  34.84 [ 3]   9.49 [ 1]
  2. Alabama                      6-0    31.30 [ 6]  22.49 [ 7]   8.81 [ 2]
  3. Oklahoma St                  6-0    29.90 [ 8]  22.34 [ 8]   7.56 [ 3]
  4. Utah                         7-0    20.33 [17]  13.13 [27]   7.20 [ 4]
  5. Penn State                   7-0    43.21 [ 2]  36.27 [ 1]   6.94 [ 5]
  6. Ball St                      7-0    22.44 [14]  15.92 [19]   6.51 [ 6]
  7. Virginia Tech                5-1    18.45 [21]  12.37 [31]   6.08 [ 7]
  8. Texas Tech                   6-0    27.80 [ 9]  22.19 [ 9]   5.61 [ 8]
  9. Boise St                     5-0    19.30 [19]  14.11 [23]   5.19 [ 9]
 10. Michigan St                  6-1    21.07 [15]  16.10 [18]   4.97 [10]

Note that there are only two teams with losses in the top ten. One surprise is Michigan State, who lost to 4-1 Cal and beat 4-2 Notre Dame and 5-1 Northwestern. Undefeated BYU and Tulsa are in the 2nd 10 as their opposition hasn't proven very sturdy.

Which teams will do the best from here? Probably the ones high in the Strength Ratings:

  #  Team                         rec    Overall     Strength     Success
  1. Penn State                   7-0    43.21 [ 2]  36.27 [ 1]   6.94 [ 5]
  2. Southern Cal                 4-1    40.43 [ 3]  35.54 [ 2]   4.88 [11]
  3. Texas                        6-0    44.33 [ 1]  34.84 [ 3]   9.49 [ 1]
  4. Oklahoma                     5-1    36.85 [ 4]  32.29 [ 4]   4.55 [12]
  5. Missouri                     5-1    35.19 [ 5]  32.03 [ 5]   3.16 [25]
  6. Florida                      5-1    30.16 [ 7]  26.82 [ 6]   3.33 [23]
  7. Alabama                      6-0    31.30 [ 6]  22.49 [ 7]   8.81 [ 2]
  8. Oklahoma St                  6-0    29.90 [ 8]  22.34 [ 8]   7.56 [ 3]
  9. Texas Tech                   6-0    27.80 [ 9]  22.19 [ 9]   5.61 [ 8]
 10. Georgia                      5-1    25.64 [10]  22.08 [10]   3.55 [19]

There are five Big Twelve teams in the top ten, so expect more big battles and upsets like we saw on Saturday from that conference. The SEC has some big powers too, with Florida, Alabama, and Georgia rating in the top ten, but LSU (59th) and Auburn (62nd) don't look very scary any more.

It's entirely possible that the best teams from those conferences—both of whom play a championship game at the end of the season—will knock each other off enough that Penn State and USC could end up playing in the national title game. Both the Nittany Lions and the Trojans have a comparatively easy schedule the rest of the way compared to the gauntlet that any SEC or Big 12 team has to face.

LA Times article ignorant about computer ratings

Here we go again. The annual article showing a complete lack of understanding about computer rating systems. It's not so much the ignorance that bothers me, but that the authors never even attempt to try to gain an understanding. They just push forth a few examples that make computer rankings look stupid or irrational, and then say "wow, can't possibly make heads or tails of that! Why even try?"

The examples in the article make complete sense to me, mostly because my two rating systems mirror the examples almost exactly. The author (Chris Dufresne) exhibits dismay that Jeff Sagarin's two systems can evaluate the same team differently. Here are his examples:

• USC: #1 but also #16
• Northwestern: #28 and #2
• Florida: #7 and #46

He doesn't even understand how Sagarin combines his ratings. The real numbers he should have cited are:

• USC: #1 (Predictor) and #16 (ELO)
• Northwestern: #35 and #2
• Florida: #2 and #46

Using my comparable ratings (Strength is similar to Predictor, and Success mirrors ELO somewhat):

• USC: #3 and #17
• Northwestern: #37 and #10
• Florida: #6 and #40

So while the differences in my ratings aren't quite as extreme, they're pretty close on the whole. So why the discrepancies between two methods of rating college football teams? There are many reasons.

One is that it's early in the season. With only four or five games to go by, any one game can have a big effect on a score-based system. Likewise, using winning and losing only, five games is a paltry amount to base a rating on. This is why I (and Mr. Sagarin) combine the two systems. When combined, the deficiencies of the two systems are lessened. USC ends up #3, Florida #8, and Northwestern #27 in my Overall ratings. In Sagarin's method, USC is #1, Florida is #7, and Northwestern #28.

Also, early ratings are sometimes handled differently until there are enough "connections" to make valid comparisons. Jeff Sagarin uses some weighing of last year's results until this year has enough data. At this point, according to his site, Predictor (the score-margin system) still is using some weight from last year, while the ELO portion does not. This is naturally going to lead to greater discrepancy between the two.

For the SportsRatings power rating, there is never any use of last year's data. But early ratings are naturally weighed heavier with the results of Strength rather than Success. The Success ratings build throughout the year as teams win games. It's difficult for a Win-Loss-only system to have good results until teams have played about 10 games; until then, the score-based system carries more weight. My motto is "you can't argue with Success" but only at the end of the season. Until then, some results, like Oregon State's upset of USC, can be at least somewhat written off as a fluke.

Dufresne does point out that USC's huge wins over Ohio State and Oregon count for a lot in a margin-based system. From there it doesn't take a lot for the ratings to make perfect sense; why should a 3-1 team by any higher than #16 anyway?

Florida's case is easier to understand. If you only knew that some "Team X" had defeated four teams that were each 2-3, and Team X had lost to another that was 3-3, wouldn't #46 seem like a possible ranking? But knowing that the score of the four wins were 56-10, 26-3, 30-6, and 38-7, and that the loss was by 1 point, you might think it reasonable that they be in the top ten. 

Just a little digging could help people understand the computer rankings, and it's an area that journalists have failed. There is so much coverage of every aspect of college football by the press—offenses, injuries, history, human-interest storylines—but the computer rankings are addresses with either pure ridicule and/or ignorance. There is no attempt to educate the fans on exactly what mechanisms they operate by. Why not? Sure, it may make for a longer, more in-depth article, but it has become a crucial area due to the BCS system. But all we get is "oh my God, look at what those computers did!"

Even for the financial meltdown in the stock market has articles that attempt to explain, step-by-step, just what happened in a technical sense. And that "system" is so complex than no one really understands how it works or where it is going since its scope is so huge. College football has a finite amount of games—it would be possible to show exactly what a computer is "thinking" every step of the way!

Sportswriters do a disservice to fans when they use the computer ratings only to take potshots at the BCS rather than to introduce fans to a very interesting nexus of sports and mathematics. The introduction of computer power ratings has clearly influenced sports and college football in particular. Spreads on games hew close to computer models these days. And pollsters, over the last ten years, have started to refer to power ratings when making their choices. In the early days of Sagarin's ratings being featured in USA Today, the difference between his ratings and the polls was huge. Then many of the teams he ranked high started to "upset" other teams. After years of this happening, you started to see these teams enter the top 25 earlier, and move up faster, than they did in the past.

Now, with spreads and the AP poll mirroring the computer rankings so much, any discrepancy is quickly noted. But does the AP poll make that much sense right now? #24 Pitt beat #19 South Florida who beat #16 Kansas; all teams have 1 loss. Michigan State is at #23 at 5-1, but the 4-1 Cal team that beat them isn't rated. There are justifiable reasons for these discrepancies perhaps. But the polls have another advantage: they are the aggregate of 70 different opinions. Combine 70 computer rankings and you wouldn't be able to pick one or two apart for having widely contrasting results.

The inconsistencies and biases that show up in the human polls are one reason I like computer power ratings. It's also why I base the SportsRatings Top 25—a non-computerized, subjective ranking—on a few basic principles so that teams are ranked in a fair manner. First and foremost, I try to explain why I put teams where I do, so that someone who takes issue with them at least knows where I'm coming from.

It's a lot easier just to mock a rating system than to try and understand it. But I think there is fan interest in knowing how these systems work, and journalists would be doing their job to go a step further in this direction.

Power Ratings for week 6

SportsRatings Power Rating updates for Week 6

The Strength rating, which measures a team's true strength using compared score margins:

  #  Team                         rec    Overall     Strength     Success
  1. Missouri                     5-0    46.40 [ 1]  39.97 [ 1]   6.43 [ 5]
  2. Penn State                   6-0    41.67 [ 3]  35.57 [ 2]   6.10 [ 7]
  3. Southern Cal                 3-1    39.17 [ 4]  35.50 [ 3]   3.67 [17]
  4. Oklahoma                     5-0    41.77 [ 2]  34.66 [ 4]   7.10 [ 2]
  5. Texas                        5-0    37.76 [ 5]  32.53 [ 5]   5.23 [ 9]
  6. Florida                      4-1    27.20 [ 8]  25.49 [ 6]   1.70 [40]
  7. Texas Tech                   5-0    28.50 [ 7]  24.51 [ 7]   3.98 [13]
  8. Georgia                      4-1    25.84 [ 9]  23.05 [ 8]   2.78 [25]
  9. Tulsa                        5-0    25.54 [10]  21.56 [ 9]   3.97 [14]
 10. North Carolina               4-1    23.99 [12]  21.09 [10]   2.90 [23]

Missouri's big win over Nebraska puts them on top, but note that USC is #3 despite their loss to Oregon State. Beating Oregon by 34 didn't hurt. Tulsa is pretty strong at #9, and North Carolina even makes the top ten.

Now the Success rating, which doesn't use score margin:

  #  Team                         rec    Overall     Strength     Success
  1. Alabama                      6-0    30.17 [ 6]  20.91 [11]   9.25 [ 1]
  2. Oklahoma                     5-0    41.77 [ 2]  34.66 [ 4]   7.10 [ 2]
  3. Vanderbilt                   5-0    20.91 [16]  13.96 [23]   6.95 [ 3]
  4. Utah                         6-0    20.08 [17]  13.37 [27]   6.70 [ 4]
  5. Missouri                     5-0    46.40 [ 1]  39.97 [ 1]   6.43 [ 5]
  6. Ball St                      6-0    25.50 [11]  19.31 [15]   6.18 [ 6]
  7. Penn State                   6-0    41.67 [ 3]  35.57 [ 2]   6.10 [ 7]
  8. Virginia Tech                5-1    18.89 [22]  13.22 [28]   5.67 [ 8]
  9. Texas                        5-0    37.76 [ 5]  32.53 [ 5]   5.23 [ 9]
 10. Northwestern                 5-0    15.93 [27]  10.87 [37]   5.06 [10]
 11. Oklahoma St                  5-0    19.99 [18]  15.40 [20]   4.59 [11]
 12. Boise St                     4-0    17.28 [25]  12.97 [29]   4.30 [12]
 13. Texas Tech                   5-0    28.50 [ 7]  24.51 [ 7]   3.98 [13]
 14. Tulsa                        5-0    25.54 [10]  21.56 [ 9]   3.97 [14]
 15. California                   4-1    19.41 [21]  15.58 [19]   3.82 [15]
 16. Ohio State                   5-1    15.84 [28]  12.05 [35]   3.79 [16]
 17. Southern Cal                 3-1    39.17 [ 4]  35.50 [ 3]   3.67 [17]
 18. Michigan St                  5-1    16.28 [26]  12.76 [30]   3.51 [18]
 19. Georgia Tech                 4-1    23.73 [13]  20.46 [12]   3.26 [19]
 20. LSU                          4-0     9.11 [46]   5.89 [50]   3.22 [20]
 21. Brigham Young                5-0    17.85 [24]  14.74 [21]   3.10 [21]

I list the top 21, which includes all the undefeated teams. Between the undefeated teams it comes down to who has the tougher schedule. Of course playing more games raises the rating too. Ball State's schedule so far has been roughly equal to Penn State's. If both teams remain undefeated, the Lions will gain many more points than the Cardinals though. Of the teams with a loss Virginia Tech stands out, as they have defeated North Carolina and Georgia Tech, who are both 4-1.

The combined rankings give a balance between raw power potential and proven ability to win games:

  #  Team                         rec    Overall     Strength     Success
  1. Missouri                     5-0    46.40 [ 1]  39.97 [ 1]   6.43 [ 5]
  2. Oklahoma                     5-0    41.77 [ 2]  34.66 [ 4]   7.10 [ 2]
  3. Penn State                   6-0    41.67 [ 3]  35.57 [ 2]   6.10 [ 7]
  4. Southern Cal                 3-1    39.17 [ 4]  35.50 [ 3]   3.67 [17]
  5. Texas                        5-0    37.76 [ 5]  32.53 [ 5]   5.23 [ 9]
  6. Alabama                      6-0    30.17 [ 6]  20.91 [11]   9.25 [ 1]
  7. Texas Tech                   5-0    28.50 [ 7]  24.51 [ 7]   3.98 [13]
  8. Florida                      4-1    27.20 [ 8]  25.49 [ 6]   1.70 [40]
  9. Georgia                      4-1    25.84 [ 9]  23.05 [ 8]   2.78 [25]
 10. Tulsa                        5-0    25.54 [10]  21.56 [ 9]   3.97 [14]
 11. Ball St                      6-0    25.50 [11]  19.31 [15]   6.18 [ 6]
 12. North Carolina               4-1    23.99 [12]  21.09 [10]   2.90 [23]
 13. Georgia Tech                 4-1    23.73 [13]  20.46 [12]   3.26 [19]
 14. TCU                          5-1    22.67 [14]  20.04 [13]   2.63 [27]
 15. Kentucky                     4-1    21.60 [15]  19.46 [14]   2.13 [34]
 16. Vanderbilt                   5-0    20.91 [16]  13.96 [23]   6.95 [ 3]
 17. Utah                         6-0    20.08 [17]  13.37 [27]   6.70 [ 4]
 18. Oklahoma St                  5-0    19.99 [18]  15.40 [20]   4.59 [11]
 19. Illinois                     3-2    19.68 [19]  18.68 [16]   0.99 [53]
 20. Oregon St                    2-3    19.45 [20]  17.99 [17]   1.46 [44]
 21. California                   4-1    19.41 [21]  15.58 [19]   3.82 [15]
 22. Virginia Tech                5-1    18.89 [22]  13.22 [28]   5.67 [ 8]
 23. Wake Forest                  3-1    18.30 [23]  15.59 [18]   2.70 [26]
 24. Brigham Young                5-0    17.85 [24]  14.74 [21]   3.10 [21]
 25. Boise St                     4-0    17.28 [25]  12.97 [29]   4.30 [12]
 26. Michigan St                  5-1    16.28 [26]  12.76 [30]   3.51 [18]
 27. Northwestern                 5-0    15.93 [27]  10.87 [37]   5.06 [10]

 46. LSU                          4-0     9.11 [46]   5.89 [50]   3.22 [20]

Some interesting things to note: Oregon State's Strength rating is high (#17) both for beating USC and staying close to Utah, but only the USC win helps their Success rating, which is dragged down by the loss to Stanford. Their other loss was to Penn State.

Louisiana State is only #46 due to a surprisingly terrible strength rating of 50th. A few reasons for this are: Their win over Auburn doesn't count for much because Auburn is ranked even lower; they played Mississippi State pretty close, and the Bulldogs are ranked 86th; a blowout over North Texas doesn't help them since the Mean Green are 118th among Division I-A teams. Finally their convincing win over I-AA power Appalachian State isn't credited much in this I-A only rating. The 727-team all-division rankings better capture that victory's signficance, and LSU is #37 in that ranking, which is still very low for a supposedly top ten team. Over the next few weeks either LSU will be exposed as a fraud, or their rating will increase significantly. The former happened to South Florida (#49 in Strength) last week vs. Pitt.

Power ratings for week 5

Fourth last week, Oklahoma takes over the top spot in the Overall (combined) power rating. Penn State remains #2, while Texas jumps from sixth to third. Missouri, though idle this week, jumps from 25th to 4th. What happened? Well the numbers just reached a "tipping point" where the computer decided the Tigers were really good. All of Missouri's opponents moved up sharply, though two of them lost.

Speaking of losing, USC fell to #7 from #3, and Florida from #5 to #6. The latter's fall was cushioned by Mississippi jumping up to #17. Georgia took the biggest fall of the three prominent losers, dropping from the #1 spot to #9, while Alabama rose to #5 from #7 for the win. Another loser, TCU, moved up a spot as losing to Oklahoma by 25 ultimately didn't hurt their rating.

Overall Power Rating (combined Strength and Success) through games of September 28, 2008.

  #  Team                         rec    Overall     Strength     Success
  1. Oklahoma                     4-0    43.54 [ 1]  37.78 [ 1]   5.75 [ 4]
  2. Penn State                   5-0    40.22 [ 2]  35.62 [ 2]   4.59 [ 8]
  3. Texas                        4-0    38.26 [ 3]  34.45 [ 3]   3.81 [16]
  4. Missouri                     4-0    36.35 [ 4]  32.49 [ 4]   3.86 [14]
  5. Alabama                      5-0    31.71 [ 5]  23.97 [ 7]   7.74 [ 1]
  6. Florida                      3-1    29.41 [ 6]  28.16 [ 5]   1.24 [48]
  7. Southern Cal                 2-1    27.62 [ 7]  25.50 [ 6]   2.11 [31]
  8. Vanderbilt                   4-0    26.78 [ 8]  21.21 [11]   5.57 [ 5]
  9. Georgia                      4-1    25.52 [ 9]  22.75 [ 9]   2.77 [23]
 10. TCU                          4-1    25.41 [10]  23.03 [ 8]   2.37 [26]
 11. Brigham Young                4-0    25.29 [11]  22.71 [10]   2.58 [25]
 12. Texas Tech                   4-0    23.72 [12]  20.61 [12]   3.10 [17]
 13. Kentucky                     4-0    22.97 [13]  18.85 [17]   4.12 [12]
 14. Tulsa                        4-0    21.62 [14]  18.90 [16]   2.71 [24]
 15. Georgia Tech                 3-1    21.21 [15]  19.72 [15]   1.49 [40]
 16. Ball St                      5-0    21.12 [16]  15.66 [24]   5.46 [ 6]
 17. Mississippi                  3-2    21.08 [17]  19.74 [14]   1.33 [44]
 18. North Carolina               3-1    21.05 [18]  19.77 [13]   1.28 [46]
 19. California                   3-1    21.01 [19]  18.02 [18]   2.98 [18]
 20. Virginia Tech                4-1    20.96 [20]  17.13 [19]   3.82 [15]
 21. Utah                         5-0    20.34 [21]  13.38 [30]   6.95 [ 2]
 22. Wake Forest                  3-1    18.67 [22]  15.85 [23]   2.82 [22]
 23. Nebraska                     3-1    18.38 [23]  16.12 [21]   2.25 [27]
 24. Oklahoma St                  4-0    17.92 [24]  14.01 [28]   3.91 [13]
 25. Boise St                     3-0    17.73 [25]  13.15 [31]   4.58 [ 9]

Other teams of note: Georgia Tech is at #15, while undefeated Ball State sits at #16. The Yellow Jackets aren't getting a lot of attention but could compete for the wide-open ACC title along with #18 North Carolina, #20 Virginia Tech, and #22 Wake Forest. The Cardinals have a chance to go undefeated this year with an easy schedule, and these results suggest they have the team to do it and are deserving of a top 25 ranking.

Success Ratings (win/loss; score margin not considered) through games of September 28, 2008.

  #  Team                         rec    Overall     Strength     Success
  1. Alabama                      5-0    31.71 [ 5]  23.97 [ 7]   7.74 [ 1]
  2. Utah                         5-0    20.34 [21]  13.38 [30]   6.95 [ 2]
  3. Northwestern                 5-0    15.68 [30]   9.84 [39]   5.83 [ 3]
  4. Oklahoma                     4-0    43.54 [ 1]  37.78 [ 1]   5.75 [ 4]
  5. Vanderbilt                   4-0    26.78 [ 8]  21.21 [11]   5.57 [ 5]
  6. Ball St                      5-0    21.12 [16]  15.66 [24]   5.46 [ 6]
  7. South Florida                5-0    13.00 [37]   7.79 [45]   5.20 [ 7]
  8. Penn State                   5-0    40.22 [ 2]  35.62 [ 2]   4.59 [ 8]
  9. Boise St                     3-0    17.73 [25]  13.15 [31]   4.58 [ 9]
 10. Connecticut                  5-0     8.96 [47]   4.82 [55]   4.14 [10]

Which team is having the best season against the strongest competition so far? That would be Alabama, which is no surprise after their win over Georgia. The team at #3 is a bit of a surprise: Northwestern is 5-0 with wins over Iowa and Duke. The Duke win may decline in value as the season progresses but right now no one else has a win over the Blue Devils. Ball State's pre-season schedule ranks favorably at the moment but their strength of schedule will have a hard time keeping up with any undefeated BCS team's.

Strength (score-based) Ratings through games of September 28, 2008.

  #  Team                         rec    Overall     Strength     Success
  1. Oklahoma                     4-0    43.54 [ 1]  37.78 [ 1]   5.75 [ 4]
  2. Penn State                   5-0    40.22 [ 2]  35.62 [ 2]   4.59 [ 8]
  3. Texas                        4-0    38.26 [ 3]  34.45 [ 3]   3.81 [16]
  4. Missouri                     4-0    36.35 [ 4]  32.49 [ 4]   3.86 [14]
  5. Florida                      3-1    29.41 [ 6]  28.16 [ 5]   1.24 [48]
  6. Southern Cal                 2-1    27.62 [ 7]  25.50 [ 6]   2.11 [31]
  7. Alabama                      5-0    31.71 [ 5]  23.97 [ 7]   7.74 [ 1]
  8. TCU                          4-1    25.41 [10]  23.03 [ 8]   2.37 [26]
  9. Georgia                      4-1    25.52 [ 9]  22.75 [ 9]   2.77 [23]
 10. Brigham Young                4-0    25.29 [11]  22.71 [10]   2.58 [25]

The Strength component is the best judge for which teams will do well in the future. As you can see it mirrors the overall ratings pretty closely, as the Success numbers haven't built up enough evidence to give teams high scores for their accomplishments yet. Florida and USC still rate high despite their close losses, and TCU and Georgia are top ten as well. BYU's schedule only gives them a Success ranking of 25th, but their back-to-back monster shutouts put them 10th in Strength.

Scoring Offense ratings through games of September 28, 2008.

  #  Team                         rec   Margin  Off    Def
  1. Missouri                     4-0   32.49  54.41  21.91 [47th]
  2. Oklahoma                     4-0   37.78  54.19  16.41 [22]
  3. Penn State                   5-0   35.62  48.03  12.40 [11]
  4. Oregon                       4-1   15.51  45.54  30.03 [91]
  5. Rice                         3-2   11.64  44.19  32.55 [103]
  6. California                   3-1   18.02  43.84  25.81 [71]
  7. Tulsa                        4-0   18.90  43.38  24.47 [66]
  8. Mississippi                  3-2   19.74  42.99  23.24 [54]
  9. Oklahoma St                  4-0   14.01  42.58  28.57 [84]
 10. Nebraska                     3-1   16.12  42.44  26.31 [75]

The offensive top 10, adjusted for opposition, has few surprises. Many of the highest-ranking offenses are paired with very poor defenses: Oregon's defense ranks 91st while Rice, which put up 77 last week on North Texas, ranks an even lower 103rd defensively. #6 offense California is 71st, #7 Tulsa is 66th, Mississippi is 54th, Oklahoma State 84th, and Nebraska 75th. Only the top three offensive teams have decent defenses, with Oklahoma ranked 22nd and Penn State 11th. Many of these teams run a spread-type offense, which could be tiring out their defense by putting them on the field too long.

Scoring Defense ratings through games of September 28, 2008.

  #  Team                         rec   Margin  Off    Def
  1. TCU                          4-1   23.03  26.72   3.69
  2. Texas                        4-0   34.45  41.31   6.85
  3. Wake Forest                  3-1   15.85  24.71   8.86
  4. Florida                      3-1   28.16  37.16   8.99
  5. Kentucky                     4-0   18.85  28.27   9.42
  6. Boston College               3-1    9.15  19.06   9.90
  7. Vanderbilt                   4-0   21.21  31.59  10.37
  8. Virginia Tech                4-1   17.13  28.15  11.01
  9. South Carolina               3-2   10.88  22.63  11.74
 10. West Virginia                2-2    1.79  13.59  11.79

The best defenses are somewhat surprising. TCU ranks #1 after giving up 35 to Oklahoma—the Sooners have scored 50+ on everyone else. And who would have thought Florida would be ranked better on defense than offense? The real shocker, though, is West Virginia, who is an alarming #117 on offense but is 10th in defense. Going into the season I would have thought that those numbers would be reversed.

And for the answer to the question, is Notre Dame's comeback for real? we look at their adjusted scoring offense and defense. They rank 55th on offense and 51st on defense so far, for an overall Strength ranking of #52. Better than average for Division I-A and much improved from last year, but not a great team by any means. But it should be enough to give them several more wins against their easy remaining schedule to get to a bowl game.

Power Ratings, week 4

It's post-week 4 and the later power ratings are offering up some clues as to who might really be the best—and worst—teams in the country. So now we attempt to use the ratings to demonstrate the following points, as of week 4:

• Florida's defense may be better, relatively, than their offense
• Washington State is not only the worst BCS team, but could the worst I-AA team, period
• Notre Dame isn't nearly as bad as last year so far, but is almost exactly average
• Ball State might not just go undefeated because of their easy schedule, but could truly be a top 25 caliber team
• TCU is playing like a top ten team

This week's Overall top ten:

  #  Team                         rec    Overall     Strength     Success
  1. Georgia                     (4-0)   40.66 [ 1]  35.54 [ 2]   5.12 [ 5]
  2. Penn State                  (4-0)   38.84 [ 2]  35.79 [ 1]   3.04 [21]
  3. Southern Cal                (2-0)   36.64 [ 3]  33.14 [ 4]   3.50 [17]
  4. Oklahoma                    (3-0)   36.35 [ 4]  33.68 [ 3]   2.66 [25]
  5. Florida                     (3-0)   34.36 [ 5]  31.01 [ 6]   3.34 [19]
  6. Texas                       (3-0)   34.13 [ 6]  31.39 [ 5]   2.73 [24]
  7. Alabama                     (4-0)   31.69 [ 7]  25.23 [ 9]   6.45 [ 2]
  8. Wisconsin                   (3-0)   31.47 [ 8]  25.82 [ 7]   5.64 [ 3]
  9. Vanderbilt                  (4-0)   28.25 [ 9]  23.01 [13]   5.24 [ 4]
 10. Kentucky                    (3-0)   27.70 [10]  23.80 [11]   3.89 [13]

Most of the names are pretty familiar there: Georgia, which started #1 in the national polls, USC number 3, Oklahoma and Florida in the top five, and Penn State is rated a bit higher at #2. The top ten has five SEC teams, though not the same ones as in the AP and USA/coaches, as Vanderbilt and Kentucky stand at #9 and #10.

Here is this week's Strength rating top ten, which uses score margin only:

  #  Team                         rec    Overall     Strength     Success
  1. Penn State                  (4-0)   38.84 [ 2]  35.79 [ 1]   3.04 [21]
  2. Georgia                     (4-0)   40.66 [ 1]  35.54 [ 2]   5.12 [ 5]
  3. Oklahoma                    (3-0)   36.35 [ 4]  33.68 [ 3]   2.66 [25]
  4. Southern Cal                (2-0)   36.64 [ 3]  33.14 [ 4]   3.50 [17]
  5. Texas                       (3-0)   34.13 [ 6]  31.39 [ 5]   2.73 [24]
  6. Florida                     (3-0)   34.36 [ 5]  31.01 [ 6]   3.34 [19]
  7. Wisconsin                   (3-0)   31.47 [ 8]  25.82 [ 7]   5.64 [ 3]
  8. South Carolina              (2-2)   26.66 [13]  25.50 [ 8]   1.16 [48]
  9. Alabama                     (4-0)   31.69 [ 7]  25.23 [ 9]   6.45 [ 2]
 10. Tulsa                       (3-0)   26.55 [14]  24.11 [10]   2.44 [27]

Penn State has been the most impressive team by this measure, crushing all their foes, but Georgia, Oklahoma, USC, and Texas are not far behind.

Here is the same Strength rating, but using the games of all 727 teams from all divisions. Note that USC is now a clear #1, but most of the same teams are represented.

  #  Team                         rec   Rating  Off   Def
  1. Southern Cal                 2-0   88.01  73.19 -14.82
  2. Penn State                   4-0   81.64  72.86  -8.77
  3. Florida                      3-0   77.26  61.61 -15.64
  4. Texas                        3-0   75.03  61.97 -13.06
  5. Oklahoma                     3-0   72.27  73.62   1.35
  6. TCU                          4-0   68.71  53.70 -15.01
  7. Kansas St                    2-1   68.63  64.32  -4.30
  8. Alabama                      4-0   68.37  56.89 -11.48
  9. California                   2-1   68.31  73.14   4.83
 10. Tulsa                        3-0   68.29  68.44   0.14

The above rating is slightly different in the early weeks, but later comes to resemble the Division I-A rating once I-A teams have played a majority I-A schedule. The Division I-A -only ratings are skewed at first by lack of knowledge of I-AA opposition; later on the I-AA games fade in significance. This is a good thing, and keeps wins over I-AA teams from impacting the ratings much by the end of the year. But right now the X-Division (cross-division) rankings workbetter.

Who has the best offense and defense? Well based on the games played so far, Oklahoma is the offense king, and Kentucky, believe it or not, is best on defense.

  #  Team                         rec    Margin  Off    Def
  1. Oklahoma                     3-0    33.68  51.80  18.11
  2. California                   2-1    16.88  51.11  34.22
  3. Tulsa                        3-0    24.11  49.06  24.95
  4. Missouri                     4-0    16.09  47.45  31.35
  5. Oregon                       3-1    13.52  45.04  31.52
  6. Penn State                   4-0    35.79  43.82   8.02
  7. North Carolina               2-1    19.91  43.30  23.38
  8. Mississippi                  2-2    16.62  43.26  26.64
  9. Georgia Tech                 3-1    20.82  42.43  21.61
 10. Utah                         4-0    19.05  42.26  23.20

  #  Team                         rec    Margin  Off    Def
  1. Kentucky                     3-0    23.80  23.74  -0.06
  2. Southern Cal                 2-0    33.14  35.09   1.94
  3. Georgia                      4-0    35.54  38.37   2.83
  4. Wake Forest                  3-0    20.85  24.40   3.55
  5. Virginia Tech                3-1    18.83  22.55   3.71
  6. Texas                        3-0    31.39  37.64   6.25
  7. Boston College               2-1     9.60  15.89   6.28
  8. Florida                      3-0    31.01  38.03   7.01
  9. Vanderbilt                   4-0    23.01  30.21   7.20
 10. Michigan St                  3-1    11.68  19.49   7.81

Again, with the games of all 700+ teams included, the answers are a bit different and probably a bit more accurate. Oklahoma is still the best offense, but USC comes in at #2 and BYU is in the top ten. Defensively, Wake Forest is #1, with TCU up at #3.

Now for the fun findings:

• Note that above, Florida's offense doesn't show up in the top ten—it ranks 19th, while the defense that held Miami to three points, Hawaii to 10, and Tennessee to 6 is ranked 8th.

• Washington State, as mentioned, is #131 out of 133 teams (13 division I-AA are listed due to playing majority I-A schedules so far; they'll be excluded in later ratings) and the lowest-rated I-A team of all. Winning last week didn't keep them out of the cellar, as they beat a lowly I-AA opponent. But they might not quite be the worst: SMU ranks lower in the X-Division Strength ratings.

• Notre Dame is doing much better than last year when they finished 80th in the country in Strength. This year so far they're 60th, which puts them smack dab in the middle of the 120 I-A teams.

• The Ball State Cardinals are undefeated and looking to remain so all year. They play in the MAC, so even if their record is unblemished, it will be hard to tell if they're really any good. What does the computer say? Well so far, they're #24 in Strength. The Strength rating uses score margin, which is the area they'll do worst in if they go undefeated. Last year, for example, Hawaii was near the top ten in Success (which only considers if you won the game) and around 50th in Strength, which was a dead giveaway about their true quality. So far Ball State looks like a legit top 25 team, but not a BCS-bowl-level squad.

• One team that might fit the BCS bill is Texas Christian. They're #12 in Strength, ranking above even BYU and Utah, both of whom make the top 25. They're even better, #6, in the X-Divisional Strength ratings. This suggests that they have a shot at winning the Mountain West instead of the two favorites. Even so, TCU would have to get past Oklahoma next week to go undefeated. If they do, however, they'll have a legit claim not only to a BCS bowl but a BCS championship game berth.

Finally, here are the Success ratings, which are based on winning and losing only—no point margins used.

  #  Team                         rec    Overall     Strength     Success
  1. Utah                        (4-0)   25.99 [15]  19.05 [21]   6.93 [ 1]
  2. Alabama                     (4-0)   31.69 [ 7]  25.23 [ 9]   6.45 [ 2]
  3. Wisconsin                   (3-0)   31.47 [ 8]  25.82 [ 7]   5.64 [ 3]
  4. Vanderbilt                  (4-0)   28.25 [ 9]  23.01 [13]   5.24 [ 4]
  5. Georgia                     (4-0)   40.66 [ 1]  35.54 [ 2]   5.12 [ 5]
  6. Boise St                    (3-0)   15.26 [32]  10.28 [42]   4.98 [ 6]
  7. Ball St                     (4-0)   22.11 [22]  17.46 [24]   4.65 [ 7]
  8. South Florida               (4-0)    2.06 [70]  -2.29 [81]   4.36 [ 8]
  9. Missouri                    (4-0)   20.31 [25]  16.09 [27]   4.21 [ 9]
 10. LSU                         (3-0)   26.99 [12]  22.85 [15]   4.13 [10]

Wisconsin might be for real this year, it seems; their opponents rank high enough to put them at #3 in Success, and they've beaten them by enough to keep their Strength rating high as well. But South Florida might be a pretender. They've beaten some good foes, like Kansas, but their stinker of a game against Florida International gives them a below-par Strength, only 81st in the country.

It's still very early in the year and really, not many good teams have played each other, so a lot will change. But the preliminary findings are sure interesting.

Three weeks in, power ratings don't make sense; blame scheduling

Back in the good old days of the 11-game schedule, after three weeks power ratings started to make some basic sense. There were still a few offbeat results, but most fell in line with expectations, and there were a few revealing insights to be gained.

No more. Now that there are 12 games, almost every I-A team is scheduling a I-AA team along with a cupcake I-A school, so after three weeks there isn't much to go by. Take Florida State. Can we make any sense of their two blowouts of I-AA teams? With many teams only playing two games so far, one against a I-AA, we'll have to wait until next week to see enough connections made to have a decent output.

  #  Team                         rec    Overall     Strength     Success
  1. Georgia Tech                (2-1)   61.07 [ 1]  60.08 [ 1]   0.99 [48]
  2. East Carolina               (3-0)   60.74 [ 2]  56.20 [ 2]   4.53 [ 2]
  3. Virginia Tech               (2-1)   55.13 [ 3]  53.63 [ 3]   1.49 [36]
  4. Boston College              (1-1)   53.03 [ 5]  52.57 [ 4]   0.46 [72]
  5. Iowa                        (3-0)   53.49 [ 4]  50.26 [ 5]   3.23 [16]
  6. Oklahoma                    (3-0)   52.27 [ 6]  49.75 [ 6]   2.52 [25]
  7. Alabama                     (3-0)   50.96 [ 7]  47.39 [ 7]   3.56 [ 9]
  8. Kansas                      (2-1)   44.31 [ 8]  43.20 [ 8]   1.11 [46]
  9. Tulane                      (0-2)   41.17 [12]  42.77 [ 9]  -1.59 [104]
 10. Iowa St                     (2-1)   43.56 [ 9]  42.65 [10]   0.90 [55]

For now, here's what's going on. Georgia Tech, at 2-1, is on top of the Strength ratings (formerly called Points). Why? For one thing, home field advantage still rates at almost 7 points (cupcakes play on the road) as opposed to the 3.5 or so that the margin will converge to later in the season. And Tech lost a close road game to Virginia Tech, who lost a close one to #2 East Carolina. These particular teams just happened to bubble to the top for reasons that could almost be called chaos theory. For example, USC doesn't rank high because their victims, Virginia and Ohio State, instead got pushed very low on the totem pole.

  #  Team                         rec    Overall     Strength     Success
  1. Auburn                      (3-0)   40.31 [13]  35.40 [13]   4.90 [ 1]
  2. East Carolina               (3-0)   60.74 [ 2]  56.20 [ 2]   4.53 [ 2]
  3. Utah                        (3-0)   31.90 [20]  27.44 [21]   4.45 [ 3]
  4. Nebraska                    (3-0)   19.73 [37]  15.48 [42]   4.24 [ 4]
  5. Vanderbilt                  (3-0)   22.63 [34]  18.60 [36]   4.02 [ 5]
  6. South Florida               (3-0)   43.23 [10]  39.22 [12]   4.01 [ 6]
  7. Wisconsin                   (3-0)   17.59 [41]  13.71 [43]   3.87 [ 7]
  8. Kentucky                    (3-0)    4.44 [62]   0.57 [71]   3.87 [ 8]
  9. Alabama                     (3-0)   50.96 [ 7]  47.39 [ 7]   3.56 [ 9]
 10. Georgia                     (3-0)   33.68 [17]  30.18 [18]   3.50 [10]

The Success ratings (formerly called Win/Loss, which doesn't consider score margin) aren't any better yet. Auburn isn't penalized for beating Mississippi State by a score of 3 to 2, but there's no major reason their schedule should be considered the most impressive. They have played three I-A teams, a rarity so far. Most of the other teams in the top ten fit this criteria, too.

The bottom line is, with the schedule played so far, there isn't enough good comparitive performance on which to judge the teams using no pre-conceived notions. That is, without looking at how good the teams were last season. Since these ratings use this season's data only, the entire picture is very fuzzy. East Carolina's high ranking might make it seem that things are starting to make sense, but that's  just a fluke at this point given the other discrepancies. It's the power rating equivalent of the VCR clock being right twice a day.

Next week, when every team will have played three games, and almost all will have had two I-A opponents, and several postponed games are made up, things will start to be clearer. At that point some insights can be drawn and perhaps a rough pecking order can be established. But it will really take until week 5 until the ratings are solid.

LSU #1 in final SportsRating College Football Power Rating

Louisiana State finished #1 in the final SportsRatings Power Rating after beating Ohio State for the BCS Championship. They were followed by West Virginia, the Points component champion, who beat Oklahoma 48-28 in the Fiesta Bowl. LSU finished fourth in the Points component and West Virginia finished fourth in the Win/Loss component.

Three Big 12 teams followed: Missouri, Oklahoma, and Kansas. The Tigers beat Arkansas 38-7 and finished second in the Win/Loss component; Oklahoma beat Missouri twice for the Tigers' only two losses but suffered three of their own including a loss to Colorado and the Fiesta Bowl defeat. Kansas lost only to Missouri and finished the season with an Orange Bowl win over Virginia Tech.

Three teams with chequered seasons followed. Southern Cal stumbled mid-season, losing to Stanford, but finished strong with a 49-17 Rose Bowl win over Illinois. Florida's offense played well all year, but the Gators lost four close games, end 14th in Win/Loss but sixth in Points. Oregon was at or near the top of the power rating with just a few weeks left when Dennis Dixon was hurt; the Ducks dropped three straight, only to recover with a 56-21 crushing of South Florida.

Georgia, who beat Hawaii 41-10, and Ohio State rounded out the top ten.

The Final SportsRatings College Football Power Rating for the 2007 season:

#   Team                       record   Rating      Points      Win/Loss
1.  LSU                         12-2    36.07 [ 1]  24.58 [ 4]  11.49 [ 1]
2.  West Virginia               11-2    34.81 [ 2]  26.84 [ 1]   7.97 [ 4]
3.  Missouri                    12-2    34.62 [ 3]  25.66 [ 3]   8.95 [ 2]
4.  Oklahoma                    11-3    33.88 [ 4]  26.53 [ 2]   7.34 [ 6]
5.  Kansas                      12-1    32.29 [ 5]  24.54 [ 5]   7.74 [ 5]
6.  Southern Cal                11-2    28.60 [ 6]  21.81 [ 8]   6.79 [ 7]
7.  Florida                      9-4    27.66 [ 7]  23.24 [ 6]   4.42 [14]
8.  Oregon                       9-4    27.61 [ 8]  22.61 [ 7]   5.00 [10]
9.  Georgia                     11-2    27.00 [ 9]  18.36 [11]   8.63 [ 3]
10. Ohio State                  11-2    26.49 [10]  21.33 [ 9]   5.15 [ 9]
11. South Florida                9-4    23.44 [11]  19.19 [10]   4.24 [15]
12. Cincinnati                  10-3    21.91 [12]  17.93 [12]   3.98 [18]
13. Virginia Tech               11-3    21.00 [13]  16.09 [14]   4.91 [11]
14. Arizona St                  10-3    20.06 [14]  15.46 [16]   4.60 [13]
15. Texas                       10-3    19.41 [15]  15.68 [15]   3.72 [21]
16. Clemson                      9-4    19.25 [16]  16.65 [13]   2.60 [29]
17. Texas Tech                   9-4    18.40 [17]  15.45 [17]   2.95 [25]
18. Tennessee                   10-4    18.02 [18]  12.24 [23]   5.77 [ 8]
19. Auburn                       9-4    17.76 [19]  13.93 [18]   3.82 [20]
20. Arkansas                     8-5    17.50 [20]  13.80 [19]   3.70 [22]
21. Oregon St                    9-4    16.03 [21]  12.43 [22]   3.59 [23]
22. Michigan                     9-4    15.97 [22]  12.53 [21]   3.43 [24]
23. Brigham Young               11-2    15.80 [23]  11.72 [25]   4.08 [17]
24. Penn State                   9-4    15.71 [24]  13.62 [20]   2.09 [35]
25. Kentucky                     8-5    15.29 [25]  11.31 [26]   3.97 [19]

#11 South Florida's Points total nearly put them in the top ten despite four losses; Big East stablemate #12 Cincinnati likewise scored well in Points. Virginia Tech played its best football in the 2nd half of the season to finish 13th. #14 Arizona State lost only to Oregon, USC, and #15 Texas, helping them to stay ahead of the Longhorns despite the Holiday Bowl loss.

Clemson's strong Points rating helps it to a #16 finish, while Tennessee's Win/Loss score puts it at #18. In between the two is the Big 12's Texas Tech, one of three teams to beat Oklahoma. SEC teams Auburn and Arkansas—one of the two teams to beat LSU—round out the top 20.

Oregon State, Michigan, BYU, Penn State, and Kentucky are 21 through 25.

Win/Loss component

The Win/Loss power rating, at the end of the year, gives a good representation of the success each team had during the year, independent of how dominant they were game-to-game.

Here LSU ends on top due to their tough schedule and many wins over quality teams; Missouri's schedule was not nearly as hard, but they lost only two games to Oklahoma, so their losses are penalized far less than LSU's. Still, the (LSU) Tigers come out on top. LSU's score is one of the lowest in any season for a #1 team in the Win/Loss component, as they are the first 2-loss team to top the rating.

Georgia's strong SEC schedule helped it to a final #3, while West Virginia's win over Oklahoma made up for the Pitt loss. Kansas had only one loss, and supplemented its meager schedule with a win over Virginia Tech.

Oklahoma's two wins over Missouri fueled the Sooners at #6 depite three losses, while USC had to overcome the Stanford loss. Tennessee again shows the strength of the SEC, with many wins offsetting four losses; Ohio State's two losses weren't to bad teams, but they didn't have many big victories to get higher than #9. Oregon had several huge wins, as well as many humbling losses, and it all shook out to to #10 in the end.

Win/Loss component of the power rating:

#   Team                       record  Win/Loss
1.  LSU                         12-2    11.49
2.  Missouri                    12-2     8.95
3.  Georgia                     11-2     8.63
4.  West Virginia               11-2     7.97
5.  Kansas                      12-1     7.74
6.  Oklahoma                    11-3     7.34
7.  Southern Cal                11-2     6.79
8.  Tennessee                   10-4     5.77
9.  Ohio State                  11-2     5.15
10. Oregon                       9-4     5.00
11. Virginia Tech               11-3     4.91
12. Hawaii                      12-1     4.71
13. Arizona St                  10-3     4.60
14. Florida                      9-4     4.42
15. South Florida                9-4     4.24
16. Boston College              11-3     4.24
17. Brigham Young               11-2     4.08
18. Cincinnati                  10-3     3.98
19. Kentucky                     8-5     3.97
20. Auburn                       9-4     3.82
21. Texas                       10-3     3.72
22. Arkansas                     8-5     3.70
23. Oregon St                    9-4     3.59
24. Michigan                     9-4     3.43
25. Texas Tech                   9-4     2.95

Hawaii places at #12, with their only loss to Georgia. The problem was their schedule didn't give them much in the way of wins to build a good rating. The highest-ranked team they beat was #42 Boise State. #19 Kentucky and #22 Arkansas finish high despite five losses because of their SEC schedule which for both teams included a win over #1 LSU.

To see the full lists click on the links above the listings.

SportsRatings College Football Power Rating - week 14

Big 12 Bonanza. Oklahoma's 2nd win over Missouri puts them at the top of the combined Power Rating this week. Missouri—with just two losses to Oklahoma on their record—remains #2. A third Big 12 team, Kansas, holds down 4th place. LSU moves up a notch to #3 despite losing ground in the Points component, and West Virginia takes a big tumble from #1 to #5 for losing to troubled Pitt. The Mountaineers lost a few points in the Points component, but got hit hard in Win/Loss, dropping from 3rd to 9th.

SportsRatings Overall Power Rating - week 14
lw  #  Team                         rec    Overall     Pts          W/L
3   1. Oklahoma                    11-2    34.66 [ 1]  27.78 [ 1]   6.87 [ 5]
2   2. Missouri                    11-2    31.78 [ 2]  24.82 [ 5]   6.95 [ 4]
4   3. LSU                         11-2    31.65 [ 3]  24.38 [ 6]   7.26 [ 3]
5   4. Kansas                      11-1    31.51 [ 4]  25.33 [ 3]   6.18 [ 7]
1   5. West Virginia               10-2    30.90 [ 5]  25.96 [ 2]   4.94 [ 9]
6   6. Ohio State                  11-1    30.54 [ 6]  23.14 [ 7]   7.39 [ 2]
7   7. Florida                      9-3    29.27 [ 7]  25.28 [ 4]   3.99 [13]
8   8. South Florida                9-3    26.58 [ 8]  22.69 [ 8]   3.89 [14]
11  9. Southern Cal                10-2    23.85 [ 9]  19.72 [ 9]   4.13 [12]
10 10. Georgia                     10-2    23.09 [10]  17.69 [13]   5.39 [ 8]
12 11. Cincinnati                   9-3    22.83 [11]  19.63 [10]   3.19 [18]
15 12. Virginia Tech               11-2    22.56 [12]  16.16 [14]   6.39 [ 6]
9  13. Oregon                       8-4    21.98 [13]  19.17 [11]   2.81 [22]
14 14. Clemson                      9-3    20.82 [14]  17.69 [12]   3.13 [19]
13 15. Arizona St                  10-2    20.64 [15]  16.16 [15]   4.47 [10]
16 16. Arkansas                     8-4    18.47 [16]  15.74 [17]   2.73 [24]
17 17. Texas Tech                   8-4    18.06 [17]  15.75 [16]   2.30 [31]
18 18. Illinois                     9-3    17.83 [18]  13.95 [21]   3.88 [15]
21 19. Texas                        9-3    17.00 [19]  14.28 [19]   2.72 [25]
20 20. Auburn                       8-4    16.65 [20]  14.04 [20]   2.60 [29]
23 21. Penn State                   8-4    16.54 [21]  14.41 [18]   2.12 [33]
24 22. Hawaii                      12-0    16.43 [22]   6.52 [48]   9.90 [ 1]
19 23. Boston College              10-3    15.71 [23]  11.46 [28]   4.24 [11]
22 24. Tennessee                    9-4    15.64 [24]  12.37 [22]   3.27 [17]
27 25. Brigham Young               10-2    14.81 [25]  11.26 [29]   3.55 [16]

The rest of the list shuffles very slightly here and there. Hawaii moves up a couple of places to #22, but in the Win/Loss component they move up to #1. The Warriors have a low rating for an undefeated team, reflecting their schedule. Ohio State's rating is pretty low for a 1-loss team, but still reflects a tougher-than-average schedule for Division I-A. Kansas' rating is even lower, despite their only loss being to Missouri; they aren't being downgraded much for the loss, they just don't have a lot of 'win points,' and the team they lost to, Missouri, suffered its second defeat, falling from 1st to 4th.

Power Rating: Win/loss component
lw  #  Team                         rec    rating
2   1. Hawaii                      12-0     9.90
4   2. Ohio State                  11-1     7.39
5   3. LSU                         11-2     7.26
1   4. Missouri                    11-2     6.95
7   5. Oklahoma                    11-2     6.87
9   6. Virginia Tech               11-2     6.39
6   7. Kansas                      11-1     6.18
8   8. Georgia                     10-2     5.39
3   9. West Virginia               10-2     4.94
13 10. Arizona St                  10-2     4.47
10 11. Boston College              10-3     4.24
16 12. Southern Cal                10-2     4.13
12 13. Florida                      9-3     3.99
11 14. South Florida                9-3     3.89
15 15. Illinois                     9-3     3.88
17 16. Brigham Young               10-2     3.55
14 17. Tennessee                    9-4     3.27
21 18. Cincinnati                   9-3     3.19
19 19. Clemson                      9-3     3.13
20 20. Connecticut                  9-3     3.09
22 21. Wisconsin                    9-3     2.89
18 22. Oregon                       8-4     2.81
23 23. Boise St                    10-2     2.77
25 24. Arkansas                     8-4     2.73
28 25. Texas                        9-3     2.72 

The "Standalone" Win/Loss component values schedule more greatly than the one that is combined with the Points rating to form the Overall Power Rating. In the Standalone ranking, LSU's schedule is sufficient to overcome its losses and take the #1 spot, even though the algorithm doesn't know how narrow its losses were.

Power Rating: "Standalone" Win/loss component
lw  #  Team                         rec    rating
3   1. LSU                         11-2    10.14
7   2. Oklahoma                    11-2     8.36
6   3. Hawaii                      12-0     7.82
1   4. Missouri                    11-2     7.75
9   5. Virginia Tech               11-2     7.57
4   6. Ohio State                  11-1     7.31
5   7. Georgia                     10-2     6.97
2   8. West Virginia               10-2     6.13
11  9. Kansas                      11-1     5.73
15 10. Southern Cal                10-2     5.67
14 11. Arizona St                  10-2     5.51
8  12. Boston College              10-3     5.18
12 13. Florida                      9-3     5.09
13 14. South Florida                9-3     4.80
10 15. Tennessee                    9-4     4.67
17 16. Illinois                     9-3     4.13
18 17. Cincinnati                   9-3     4.05
16 18. Oregon                       8-4     3.93
20 19. Arkansas                     8-4     3.82
21 20. Connecticut                  9-3     3.57
19 21. Clemson                      9-3     3.53
25 22. Kentucky                     7-5     3.39
22 23. Brigham Young               10-2     3.36
23 24. Auburn                       8-4     3.32
24 25. Virginia                     9-3     3.09 

A two-loss team has never finished at the top of the standalone Win/Loss Power Rating, as far as I can recall. Even for 1990, a year where almost every team had 2 losses, 11-1-1 Colorado and 11-0-1 Georgia Tech finish #1 and #2.

In the Points component:

Power Rating: Points (score margin) component
lw  #  Team                         rec    rating  off    def
2   1. Oklahoma                    11-2    27.78  43.01  15.22
1   2. West Virginia               10-2    25.96  41.62  15.66
6   3. Kansas                      11-1    25.33  40.32  14.99
4   4. Florida                      9-3    25.28  47.75  22.47
3   5. Missouri                    11-2    24.82  44.34  19.51
5   6. LSU                         11-2    24.38  42.18  17.79
7   7. Ohio State                  11-1    23.14  32.16   9.02
8   8. South Florida                9-3    22.69  41.10  18.40
11  9. Southern Cal                10-2    19.72  33.07  13.35
10 10. Cincinnati                   9-3    19.63  37.79  18.15

Oklahoma and West Virginia switch places, and the Sooners have a big lead. Last week, the Mountaineers had a sizeable lead, and Oklahoma and Missouri were nearly tied for 2nd place. Kansas' rating is unchanged, but they move up as Missouri falls, and Florida loses a small amount of points from random reshuffling. It's not random, of course, but since every game has a slight pull on everyone, what's remarkable is Kansas' rating not moving, since one of their opponents lost points itself. Two of Florida's opponents (Troy, Tennessee) lost, but they were beaten by two other Gator opponents (LSU, Florida Atlantic).

719-team Power Rating: Division I-A, I-AA, II, III, NAIA

In order to rate the Division I-AA teams during the playoff season, I ran my Points-based power rating subcomponent on a file containing all games among teams in all divisions. The resultant rating has 719 teams, and the additional rating affects the position of some Division I-A teams as well.

For the SportsRatings College Football Power Rating, also referred to as the "overall" power rating, I used only games involving Division I-A teams. This introduces a subtle penalty for teams that play Division I-AA teams, which is normally a given win. It's difficult to gain points by playing such teams under this system, which is how it should be.

When the file is expanded to include the other divisions, a team's rating may adjust somewhat, as information is added about the lower division opponent's quality.

Here is an example using two prominent teams, Michigan and Hawaii:

num  Team                          rec   rating  off     def
35.  Michigan                      8-4    50.40  47.17  -3.22
36.  Hawaii                       11-0    49.57  60.14  10.57

As it turns out, they end up right next to each other in the expanded rating. In the Division I only rating, Michigan is 27th and Hawaii is 51st, so there is quite an effect for these specific teams. Adding more information actually diminishes Appalachian State, making Michigan's loss to them worse. This makes sense if you think about it: the Michigan game was certainly their best win, and adding in their other performances can only draw them down, thereby pulling the Wolverines' rating down slightly.

Hawaii played and defeated two Division I-AA opponents, beating Northern Colorado 63-6, and Charleston Southern 66-10. While it would seem that these lopsided victories would inflate Hawaii's rating, when taken out of context there is an equal reaction to degrade the opposition. These were clearly the worst defeat either team suffered; when their other games are added in, their rating improves, thus Hawaii's wins look more impressive.

The upshot? Without the full Division I-AA information, Michigan rates a 5.5 point favorite on a neutral field over the Warriors; with the extra info, they're about a 1 point favorite.

Another interesting result is that average scores become exagerrated since the "average team" is a poor Division II team, so large offensive scores and negative defensive averages are common.

At a certain point in the ratings, non-Division I-A teams start to show up:

num  Team                         rec   rating  off     def
58. Northern Iowa                12-0    43.78  46.81   3.02
59. Fresno St                     7-4    43.16  49.74   6.58
60. Mississippi St                7-5    42.65  44.70   2.04
61. Nebraska-Omaha               10-1    42.10  47.27   5.17
62. Indiana                       7-5    41.62  49.17   7.54
63. North Carolina                4-8    41.58  44.91   3.32
64. Washington St                 5-7    41.41  49.09   7.67
65. New Mexico                    8-4    41.38  43.69   2.31
66. Pittsburgh                    4-7    40.98  43.39   2.41
67. Ball St                       7-5    39.86  46.13   6.26
68. Iowa                          6-6    39.60  37.56  -2.03
69. Tulsa                         9-3    39.41  55.49  16.08
70. Appalachian St               10-2    39.23  52.30  13.07
71. Massachusetts                10-2    39.10  44.20   5.09
72. NW Missouri St               10-1    39.04  48.18   9.13
73. North Dakota                 10-2    38.98  48.51   9.53
74. Richmond                     10-2    38.82  48.66   9.83
75. Mississippi                   3-9    38.66  42.93   4.27
76. North Dakota St              10-1    38.45  48.87  10.42
77. Southern Illinois            11-1    38.14  47.86   9.72

Northern Iowa, the top-ranked team in the polls and the #1 seed in the Division I-AA playoff, is also the highest-rated team in the power rating at #58. Note that they rate higher than the two Division I-A teams in the state, Iowa at #68 and Iowa State at #92, whom the Panthers defeated earlier in the year.

Other non-Division I-A teams in this range are highlighted. Nebraska-Omaha is a Division II team, and they recently lost in the Division II playoffs when #103 Central Washington scored two touchdowns in the final minutes. Northwest Missouri State is also Division II, along with North Dakota. Thus there isn't too much difference between the top Division II teams and the best of Division I-AA. At #70, of course, is defending champ Appalachian State, a team that looked like the sure #1 after beating Michigan. They lost twice this season, but are in the final eight in the playoffs.

Another interesting comparison: The top Division II team in the polls is undefeated Grand Valley State. They rank three spots higher than Notre Dame this year, a result which has to be a first for the Irish.

num  Team                         rec   rating  off     def
93. Grand Valley St              11-0    34.68  42.20   7.51
94. Nevada                        5-6    34.44  50.48  16.03
95. McNeese St                   11-1    34.35  42.31   7.96
96. Notre Dame                    3-9    34.27  39.20   4.93

And finally, what everybody wants to see after viewing any ratings for a few minutes, is, who are the lowest-rated teams? So here are the bottom ten. Remember, only score margin is used for this power rating so winning and losing per sé doesn't influence a team's rating.

num  Team                          rec   rating  off     def
710. Lewis & Clark                 0-9   -62.23   4.39  66.63
711. SUNY-Maritime                 2-8   -63.18  -1.99  61.19
712. Gallaudet                     2-6   -63.68  -7.80  55.87
713. St Vincent                    0-10  -66.75  -1.05  65.69
714. Eureka                        2-8   -70.74  -7.58  63.16
715. Blackburn                     4-6   -70.84  -3.47  67.37
716. MacMurray                     0-10  -79.04 -10.82  68.22
717. Maranatha Baptist             2-7   -81.24 -15.97  65.26
718. Trinity Bible                 1-8   -86.39 -18.29  68.10
719. Principia                     1-8   -96.54 -17.91  78.63

Principia, a St. Louis-area college, rates as the worst team in America. They did manage one victory this season—over #718 Trinity Bible at home, but they lost a rematch on a neutral field, 45-20. That was the sixth-fewest points they allowed all season, as well. According to these numbers, the average team is nearly 100 points better than Principia, though they'd only be favored by 79 since the Panthers offense  score is deeply negative (suggesting a team with a defense 18 points worse than average could still expect to shut them out).

Who is this "average" team then? The closest to meeting that designation is #321 Muhlenberg, with a 0.07 overall rating. The Mules are a Division III team, so being the average team among all divisions is an achievement. As expected, they outperformed in their Division, and finished their season 11-1, losing in the second round of the Division III playoffs.

Division I-AA Playoffs - first round recap

The Division I-AA playoffs have finished their first round, and so far there has been only one upset, Eastern Washington destroying McNeese State. This is according to a power rating analysis of the Division I-AA teams. If you go by the seedings, there were two upsets; the aforementioned game, and also when 3rd seeded Montana lost to Wofford. Using the polls as your guide, their was a third, minor upset, when Delaware beat Delaware State.

The top seed in the 16-team playoff is Northern Iowa (11-0 going in), and they narrowly escaped against  New Hampshire, 38-35. This was played in the Uni-dome in Cedar Falls, Iowa, as the top seeds plays at home. The home field might have been the difference here. So what happened? As it turns out, New Hampshire, only 7-4 before the tournament, is much better than their record. Northern Iowa simply got a bad draw. Three of their four losses were to other playoff teams, and they had a win against a fourth, Delaware, as well as a win over Division I-A Marshall.

Here's a list of the 16 teams that started the tournament, and their power ratings (pre-first round):

#   rnk Team                         rec   rating
1   57. Northern Iowa               11-0   46.02 
2   71. Massachusetts                9-2   40.40 
3   72. Appalachian St               9-2   40.39 
4   74. Southern Illinois           10-1   39.92 
5   77. Richmond                     9-2   39.30 
6   79. McNeese St                  11-0   38.86 
7   86. James Madison                8-3   36.87 
8   88. Delaware                     8-3   36.61 
9   99. Wofford                      8-3   34.42 
10 110. New Hampshire                7-4   32.36 
11 133. Montana                     11-0   29.19 
12 142. Eastern Washington           8-3   27.57 
13 165. Eastern Kentucky             9-2   23.45 
14 201. Fordham                      8-3   16.88 
15 208. Eastern Illinois             8-3   16.17 
16 214. Delaware St                 10-1   15.60 

Delaware (8-3) played the worst team in the draw, Delaware State (10-1) and beat them 44-7, in the first ever meeting (weird, huh?) between the two schools. The teams played because there is a regional aspect to the playoffs, and only 4 teams are seeded. Delaware will play at UNI on Saturday.

4th seed Southern Illinois beat Eastern Illinois as expected, 30-11, and Massachusetts took out Fordham 49-35; that takes care of the upper bracket.

The lower bracket had a few surprises. McNeese State, playing at home, fell to Eastern Washington 44-15. Though McNeese was #6 in the field according to the power rating, they were essentially even with teams #2 through #5, and they were undefeated, while Eastern Washington was 12th best in the playoff. Definitely an upset, especially the magnitude of the victory.

Eastern Washington will face another tough team in Appalachian State, famous upset winner over Michigan to open the season, who survived a scare from James Madison. Both are among the top-rated teams in the playoff, so this was a key game, and went down to the wire, 28-27.

Though undefeated, Montana was a truly weak 3rd seed, and should have been a very slight underdog to Wofford even with home field advantage. This turned out to be true as the final was 23-22 on a wild finish, a Wofford touchdown with 30 seconds to go, and a missed Montana field goal at the finish.

Richmond is the last team in the final eight after beating Eastern Kentucky 31-14. This leaves, in the final eight, teams that rate 1, 2, 3, 4, 5, 8, 9, and 12 among the full 16. These comparative rankings may change with the updated power rating. The full 662-team power rating can be found in the sidebar.

SportsRatings College Football Power Rating - week 13

West Virginia #1, Missouri #2. The Mountaineers' domination of Connecticut allows them to inherit the #1 spot vacated by Kansas, who falls to #5. Missouri is a close second, and Oklahoma (who also benefits from Missouri's win) jumps to third. LSU slips to fourth after their loss to Arkansas.

SportsRatings Overall Power Rating - week 13
lw  #  Team                         rec    Overall     Pts          W/L
4   1. West Virginia               10-1    36.60 [ 1]  28.83 [ 1]   7.77 [ 3]
3   2. Missouri                    11-1    36.05 [ 2]  26.58 [ 3]   9.47 [ 1]
6   3. Oklahoma                    10-2    32.41 [ 3]  26.67 [ 2]   5.73 [ 7]
2   4. LSU                         10-2    31.58 [ 4]  25.38 [ 5]   6.19 [ 5]
1   5. Kansas                      11-1    31.50 [ 5]  25.33 [ 6]   6.17 [ 6]
5   6. Ohio State                  11-1    30.56 [ 6]  23.10 [ 7]   7.46 [ 4]
7   7. Florida                      9-3    29.89 [ 7]  25.68 [ 4]   4.21 [12]
9   8. South Florida                9-3    26.72 [ 8]  22.49 [ 8]   4.22 [11]
8   9. Oregon                       8-3    24.53 [ 9]  21.11 [ 9]   3.41 [18]
12 10. Georgia                     10-2    23.85 [10]  18.19 [12]   5.66 [ 8]
13 11. Southern Cal                 9-2    23.19 [11]  19.62 [11]   3.57 [16]
11 12. Cincinnati                   9-3    22.65 [12]  19.67 [10]   2.97 [21]
10 13. Arizona St                   9-2    21.28 [13]  17.13 [14]   4.15 [13]
14 14. Clemson                      9-3    20.89 [14]  17.69 [13]   3.20 [19]
15 15. Virginia Tech               10-2    20.77 [15]  15.52 [17]   5.25 [ 9]
23 16. Arkansas                     8-4    19.01 [16]  16.30 [15]   2.70 [25]
18 17. Texas Tech                   8-4    18.11 [17]  15.96 [16]   2.15 [31]
17 18. Illinois                     9-3    17.93 [18]  14.01 [21]   3.91 [15]
19 19. Boston College              10-2    17.60 [19]  12.49 [22]   5.10 [10]
21 20. Auburn                       8-4    17.10 [20]  14.43 [18]   2.66 [26]
16 21. Texas                        9-3    16.86 [21]  14.23 [20]   2.62 [28]
26 22. Tennessee                    9-3    16.47 [22]  12.42 [23]   4.04 [14]
24 23. Penn State                   8-4    16.38 [23]  14.25 [19]   2.12 [33]
38 24. Hawaii                      11-0    15.25 [24]   6.15 [51]   9.09 [ 2]
22 25. Boise St                    10-2    14.67 [25]  11.82 [25]   2.84 [23]
 
Hawaii jumps 14 spots and lands at 24th, one position ahead of the Boise State Broncos, who they defeated on Friday. Arkansas also climbs quite a bit for beating LSU, while Connecticut falls nine spots to 29th after the disastrous West Virginia game.

Power Rating: Win-loss component
lw  #  Team                         rec    rating
5   1. Missouri                     11-1     9.47
4   2. Hawaii                       11-0     9.09
6   3. West Virginia                10-1     7.77
3   4. Ohio State                   11-1     7.46
2   5. LSU                          10-2     6.19
1   6. Kansas                       11-1     6.17
9   7. Oklahoma                     10-2     5.73
8   8. Georgia                      10-2     5.66
12  9. Virginia Tech                10-2     5.25
11 10. Boston College               10-2     5.10
18 11. South Florida                 9-3     4.22
16 12. Florida                       9-3     4.21
7  13. Arizona St                    9-2     4.15
20 14. Tennessee                     9-3     4.04
13 15. Illinois                      9-3     3.91
24 16. Southern Cal                  9-2     3.57
23 17. Brigham Young                 9-2     3.45
10 18. Oregon                        8-3     3.41
27 19. Clemson                       9-3     3.20
14 20. Connecticut                   9-3     3.11
21 21. Cincinnati                    9-3     2.97
22 22. Wisconsin                     9-3     2.90
15 23. Boise St                     10-2     2.84
25 24. Michigan                      8-4     2.74
40 25. Arkansas                      8-4     2.70 

Undefeated Hawaii ranks #2 in the Win/Loss component. The purpose of the Win/Loss component is to smooth out the results of the Points rating, insuring that teams that are winning get rewarded while teams with losses are penalized. While it doesn't push Hawaii very high in the overall power rating, it does move them up to a respectable top 25 placing.

Last week's #1 Kansas takes a big fall as they lose their undefeated status, and Arkansas gains a lot of ground for taking out LSU, a win that was valuable in both components.

While the above component is structured to be combined with the Points component, with a slight modification, it can be used (later in the season) to form a good "standalone" rating.

Power Rating: Modified "Standalone" Win-loss component
#   Team                         rec    rating
1.  Missouri                    11-1    10.28
2.  West Virginia               10-1     9.12
3.  LSU                         10-2     8.34
4.  Ohio State                  11-1     7.45
5.  Georgia                     10-2     7.35
6.  Hawaii                      11-0     7.10
7.  Oklahoma                    10-2     6.68
8.  Boston College              10-2     6.16
9.  Virginia Tech               10-2     5.97
10. Tennessee                    9-3     5.72
11. Kansas                      11-1     5.64
12. Florida                      9-3     5.46
13. South Florida                9-3     5.22
14. Arizona St                   9-2     5.21
15. Southern Cal                 9-2     5.05
16. Oregon                       8-3     4.79
17. Illinois                     9-3     4.21
18. Cincinnati                   9-3     3.70
19. Clemson                      9-3     3.70
20. Arkansas                     8-4     3.64
21. Connecticut                  9-3     3.50
22. Brigham Young                9-2     3.45
23. Auburn                       8-4     3.43
24. Virginia                     9-3     3.03
25. Kentucky                     7-5     3.01 

In this version Hawaii is at #6 and Kansas at #11. These two are pretty much reversed in the national polls, reflecting how voters incorporate margin of victory into their analyses. Kansas does not get a lot of "win points" for its schedule; neither does Hawaii, but they aren't hampered by a loss, and the Boise State win did as much for them as the rest of their schedule combined. Also note LSU rates ahead of Ohio State, and Georgia ahead of Hawaii, reflecting strength of schedule as well. The SEC fares well in general, with 3-loss Tennessee and Florida both in the top 12, and seven teams overall.

But back to the main power rating, and the Points component

Power Rating: Points component
lw  #  Team                         rec    rating  off    def
3   1. West Virginia               10-1    28.83  44.53  15.69
5   2. Oklahoma                    10-2    26.67  42.94  16.26
4   3. Missouri                    11-1    26.58  45.84  19.25
6   4. Florida                      9-3    25.68  47.87  22.18
1   5. LSU                         10-2    25.38  44.15  18.76
2   6. Kansas                      11-1    25.33  40.25  14.92
9   7. Ohio State                  11-1    23.10  32.19   9.08
8   8. South Florida                9-3    22.49  41.02  18.52
7   9. Oregon                       8-3    21.11  40.84  19.72
10 10. Cincinnati                   9-3    19.67  37.95  18.28

LSU and Kansas both suffer big drops for losing, but still rank among the top teams. Their new ratings reflect the new information gained last week, which doesn't drastically alter a reasonable opinion of how good they are. Oregon slips as well, but in their case it's a different matter. Clearly the Oregon of today isn't the same team as it was for most of the year. This is, of course, a drawback of power ratings. Each week Oregon plays in its hobbled state, the ratings are slightly "polluted" by basically having a mediocre team masquerade as a good one. UCLA gets much more credit than it should for beating the Ducks 16-0, as did Arizona for their win. To a lesser extent, teams that lost to Oregon get penalized too much, as Oregon's overall rating goes down. To account for this, either more information would need to be available to the rating, or manual adjustments (such as splitting the post-Dixon Oregon into its own team) would need to be made, but this leads to subjective calls and is rather unwieldy. It's just one of the many imperfections in any real-world rating system that have to be lived with.

Offense
#   Team                         rec    rating  off    def
1.  Florida                      9-3    25.68  47.87  22.18
2.  Missouri                    11-1    26.58  45.84  19.25
3.  West Virginia               10-1    28.83  44.53  15.69
4.  LSU                         10-2    25.38  44.15  18.76
5.  Oklahoma                    10-2    26.67  42.94  16.26
6.  Arkansas                     8-4    16.30  41.41  25.10
7.  South Florida                9-3    22.49  41.02  18.52
8.  Oregon                       8-3    21.11  40.84  19.72
9.  Kansas                      11-1    25.33  40.25  14.92
10. Texas Tech                   8-4    15.96  39.01  23.04

Oregon's offense takes a dive for being shut out. As said above, the current rating doesn't reflect Oregon's current offensive capability, and the downgrade is unfair to the earlier version of the team. Another systemic flaw: Arkansas' offense gets a boost for scoring 50 points on LSU, though of course a large amount of these were in the three overtimes. This one could be corrected, but most score sources don't even note overtime games, let alone the score at the end of regulation.

Defense
#   Team                         rec    rating  off    def
1.  Ohio State                  11-1    23.10  32.19   9.08
2.  Auburn                       8-4    14.43  26.89  12.45
3.  Southern Cal                 9-2    19.62  33.33  13.71
4.  Kansas                      11-1    25.33  40.25  14.92
5.  Virginia Tech               10-2    15.52  30.75  15.23
6.  West Virginia               10-1    28.83  44.53  15.69
7.  Michigan                     8-4    11.65  27.90  16.24
8.  Oklahoma                    10-2    26.67  42.94  16.26
9.  Utah                         8-4    10.03  27.00  16.97
10. Georgia                     10-2    18.19  35.97  17.78

A new face on the best-scoring-defense list is Utah, who held BYU to 17 points in a losing effort. I haven't looked at the yardage stats for that game but it should be interesting to see whether BYU had its usual dominating spread in total yardage or not.

SportsRatings College Football Power Rating - week 12: Neck and neck

KU back on top. LSU edged past Kansas last week, and KU returns the favor this week. LSU maintains a tiny edge in Points while Kansas opens a larger lead in Win/Loss, giving them just over one point in overall advantage. The Jayhawks' next two opponents are #3 Missouri, and (if they win that) #6 Oklahoma, so if they win out they'll easily retain the #1 spot. LSU faces #23 Arkansas, then either #26 Tennessee or #12 Georgia.

SportsRatings Overall Power Rating - week 12
lw  #  Team                         rec    Overall     Pts         W/L
2   1. Kansas                     (11-0)   36.63 [ 1]  26.78 [ 2]   9.84 [ 1]
1   2. LSU                        (10-1)   35.27 [ 2]  26.81 [ 1]   8.46 [ 2]
5   3. Missouri                   (10-1)   33.04 [ 3]  26.02 [ 4]   7.01 [ 5]
6   4. West Virginia               (9-1)   32.71 [ 4]  26.50 [ 3]   6.20 [ 6]
7   5. Ohio State                 (11-1)   31.00 [ 5]  23.32 [ 9]   7.68 [ 3]
4   6. Oklahoma                    (9-2)   30.11 [ 6]  25.32 [ 5]   4.79 [ 9]
8   7. Florida                     (8-3)   28.50 [ 7]  24.74 [ 6]   3.75 [16]
3   8. Oregon                      (8-2)   28.26 [ 8]  23.56 [ 7]   4.70 [10]
10  9. South Florida               (8-3)   27.06 [ 9]  23.48 [ 8]   3.57 [18]
9  10. Arizona St                  (9-1)   25.18 [10]  19.33 [11]   5.84 [ 7]
11 11. Cincinnati                  (8-3)   23.51 [11]  20.56 [10]   2.94 [21]
13 12. Georgia                     (9-2)   22.79 [12]  17.64 [14]   5.14 [ 8]
14 13. Southern Cal                (8-2)   20.69 [13]  17.81 [13]   2.88 [24]
12 14. Clemson                     (8-3)   20.67 [14]  17.93 [12]   2.74 [27]
16 15. Virginia Tech               (9-2)   19.53 [15]  15.16 [17]   4.36 [12]
15 16. Texas                       (9-2)   19.09 [16]  15.56 [16]   3.53 [19]
17 17. Illinois                    (9-3)   18.10 [17]  14.12 [21]   3.98 [13]
23 18. Texas Tech                  (8-4)   17.85 [18]  15.84 [15]   2.01 [32]
28 19. Boston College              (9-2)   17.27 [19]  12.79 [25]   4.47 [11]
19 20. Connecticut                 (9-2)   16.98 [20]  13.08 [23]   3.89 [14]
20 21. Auburn                      (7-4)   16.96 [21]  14.81 [18]   2.15 [29]
25 22. Boise St                   (10-1)   16.87 [22]  12.99 [24]   3.88 [15]
21 23. Arkansas                    (7-4)   16.50 [23]  14.78 [19]   1.71 [40]
18 24. Penn State                  (8-4)   16.39 [24]  14.27 [20]   2.12 [31]
22 25. Kentucky                    (7-4)   15.29 [25]  12.54 [26]   2.74 [26]
 

Texas Tech and Boston College make the biggest moves among the highest rated of the 120 Div I-A teams. Tech's upset of Oklahoma garners them five spots and makes them the highest-rated 4-loss team. Boston College greatly improves its Points score with a win over Clemson.

Power Rating: Win-loss component
lw  #  Team                         rec    rating
1   1. Kansas                     (11-0)    9.84
2   2. LSU                        (10-1)    8.46
4   3. Ohio State                 (11-1)    7.68
6   4. Hawaii                     (10-0)    7.06
7   5. Missouri                   (10-1)    7.01
9   6. West Virginia               (9-1)    6.20
8   7. Arizona St                  (9-1)    5.84
10  8. Georgia                     (9-2)    5.14
5   9. Oklahoma                    (9-2)    4.79
3  10. Oregon                      (8-2)    4.70
14 11. Boston College              (9-2)    4.47
11 12. Virginia Tech               (9-2)    4.36
12 13. Illinois                    (9-3)    3.98
15 14. Connecticut                 (9-2)    3.89
22 15. Boise St                   (10-1)    3.88
17 16. Florida                     (8-3)    3.75
13 17. Virginia                    (9-2)    3.66
24 18. South Florida               (8-3)    3.57
20 19. Texas                       (9-2)    3.53
23 20. Tennessee                   (8-3)    3.32
18 21. Cincinnati                  (8-3)    2.94
25 22. Wisconsin                   (9-3)    2.92
29 23. Brigham Young               (8-2)    2.89
27 24. Southern Cal                (8-2)    2.88
16 25. Michigan                    (8-4)    2.78 

The Win-Loss rating looks basically similar to the AP or coaches poll with some slight variations. One exception is Hawaii's ranking, which is much higher than in the national polls. With two games left—one against 10-1 Boise State, Hawaii could conceivably finish at the top of this component if a few other teams lose. Clearly, no one considers Hawaii to be this good.

This example is exactly why I use a combined system for my overall ratings. In the overall (combined) power rating, Hawaii is #38. The Win/Loss component I use is "pure" in terms of how it weights wins; the resultant ranking is really not meant to be used alone, but for a combined system. But the national polls don't leave Hawaii unranked, nor do the put them in the top five. Using a slightly modified Win/Loss system does an excellent job as a "standalone" ranking, and back-tested through the years yields very agreeable results. Where does Hawaii rank in that system?

Power Rating: Modified "Standalone" Win-loss component
#   Team                         rec    rating
1.  LSU                        (10-1)   10.64
2.  Kansas                     (11-0)    8.86
3.  Ohio State                 (11-1)    7.93
4.  Missouri                   (10-1)    7.56
5.  West Virginia               (9-1)    7.17
6.  Arizona St                  (9-1)    7.04
7.  Georgia                     (9-2)    6.57
8.  Oregon                      (8-2)    6.42
9.  Oklahoma                    (9-2)    5.51
10. Boston College              (9-2)    5.32
11. Hawaii                     (10-0)    5.17
12. Virginia Tech               (9-2)    4.74
13. Florida                     (8-3)    4.72
14. Illinois                    (9-3)    4.45
15. Tennessee                   (8-3)    4.38
16. South Florida               (8-3)    4.30
17. Virginia                    (9-2)    4.29
18. Connecticut                 (9-2)    4.29
19. Southern Cal                (8-2)    3.77
20. Cincinnati                  (8-3)    3.74
21. Kentucky                    (7-4)    3.71
22. Texas                       (9-2)    3.49
23. Boise St                   (10-1)    3.25
24. Clemson                     (8-3)    3.06
25. Brigham Young               (8-2)    3.02 

Hawaii lands at #11, which is slightly higher than in the national polls, but much more in line with where people think they should be ranked than in either my overall power rating or its subcomponents.

The problem with Hawaii is that they're extremely hard to rate fairly. A system that doesn't use margin of victory isn't going to see how they're just scraping by against a very easy schedule, and might rate them too high; but if they were blowing out every opponent and truly looked impressive, that would have no effect on their rating, and systems that choose to penalize them for their poor schedule would end up ranking them too low.

Using point spreads, we get an idea of how good Hawaii actually is: #57, or just above average among the 120 Division I-A teams. But they keep winning, and are now 10-0. We'll see what happens when they meet Boise State, another team that's hard to rank but is probably a lot better than Hawaii. Should the Warriors win that contest (which is played in Honolulu) and win out, they have a good chance of getting in a BCS bowl game, which they might deserve simply for going undefeated. I'm not sure they'll make a good game of it like Boise State-Oklahoma last year.

Power Rating: Points component
lw  #  Team                         rec    rating  off    def
1   1. LSU                        (10-1)   26.81  43.48  16.67
2   2. Kansas                     (11-0)   26.78  41.36  14.57
5   3. West Virginia               (9-1)   26.50  41.58  15.08
6   4. Missouri                   (10-1)   26.02  45.77  19.75
4   5. Oklahoma                    (9-2)   25.32  42.24  16.92
7   6. Florida                     (8-3)   24.74  47.95  23.20
3   7. Oregon                      (8-2)   23.56  44.37  20.81
9   8. South Florida               (8-3)   23.48  39.94  16.46
8   9. Ohio State                 (11-1)   23.32  32.32   8.99
10 10. Cincinnati                  (8-3)   20.56  37.69  17.12

LSU tops the Points component by a hair. Oregon falls quite a bit more than Oklahoma for their loss.

Offense
#   Team                         rec    rating  off    def
1.  Florida                     (8-3)   24.74  47.95  23.20
2.  Missouri                   (10-1)   26.02  45.77  19.75
3.  Oregon                      (8-2)   23.56  44.37  20.81
4.  LSU                        (10-1)   26.81  43.48  16.67
5.  Oklahoma                    (9-2)   25.32  42.24  16.92
6.  West Virginia               (9-1)   26.50  41.58  15.08
7.  Kansas                     (11-0)   26.78  41.36  14.57
8.  South Florida               (8-3)   23.48  39.94  16.46
9.  Arkansas                    (7-4)   14.78  39.56  24.78
10. Texas Tech                  (8-4)   15.84  39.22  23.38

Oregon's offense slides a notch, while Texas Tech jumps into the top ten after shredding Oklahoma's D.

Defense
#   Team                         rec    rating  off    def
1.  Ohio State                 (11-1)   23.32  32.32   8.99
2.  Auburn                      (7-4)   14.81  27.62  12.81
3.  Southern Cal                (8-2)   17.81  31.25  13.44
4.  Virginia Tech               (9-2)   15.16  29.67  14.50
5.  Kansas                     (11-0)   26.78  41.36  14.57
6.  West Virginia               (9-1)   26.50  41.58  15.08
7.  Michigan                    (8-4)   11.78  27.66  15.87
8.  Connecticut                 (9-2)   13.08  29.47  16.38
9.  South Florida               (8-3)   23.48  39.94  16.46
10. Arizona St                  (9-1)   19.33  35.81  16.47

Ohio State held Michigan to three points in Ann Arbor, so there was little doubt they'd stay the #1 scoring defense. But Michigan's D performed well, too, allowing only two touchdowns, and they move up to #7. LSU was #6 last week but fell to eleventh after allowing Mississippi to score 24 points.

SportsRatings College Football Power Rating - week 11

LSU back at #1. Since last week's #2 Ohio State fell, it's no surprise that former #3 LSU moved up, but they also inched past Kansas to take over the #1 spot in this week's SportsRatings Power Rating. With the top four from last week in a virtual dead heat and Oregon idle, the top spot was a battle between the undefeated Jayhawks and the once-beaten Tigers. LSU clobbered a weak Lousiana Tech team, which didn't gain them a lot of points in either component, while Kansas beat Oklahoma State 43-28. Though they gained in the Win/Loss component, they lost a bit of ground in the Points component and now trail LSU by just .08.

Ohio State lost rating points in both components, but moreso, naturally, in the Win/Loss subcomponent, where last week they were #1. Oregon, Oklahoma, Missouri, and West Virginia all slide up  a notch, while Florida, Arizona State, and South Florida remain in place to round out the top ten.

SportsRatings Overall Power Rating - week 11
lw  #  Team                         rec    Overall     Pts         W/L
3   1. LSU                         (9-1)   36.30 [ 1]  27.97 [ 1]  8.32 [ 2]
1   2. Kansas                     (10-0)   36.22 [ 2]  26.75 [ 2]  9.47 [ 1]
4   3. Oregon                      (8-1)   34.21 [ 3]  26.62 [ 3]  7.59 [ 3]
5   4. Oklahoma                    (9-1)   33.14 [ 4]  26.62 [ 4]  6.51 [ 5]
6   5. Missouri                    (9-1)   31.61 [ 5]  25.37 [ 6]  6.24 [ 7]
7   6. West Virginia               (8-1)   30.99 [ 6]  25.98 [ 5]  5.00 [ 9]
2   7. Ohio State                 (10-1)   29.60 [ 7]  22.91 [ 8]  6.69 [ 4]
8   8. Florida                     (7-3)   27.91 [ 8]  24.42 [ 7]  3.49 [17]
9   9. Arizona St                  (9-1)   25.88 [ 9]  19.82 [11]  6.06 [ 8]
10 10. South Florida               (7-3)   24.15 [10]  21.12 [ 9]  3.02 [24]
14 11. Cincinnati                  (8-2)   24.09 [11]  20.62 [10]  3.47 [18]
12 12. Clemson                     (8-2)   22.77 [12]  19.31 [12]  3.46 [19]
15 13. Georgia                     (8-2)   21.93 [13]  17.60 [14]  4.32 [10]
19 14. Southern Cal                (8-2)   20.44 [14]  17.71 [13]  2.72 [27]
22 15. Texas                       (9-2)   19.14 [15]  15.75 [15]  3.38 [20]
24 16. Virginia Tech               (8-2)   18.57 [16]  14.53 [20]  4.04 [11]
28 17. Illinois                    (8-3)   18.32 [17]  14.51 [21]  3.80 [12]
23 18. Penn State                  (8-3)   17.77 [18]  14.94 [18]  2.82 [26]
16 19. Connecticut                 (8-2)   17.07 [19]  13.50 [22]  3.56 [15]
11 20. Auburn                      (7-4)   16.98 [20]  14.82 [19]  2.15 [30]
13 21. Arkansas                    (6-4)   16.45 [21]  15.04 [17]  1.40 [45]
26 22. Kentucky                    (7-3)   16.36 [22]  13.03 [25]  3.33 [21]
17 23. Texas Tech                  (7-4)   16.35 [23]  15.10 [16]  1.25 [46]
31 24. Tennessee                   (7-3)   16.07 [24]  13.03 [24]  3.04 [23]
29 25. Boise St                    (9-1)   15.80 [25]  12.49 [26]  3.30 [22] 

Southern Cal and Texas both make a large jump, primarily from Points. USC is still only 27th in Win/Loss.  Virginia Tech moves up eight spots, while Illinois' win over the Buckeyes catapults them from #28 to #17. Boise State moves into the top 25; their undefeated conference-mate Hawaii is behind them at #33.

Power Rating: Win-loss component
lw  #  Team                         rec    rating
2   1. Kansas                     (10-0)    9.47
4   2. LSU                         (9-1)    8.32
3   3. Oregon                      (8-1)    7.59
1   4. Ohio State                 (10-1)    6.69
5   5. Oklahoma                    (9-1)    6.51
8   6. Hawaii                      (9-0)    6.29
9   7. Missouri                    (9-1)    6.24
6   8. Arizona St                  (9-1)    6.06
12  9. West Virginia               (8-1)    5.00
14 10. Georgia                     (8-2)    4.32
16 11. Virginia Tech               (8-2)    4.04
29 12. Illinois                    (8-3)    3.80
13 13. Virginia                    (9-2)    3.78
7  14. Boston College              (8-2)    3.59
10 15. Connecticut                 (8-2)    3.56
11 16. Michigan                    (8-3)    3.53
21 17. Florida                     (7-3)    3.49
28 18. Cincinnati                  (8-2)    3.47
19 19. Clemson                     (8-2)    3.46
17 20. Texas                       (9-2)    3.38
18 21. Kentucky                    (7-3)    3.33
15 22. Boise St                    (9-1)    3.30
25 23. Tennessee                   (7-3)    3.04
23 24. South Florida               (7-3)    3.02
34 25. Wisconsin                   (8-3)    2.88 

Ohio State falls to #4 on the score-doesn't-matter Win/Loss rating. Their loss to Illinois doesn't hurt them as much as the Sooners' loss to Colorado, and their 10 wins add up to more points for them than Missouri and Arizona State's 9 apiece. West Virginia is even farther behind, and Hawaii, despite having no losses, doesn't play good enough opposition to place higher than sixth even this late in the season.

Note that Oregon, with only eight wins, still ranks ahead of the Buckeyes. This is due to wins over teams such as USC, Arizona State, and Michigan. Their loss to California hurts them more than Ohio State's loss to Illinois, but their win points more than make up for it. It doesn't quite approach that of LSU, who is surprisingly close even to Kansas' undefeated rating.

Boise State's rating falls slightly after they beat winless Utah State. It's not because their strength of schedule goes down—the system is additive, not averaged—but because beating weak opponents gives only a very small gain that can easily be offset by the wins and losses of a team's previous opponents. In this instance, Boise State's past foes went 1-6 last week.

Power Rating: Points component
lw  #  Team                         rec    rating  off    def
2   1. LSU                         (9-1)   27.97  43.47  15.50
4   2. Kansas                     (10-0)   26.75  41.28  14.53
3   3. Oregon                      (8-1)   26.62  46.35  19.72
1   4. Oklahoma                    (9-1)   26.62  43.34  16.72
5   5. West Virginia               (8-1)   25.98  41.92  15.93
6   6. Missouri                    (9-1)   25.37  45.32  19.95
8   7. Florida                     (7-3)   24.42  47.20  22.78
7   8. Ohio State                 (10-1)   22.91  32.82   9.90
8   9. South Florida               (7-3)   21.12  38.39  17.26
14 10. Cincinnati                  (8-2)   20.62  38.18  17.56

Florida continues to rank high in the Points component. All of their losses have been narrow and to good teams. In case you were wondering, Kansas State fell from #17 to #34, and Nebraska moved up from #62 to #52.

Offense
#   Team                         rec    rating  off    def
1.  Florida                     (7-3)   24.42  47.20  22.78
2.  Oregon                      (8-1)   26.62  46.35  19.72
3.  Missouri                    (9-1)   25.37  45.32  19.95
4.  LSU                         (9-1)   27.97  43.47  15.50
5.  Oklahoma                    (9-1)   26.62  43.34  16.72
6.  West Virginia               (8-1)   25.98  41.92  15.93
7.  Kansas                     (10-0)   26.75  41.28  14.53
8.  Hawaii                      (9-0)    5.43  40.52  35.09
9.  Kentucky                    (7-3)   13.03  39.51  26.48
10. Clemson                     (8-2)   19.31  39.36  20.05

Florida's offense is their bread and butter, and their whupping of South Carolina makes it the #1 scoring machine in the country.

Defense
#   Team                         rec    rating  off    def
1.  Ohio State                 (10-1)   22.91  32.82   9.90 
2.  Auburn                      (7-4)   14.82  27.99  13.16
3.  Southern Cal                (8-2)   17.71  31.26  13.55
4.  Virginia Tech               (8-2)   14.53  28.13  13.60
5.  Kansas                     (10-0)   26.75  41.28  14.53
6.  LSU                         (9-1)   27.97  43.47  15.50
7.  West Virginia               (8-1)   25.98  41.92  15.93
8.  Connecticut                 (8-2)   13.50  29.67  16.16
9.  Arizona St                  (9-1)   19.82  36.40  16.58
10. Michigan                    (8-3)   11.97  28.63  16.66

Despite their loss and giving up a season-high 28 points, Ohio State retains the top spot defensively. Helping out was the fact that Georgia scored 45 points on Auburn, preventing the Tigers from taking 1st place. USC continues to move up the defensive chart with a good performance against Cal on Saturday.

SportsRatings College Football Power Rating - week 10

Kansas remains #1; Ohio State up to #2. The two major undefeated teams are 1-2 in the latest SportsRatings College Football Power Rating. Kansas holds on to the #1 spot after demolishing Nebraska 76-39, while Ohio State captures the #2 ranking by handling Wisconsin 38-17. LSU, despite coming back to beat Alabama, slips to #3, while Oregon, victor over Arizona State, remains in fourth place. The difference between the top 4 teams is less than one point overall.

Two Big 12 teams follow at #5 and #6. Oklahoma moves up a spot, and Missouri makes a leap from #11. Both made big gains on the Points component after blowing out Texas A&M and Colorado, respectively, putting three Big 12 teams among the top six. Florida moved up a spot to #8 while idle West Virginia fell to #7. Arizona State and South Florida both fell two spots after losing on Saturday.

SportsRatings Power Rating week 10 Overall
lw  #  Team                         rec    Overall     Pts         W/L
1   1. Kansas                      (9-0)  36.02 [ 1]  27.12 [ 4]   8.89 [ 2]
3   2. Ohio State                 (10-0)  35.85 [ 2]  25.17 [ 7]  10.67 [ 1]
2   3. LSU                         (8-1)  35.55 [ 3]  27.74 [ 2]   7.80 [ 4]
4   4. Oregon                      (8-1)  35.38 [ 4]  27.45 [ 3]   7.93 [ 3]
6   5. Oklahoma                    (8-1)  34.65 [ 5]  28.38 [ 1]   6.27 [ 5]
11  6. Missouri                    (8-1)  31.69 [ 6]  26.43 [ 6]   5.26 [ 9]
5   7. West Virginia               (7-1)  30.91 [ 7]  26.83 [ 5]   4.08 [12]
9   8. Florida                     (6-3)  26.37 [ 8]  23.50 [ 8]   2.87 [21]
7   9. Arizona St                  (8-1)  26.05 [ 9]  20.35 [10]   5.69 [ 6]
8  10. South Florida               (6-3)  23.56 [10]  20.91 [ 9]   2.65 [23]
12 11. Auburn                      (7-3)  21.04 [11]  18.42 [11]   2.61 [24]
21 12. Clemson                     (7-2)  20.30 [12]  17.40 [13]   2.89 [19]
14 13. Arkansas                    (6-3)  20.14 [13]  18.38 [12]   1.75 [36]
18 14. Cincinnati                  (7-2)  19.78 [14]  17.36 [14]   2.42 [28]
15 15. Georgia                     (7-2)  19.46 [15]  15.74 [18]   3.71 [14]
17 16. Connecticut                 (8-1)  19.09 [16]  14.29 [22]   4.79 [10]
28 17. Texas Tech                  (7-3)  18.93 [17]  17.30 [15]   1.63 [39]
10 18. Boston College              (8-1)  18.86 [18]  13.33 [25]   5.52 [ 7]
20 19. Southern Cal                (7-2)  18.84 [19]  16.71 [16]   2.12 [30]
19 20. Michigan                    (8-2)  18.10 [20]  13.90 [24]   4.20 [11]
16 21. Alabama                     (6-3)  17.95 [21]  15.50 [19]   2.45 [27]
24 22. Texas                       (8-2)  17.89 [22]  14.90 [21]   2.99 [17]
23 23. Penn State                  (7-3)  17.87 [23]  15.40 [20]   2.47 [26]
27 24. Virginia Tech               (7-2)  17.64 [24]  14.23 [23]   3.40 [16]
13 25. Kansas St                   (5-4)  16.86 [25]  15.75 [17]   1.11 [57]

Clemson jumps up nine to #12 after crushing Duke, though their move is as much the result of past opponent Florida State upsetting Boston College. The Eagles' first loss drops them from #10 to #18, past Auburn, Arkansas, Cincinnati, Georgia, Connecticut, and Texas Tech, all of whom posted wins over the weekend.

The Red Raiders were a big gainer, jumping from 28 to 17, while another Big 12 school, Kansas State, plunged after being upset by Iowa State. In all there are six Big 12 teams among the first 25 schools, joining six SEC teams, four from the Big East, and three each from the Big Ten, Pac-10, and ACC.

Once again, the other undefeated team in Division I-A, Hawaii, is outside the top 25 at #33, though on the Win/Loss subcomponent, they are #8.

Power Rating: Win-loss component
lw  #  Team                         rec    rating
2   1. Ohio State                 (10-0)   10.67
4   2. Kansas                      (9-0)    8.89
7   3. Oregon                      (8-1)    7.93
5   4. LSU                         (8-1)    7.80
8   5. Oklahoma                    (8-1)    6.27
3   6. Arizona St                  (8-1)    5.69
1   7. Boston College              (8-1)    5.52
6   8. Hawaii                      (8-0)    5.27
11  9. Missouri                    (8-1)    5.26
10 10. Connecticut                 (8-1)    4.79
12 11. Michigan                    (8-2)    4.20
9  12. West Virginia               (7-1)    4.08
15 13. Virginia                    (8-2)    3.71
14 14. Georgia                     (7-2)    3.71
17 15. Boise St                    (8-1)    3.40
18 16. Virginia Tech               (7-2)    3.40
21 17. Texas                       (8-2)    2.99
19 18. Kentucky                    (6-3)    2.95
20 19. Clemson                     (7-2)    2.89
28 20. California                  (6-3)    2.89
25 21. Florida                     (6-3)    2.87
33 22. Florida St                  (6-3)    2.83
13 23. South Florida               (6-3)    2.65
24 24. Auburn                      (7-3)    2.61
26 25. Tennessee                   (6-3)    2.58

When point spreads aren't considered, Ohio State comes up ahead of Kansas. The Buckeyes' schedule is harder and they've played one more game than the Jayhawks, giving them a fairly wide margin at the top. Oregon rockets to #3 and in the process knocks ASU to #6, becoming the top-rated one-loss team in this subcomponent. LSU inches ahead after another close win; the Tigers' loss to Kentucky scores roughly equal to Oregon's loss to California, and their wins over Florida, Auburn, and Virginia Tech are very close to the Ducks' wins over ASU, Cal, and USC. When all is said and done, Oregon retains a razor-thin margin that could change depending on how their past opponents fare, let alone what them teams themselves do.

Last week's #1, Boston College, falls to #7, still a spot ahead of the third undefeated team, Hawaii. Michigan's fortunes increase with Oregon's, as the Wolverines' loss to the Ducks hardly hurts them anymore, especially since the Win/Loss algorithm doesn't see the 32 point margin. Boise State is the lowest-ranked 1-loss team, held back by a weak schedule and a loss to Washington. Kentucky is tops among 3-loss teams, with their win over LSU bouying them.

Power Rating: Points component
lw  #  Team                         rec    rating  off    def
5   1. Oklahoma                    (8-1)   28.38  43.51  15.12
1   2. LSU                         (8-1)   27.74  42.59  14.85
2   3. Oregon                      (8-1)   27.45  46.40  18.95
4   4. Kansas                      (9-0)   27.12  41.18  14.05
3   5. West Virginia               (7-1)   26.83  41.64  14.81
13  6. Missouri                    (8-1)   26.43  46.05  19.62
6   7. Ohio State                 (10-0)   25.17  33.43   8.25
7   8. Florida                     (6-3)   23.50  45.42  21.92
8   9. South Florida               (6-3)   20.91  38.52  17.61
9  10. Arizona St                  (8-1)   20.35  37.34  16.98

For all the attention given Kansas' blowout of Nebraska, two other Big 12 teams made bigger moves on the pure Points power rating subcomponent. Oklahoma retakes the #1 crown it held earlier in the year after beating Texas A&M 42-14, and Missouri leaps into #6 by crushing Colorado by 45 on the road. By comparison, Kansas beat these teams 19-11 and 19-14.

Offense
#   Team                         rec    rating  off    def
1.  Oregon                      (8-1)   27.45  46.40  18.95
2.  Missouri                    (8-1)   26.43  46.05  19.62
3.  Florida                     (6-3)   23.50  45.42  21.92
4.  Oklahoma                    (8-1)   28.38  43.51  15.12
5.  Hawaii                      (8-0)    7.23  42.71  35.48
6.  LSU                         (8-1)   27.74  42.59  14.85
7.  Arkansas                    (6-3)   18.38  41.72  23.33
8.  West Virginia               (7-1)   26.83  41.64  14.81
9.  Kansas                      (9-0)   27.12  41.18  14.05
10. Kentucky                    (6-3)   12.98  40.03  27.05

Kansas' offense does get quite a boost for their 76 points Saturday, a move from #15 to #9 on the list of top offenses. Their defense, however, slips a notch for allowing 39 points, and now is giving up an average of 14 points a game instead of 10.

Defense
#   Team                         rec    rating  off    def
1.  Ohio State                 (10-0)   25.17  33.43   8.25
2.  Auburn                      (7-3)   18.42  28.35   9.93
3.  Virginia Tech               (7-2)   14.23  26.49  12.25
4.  Kansas                      (9-0)   27.12  41.18  14.05
5.  Southern Cal                (7-2)   16.71  31.24  14.52
6.  West Virginia               (7-1)   26.83  41.64  14.81
7.  LSU                         (8-1)   27.74  42.59  14.85
8.  Michigan                    (8-2)   13.90  28.92  15.02
9.  Oklahoma                    (8-1)   28.38  43.51  15.12
10. Connecticut                 (8-1)   14.29  31.16  16.86

As always, Ohio State has the best defense, though Auburn is getting closer. Southern Cal's performance Saturday—holding Oregon State to a field goal—puts them in the top five, and should give them some hope that down the stretch their offensive inconsistency won't hurt them as much.

SportsRatings College Football Power Rating week 9

A bit of a surprise at #1 as the Kansas Jayhawks move into the pole position. Not too shocking, as they have rated high all year here, and top the charts this week in other well-known power ratings as well. Interestingly they gain the top spot while only 4th in both the subcomponents, ranking three rungs lower than LSU in Points and one place higher in Win/Loss. Ohio State also nudges Oregon out of third place despite the Ducks win over USC, which was actually closer than the Points system expected. West Virginia's win over Rutgers was more impressive than expected, and they jump to #5.

As the Win/Loss component gains in importance, four undefeated teams make advances in the top ten: Kansas and Ohio State up one place each, Arizona State and Boston College up three each for wins over strong teams. Meanwhile being idle hurts LSU and Oklahoma, as they can't gain any significant Win/Loss points. South Florida and Florida slide just a bit for losing to good teams; Florida remains in the top  ten despite three losses, since all their losses were close games to top 25 teams.

SportsRatings Power Rating week 9 Overall
lw  #  Team                         rec    Overall     Pts         W/L
2   1. Kansas                      (8-0)   35.84 [ 1]  26.95 [ 4]  8.88 [ 4]
1   2. LSU                         (7-1)   34.93 [ 2]  28.48 [ 1]  6.45 [ 5]
4   3. Ohio State                  (9-0)   34.39 [ 3]  25.02 [ 6]  9.36 [ 2]
3   4. Oregon                      (7-1)   33.11 [ 4]  27.73 [ 2]  5.37 [ 7]
8   5. West Virginia               (7-1)   31.81 [ 5]  27.32 [ 3]  4.48 [ 9]
5   6. Oklahoma                    (7-1)   31.78 [ 6]  26.50 [ 5]  5.27 [ 8]
10  7. Arizona St                  (8-0)   30.86 [ 7]  21.60 [ 9]  9.25 [ 3]
6   8. South Florida               (6-2)   26.50 [ 8]  22.92 [ 8]  3.57 [13]
7   9. Florida                     (5-3)   25.52 [ 9]  23.09 [ 7]  2.43 [25]
13 10. Boston College              (8-0)   25.44 [10]  15.84 [18]  9.59 [ 1]
9  11. Missouri                    (7-1)   24.29 [11]  20.07 [13]  4.21 [11]
11 12. Auburn                      (6-3)   22.75 [12]  20.20 [12]  2.55 [24]
15 13. Kansas St                   (5-3)   22.66 [13]  20.79 [10]  1.87 [36]
14 14. Arkansas                    (5-3)   21.86 [14]  20.56 [11]  1.29 [52]
25 15. Georgia                     (6-2)   20.13 [15]  16.68 [15]  3.45 [14]
18 16. Alabama                     (6-2)   19.11 [16]  16.18 [16]  2.93 [16]
23 17. Connecticut                 (7-1)   18.53 [17]  14.20 [22]  4.33 [10]
16 18. Cincinnati                  (6-2)   18.44 [18]  16.68 [14]  1.75 [40]
19 19. Michigan                    (7-2)   17.84 [19]  14.06 [23]  3.77 [12]
22 20. Southern Cal                (6-2)   17.78 [20]  16.03 [17]  1.74 [41]
29 21. Clemson                     (6-2)   17.73 [21]  15.16 [19]  2.56 [20]
20 22. California                  (5-3)   16.59 [22]  14.23 [21]  2.35 [28]
17 23. Penn State                  (6-3)   16.40 [23]  14.47 [20]  1.92 [32]
21 24. Texas                       (7-2)   16.01 [24]  13.45 [25]  2.56 [21]
12 25. Kentucky                    (6-3)   15.95 [25]  13.27 [26]  2.68 [19]

Georgia takes a big leap from #25 to #15 after beating Florida, and Connecticut is up 6 places to #17 for beating South Florida. Clemson jumps 8 places to #21 after dominating Maryland. Interestingly, USC moves up after losing to Oregon which, as I said, was a more impressive showing for the Trojans than expected by the Points component. Kentucky takes a dive from 12 to 25 after being crushed by Mississippi State 31-14.

The fifth undefeated team in Division I-A, Hawaii, is up nine places to #31 overall. But on the Win/Loss side, they are up to #6.

Power Rating: Win-loss component
lw  #  Team                         rec    rating
2   1. Boston College              (8-0)    9.59
1   2. Ohio State                  (9-0)    9.36
3   3. Arizona St                  (8-0)    9.25
5   4. Kansas                      (8-0)    8.88
4   5. LSU                         (7-1)    6.45
9   6. Hawaii                      (8-0)    5.44
10  7. Oregon                      (7-1)    5.37
6   8. Oklahoma                    (7-1)    5.27
13  9. West Virginia               (7-1)    4.48
18 10. Connecticut                 (7-1)    4.33
11 11. Missouri                    (7-1)    4.21
14 12. Michigan                    (7-2)    3.77
8  13. South Florida               (6-2)    3.57
22 14. Georgia                     (6-2)    3.45
7  15. Virginia                    (7-2)    3.14
23 16. Alabama                     (6-2)    2.93
19 17. Boise St                    (7-1)    2.84
15 18. Virginia Tech               (6-2)    2.70
12 19. Kentucky                    (6-3)    2.68
38 20. Clemson                     (6-2)    2.56
32 21. Texas                       (7-2)    2.56
31 22. Purdue                      (7-2)    2.55
16 23. South Carolina              (6-3)    2.55
29 24. Auburn                      (6-3)    2.55
17 25. Florida                     (5-3)    2.43

Boston College nudged Ohio State out of the top spot in the Win/Loss component, even though the Buckeyes have played more games. Kansas gained far more points than a win over Texas A&M would have suggested; their priors went 4-2 on Saturday and they bumped LSU out of #4. Hawaii's opponents fared poorly as usual, but the Warriors are up to #6 with 5.44 points, a very low score for an 8-0 team. Still they're lucky that the Win/Loss algorithm doesn't know how close some of those games were. Connecticut and Georgia jump eight spots each, and though they were idle this week, Alabama takes almost as big a jump; their prior opponents went 5-1 on Saturday, and they move past a lot of week 9 losers (Virginia Tech, Kentucky, South Carolina, Florida). The Gators now sit one spot behind Auburn, about where they logically should be. It's nice when things work out like that.

Power Rating: Points component
lw  #  Team                         rec    rating  off    def
1   1. LSU                         (7-1)   28.48  42.26  13.77
2   2. Oregon                      (7-1)   27.73  45.88  18.15
7   3. West Virginia               (7-1)   27.32  41.91  14.58
4   4. Kansas                      (8-0)   26.95  37.07  10.11
5   5. Oklahoma                    (7-1)   26.50  42.41  15.90
8   6. Ohio State                  (9-0)   25.02  31.72   6.70
3   7. Florida                     (5-3)   23.09  44.58  21.49
6   8. South Florida               (6-2)   22.92  38.47  15.55
12  9. Arizona St                  (8-0)   21.60  37.45  15.85
13 10. Kansas St                   (5-3)   20.79  42.34  21.55

West Virginia made the biggest gain at the top of the Points component list after beating Rutgers on the road by four touchdowns. Conversely Florida took a small spill but still rates in the top ten, mainly due to their offense ranking #2 in the nation in scoring:

Offense
#   Team                         rec    rating  off    def
1.  Oregon                      (7-1)   27.73  45.88  18.15
2.  Florida                     (5-3)   23.09  44.58  21.49
3.  Hawaii                      (8-0)    8.15  43.09  34.93
4.  Missouri                    (7-1)   20.07  42.69  22.61
5.  Oklahoma                    (7-1)   26.50  42.41  15.90
6.  Kansas St                   (5-3)   20.79  42.34  21.55
7.  LSU                         (7-1)   28.48  42.26  13.77
8.  West Virginia               (7-1)   27.32  41.91  14.58
9.  Arkansas                    (5-3)   20.56  41.36  20.79
10. Kentucky                    (6-3)   13.27  40.45  27.18

Defense
#   Team                         rec    rating  off    def
1.  Ohio State                  (9-0)   25.02  31.72   6.70
2.  Auburn                      (6-3)   20.20  29.41   9.20
3.  Kansas                      (8-0)   26.95  37.07  10.11
4.  Virginia Tech               (6-2)   11.57  24.43  12.85
5.  LSU                         (7-1)   28.48  42.26  13.77
6.  West Virginia               (7-1)   27.32  41.91  14.58
7.  Michigan                    (7-2)   14.06  29.10  15.03
8.  Georgia Tech                (5-3)   11.14  26.24  15.09
9.  Connecticut                 (7-1)   14.20  29.52  15.32
10. South Florida               (6-2)   22.92  38.47  15.55

Another week, another scoring defense list with Ohio State on top. But how about those Wolverines? After the first two games, when Michigan gave up 34 and 39, who would have thought they'd eventually rank in the top ten when adjusted for opposition?

SportsRatings Power Rating - week 8

LSU remains #1 in the overall power rating after the win over Auburn. South Florida's first loss drops them to #6, allowing Kansas to move into #2. Oregon leaps up to #3, and Ohio State nudges Oklahoma out of #4. Florida moves up one spot after beating Kentucky, while West Virgina slides down two despite beating Mississippi State handily, and Missouri's throttling of Texas Tech nudges idle Arizona State out of the ninth position.

SportsRating Power Rating week 8 Overall
lw  #  Team                         rec     Overall     Pts          W/L
1   1. LSU                         (7-1)   37.07[ 1]  29.97[ 1]   7.09[ 4]
3   2. Kansas                      (7-0)   33.79[ 2]  27.07[ 4]   6.71[ 5]
7   3. Oregon                      (6-1)   33.21[ 3]  28.72[ 2]   4.49[10]
5   4. Ohio State                  (8-0)   32.50[ 4]  24.30[ 8]   8.19[ 1]
4   5. Oklahoma                    (7-1)   31.67[ 5]  26.78[ 5]   4.89[ 6]
2   6. South Florida               (6-1)   31.27[ 6]  26.57[ 6]   4.70[ 8]
8   7. Florida                     (5-2)   30.64[ 7]  27.51[ 3]   3.13[17]
6   8. West Virginia               (6-1)   29.70[ 8]  26.17[ 7]   3.52[13]
10  9. Missouri                    (6-1)   27.76[ 9]  23.65[ 9]   4.11[11]
9  10. Arizona St                  (7-0)   27.65[10]  20.54[12]   7.11[ 3]
14 11. Auburn                      (5-3)   24.51[11]  22.25[10]   2.25[29]
11 12. Kentucky                    (6-2)   22.96[12]  19.08[14]   3.88[12]
16 13. Boston College              (7-0)   22.47[13]  15.03[18]   7.43[ 2]
21 14. Arkansas                    (4-3)   22.34[14]  21.26[11]   1.08[57]
12 15. Kansas St                   (4-3)   21.71[15]  20.28[13]   1.43[49]
13 16. Cincinnati                  (6-2)   18.99[16]  17.06[15]   1.93[34]
18 17. Penn State                  (6-2)   18.42[17]  16.15[16]   2.27[27]
32 18. Alabama                     (6-2)   17.96[18]  15.48[17]   2.47[23]
23 19. Michigan                    (6-2)   17.47[19]  14.05[21]   3.41[14]
17 20. California                  (5-2)   16.87[20]  14.33[20]   2.53[21]
20 21. Texas                       (6-2)   16.75[21]  14.76[19]   1.99[32]
30 22. Southern Cal                (6-1)   16.21[22]  13.75[22]   2.45[25]
22 23. Connecticut                 (6-1)   16.03[23]  13.04[25]   2.99[18]
19 24. Texas Tech                  (6-2)   15.77[24]  13.46[23]   2.30[26]
25 25. Georgia                     (5-2)   15.65[25]  13.12[24]   2.53[22]

Note that Auburn actually moves up three spots despite losing, and Kentucky only falls one spot. Auburn, Arkansas, and Kansas State remain high in the ratings despite having three losses; conversely, undefeated Hawaii comes in at #40. See the full ratings here.

Power Rating: Win-loss component
lw  #  Team                         rec    rating
4   1. Ohio State                  (8-0)    8.19
3   2. Boston College              (7-0)    7.43
2   3. Arizona St                  (7-0)    7.11
5   4. LSU                         (7-1)    7.09
6   5. Kansas                      (7-0)    6.71
10  6. Oklahoma                    (7-1)    4.89
13  7. Virginia                    (7-1)    4.81
1   8. South Florida               (6-1)    4.70
9   9. Hawaii                      (7-0)    4.61
12 10. Oregon                      (6-1)    4.49
15 11. Missouri                    (6-1)    4.11
8  12. Kentucky                    (6-2)    3.88
16 13. West Virginia               (6-1)    3.52
20 14. Michigan                    (6-2)    3.41
14 15. Virginia Tech               (6-1)    3.36
7  16. South Carolina              (6-2)    3.24
12 17. Florida                     (5-2)    3.13
23 18. Connecticut                 (6-1)    2.99
21 19. Boise St                    (6-1)    2.56
47 20. Rutgers                     (5-2)    2.56
11 21. California                  (5-2)    2.53
22 22. Georgia                     (5-2)    2.53
38 23. Alabama                     (6-2)    2.47
32 24. Texas A&M                   (6-2)    2.47
24 25. Southern Cal                (6-1)    2.45

Ohio State, the only 8-0 team in the country, takes over the Win/Loss #1 spot from South Florida. It helps that Boston College and Arizona State were idle. Note that despite this, BC passes ASU. This is due to how their previous opponents fared on Saturday.  The Eagles rating increased slightly from  7.02 to 7.43 while the Sun Devils rating slipped from 7.67 to 7.11. Oklahoma gained four spots after beating Iowa State, but their rating didn't increase much: from 4.39 to 4.89. Instead, three teams ahead of them lost, and Hawaii was idle. Note that USC slipped a spot despite beating Notre Dame, which like beating Iowa State doesn't carry a lot of point value.

Power Rating: Points component
lw  #  Team                         rec   overall  off    def
3   1. LSU                         (7-1)   29.97  42.97  12.99
8   2. Oregon                      (6-1)   28.72  47.62  18.90
6   3. Florida                     (5-2)   27.51  44.84  17.32
1   4. Kansas                      (7-0)   27.07  39.28  12.20
2   5. Oklahoma                    (7-1)   26.78  43.08  16.30
4   6. South Florida               (6-1)   26.57  40.62  14.04
5   7. West Virginia               (6-1)   26.17  41.75  15.57
7   8. Ohio State                  (8-0)   24.30  30.83   6.52
10  9. Missouri                    (6-1)   23.65  42.62  18.97
11 10. Auburn                      (5-3)   22.25  31.56   9.30
18 11. Arkansas                    (4-3)   21.26  39.93  18.66
13 12. Arizona St                  (7-0)   20.54  37.56  17.02
9  13. Kansas St                   (4-3)   20.28  41.93  21.65
14 14. Kentucky                    (6-2)   19.08  45.24  26.16
12 15. Cincinnati                  (6-2)   17.06  35.83  18.77

LSU takes the #1 spot in the Points component not by improving their score, which actually declines slightly, but because both Kansas and Oklahoma struggled on Saturday and fell to #4 and #5. Auburn's score improves from the close game at LSU (home field advantage this week is about 3.25). Kentucky, too, doesn't see any decline as their margin of loss to Florida was within expected bounds. With a points-based system, you will more regularly see things such as 3-loss Auburn and Arkansas ahead of undefeated Arizona State. The Tigers and Razorbacks losses have all been close games.

The full Points component ratings include surprises such as Boston College at #18 and Hawaii at #51. The latter probably truly reflects how average of a team Hawaii is. Their offense and defense are completely different birds, as we can see by reviewing the offensive and defensive scoring charts:

Offense
#   Team                         rec    rating  off    def
1.  Oregon                      (6-1)   28.72  47.62  18.90
2.  Kentucky                    (6-2)   19.08  45.24  26.16
3.  Florida                     (5-2)   27.51  44.84  17.32
4.  Hawaii                      (7-0)    5.20  44.48  39.27
5.  Oklahoma                    (7-1)   26.78  43.08  16.30
6.  LSU                         (7-1)   29.97  42.97  12.99
7.  Missouri                    (6-1)   23.65  42.62  18.97
8.  Kansas St                   (4-3)   20.28  41.93  21.65
9.  West Virginia               (6-1)   26.17  41.75  15.57
10. South Florida               (6-1)   26.57  40.62  14.04

About Oregon, Washington defensive coordinator Kent Baer said "I don't think there is a better offense in the country right now," and he's right. As for Hawaii, their pass-based offense ranks fourth in scoring when adjusted for opposition; and the opposition ratings continue to adjust, as Hawaii was 4th last week and didn't play, so their opponents were downgraded this Saturday. They rank just 113th, however, in defense, leading to their combined rating of #51 in the Points power rating.

Defense
#   Team                         rec    rating  off    def
1.  Ohio State                  (8-0)   24.30  30.83   6.52
2.  Auburn                      (5-3)   22.25  31.56   9.30
3.  Kansas                      (7-0)   27.07  39.28  12.20
4.  LSU                         (7-1)   29.97  42.97  12.99
5.  Virginia Tech               (6-1)   11.70  25.19  13.49
6.  South Florida               (6-1)   26.57  40.62  14.04
7.  Georgia Tech                (5-3)   11.30  26.07  14.77
8.  South Carolina              (6-2)   11.99  27.18  15.19
9.  West Virginia               (6-1)   26.17  41.75  15.57
10. Michigan                    (6-2)   14.05  29.64  15.58

Ohio State still rules on defense, even with their offense giving up 14 points to Michigan State on Saturday. Offensive ineptitude gets factored into the defense numbers, of course, just as interceptions (and punts) returned for touchdowns get factored into the offensive numbers. Michigan also rates much higher when numbers are adjusted (i.e., those 39 points given up to Oregon—the #1 scoring team—don't look as bad then).

SportsRatings Power Rating - Week 7: LSU remains on top

First off, let's look at the Win-Loss component, to see where the teams are ranked in terms of their resume for the year, irrespective of points scored:

Power Rating: Win-loss component
lw  #  Team                         rec    rating
3   1. South Florida               (6-0)    8.06
5   2. Arizona St                  (7-0)    7.67
6   3. Boston College              (7-0)    7.02
4   4. Ohio State                  (7-0)    6.86
1   5. LSU                         (6-1)    6.81
9   6. Kansas                      (6-0)    5.96
11  7. South Carolina              (6-1)    5.71
18  8. Kentucky                    (6-1)    5.63
13  9. Hawaii                      (7-0)    4.90
14 10. Oklahoma                    (6-1)    4.39
7  11. California                  (5-1)    3.67
16 12. Oregon                      (5-1)    3.61
22 13. Virginia                    (6-1)    3.58
15 14. Virginia Tech               (6-1)    3.42
2  15. Missouri                    (5-1)    3.25
17 16. West Virginia               (5-1)    3.07
   17. Texas Tech                  (6-1)    2.98
8  18. Cincinnati                  (6-1)    2.86
   19. Auburn                      (5-2)    2.64
   20. Michigan                    (5-2)    2.57
   21. Boise St                    (5-1)    2.39
   22. Georgia                     (5-2)    2.26
12 23. Connecticut                 (5-1)    2.22
   24. Southern Cal                (5-1)    2.11
   25. Tennessee                   (4-2)    2.07

Of the undefeated teams, South Florida's schedule has been the toughest. This power rating uses "additive" schedule strength rather than "average" schedule strength, so playing a weak opponent here and there doesn't detract from a team's rating as it would in an averaged system. LSU's schedule has been so tough that they stay ahead of two undefeated teams; Hawaii's schedule is slight enough that two more teams with a loss, South Carolina and Kentucky, are ahead of the Warriors. The Gamecocks and Wildcats both feed off the upset of LSU. At the bottom of the list, some 2-loss teams move in; USC is actually the lowest rated of the 1-loss teams, as the loss to Stanford hurts them the most.

Now the top ten of the Points component, a truer measure of how powerful a team is:

Power Rating: Points component
lw  #  Team                         rec   overall  off    def
3   1. Kansas                      (6-0)   30.98  43.35  12.36
2   2. Oklahoma                    (6-1)   30.95  46.13  15.18
1   3. LSU                         (6-1)   30.57  42.13  11.56
10  4. South Florida               (6-0)   28.91  41.23  12.31
5   5. West Virginia               (5-1)   28.35  42.84  14.49
6   6. Florida                     (4-2)   26.57  44.27  17.70
4   7. Ohio State                  (7-0)   26.37  32.05   5.68
8   8. Oregon                      (5-1)   26.34  43.82  17.47
17  9. Kansas St                   (4-2)   22.50  41.38  18.87
7  10. Missouri                    (5-1)   22.00  43.24  21.23

Kansas takes a slight lead that may or may not last as they play tougher teams, but the difference between them and #3 LSU is very small. Ohio State's slip is probably due to former opponents Purdue and Washington getting beat badly during the week, and OSU's blowout of Kent State not making much of a difference.

So we see that the two lists are very out of synch. When we combine them into the Overall results:

SportsRating Power Rating week 7 Overall
lw  #  Team                         rec     Overall     Pts          W/L
1   1. LSU                         (6-1)   37.39[ 1]  30.57[ 3]   6.81[ 5]
8   2. South Florida               (6-0)   36.97[ 2]  28.91[ 4]   8.06[ 1]
3   3. Kansas                      (6-0)   36.95[ 3]  30.98[ 1]   5.96[ 6]
4   4. Oklahoma                    (6-1)   35.34[ 4]  30.95[ 2]   4.39[10]
2   5. Ohio State                  (7-0)   33.23[ 5]  26.37[ 7]   6.86[ 4]
6   6. West Virginia               (5-1)   31.42[ 6]  28.35[ 5]   3.07[16]
10  7. Oregon                      (5-1)   29.96[ 7]  26.34[ 8]   3.61[12]
9   8. Florida                     (4-2)   28.59[ 8]  26.57[ 6]   2.02[28]
12  9. Arizona St                  (7-0)   27.85[ 9]  20.17[13]   7.67[ 2]
5  10. Missouri                    (5-1)   25.25[10]  22.00[10]   3.25[15]
19 11. Kentucky                    (6-1)   24.60[11]  18.96[14]   5.63[ 8]
20 12. Kansas St                   (4-2)   24.54[12]  22.50[ 9]   2.03[27]
7  13. Cincinnati                  (6-1)   23.29[13]  20.42[12]   2.86[18]
16 14. Auburn                      (5-2)   23.23[14]  20.58[11]   2.64[19]
17 15. South Carolina              (6-1)   20.77[15]  15.06[20]   5.71[ 7]
13 16. Boston College              (7-0)   19.77[16]  12.75[21]   7.02[ 3]
11 17. California                  (5-1)   19.54[17]  15.87[17]   3.67[11]
22 18. Penn State                  (5-2)   19.31[18]  17.42[15]   1.89[31]
   19. Texas Tech                  (6-1)   19.07[19]  16.09[16]   2.98[17]
   20. Texas                       (5-2)   17.19[20]  15.44[19]   1.75[36]
24 21. Arkansas                    (3-3)   16.60[21]  15.80[18]   0.80[66]
15 22. Connecticut                 (5-1)   14.88[22]  12.65[22]   2.22[23]
   23. Michigan                    (5-2)   14.64[23]  12.07[24]   2.57[20]
14 24. Illinois                    (5-2)   14.38[24]  12.36[23]   2.01[29]
   25. Georgia                     (5-2)   14.29[25]  12.02[25]   2.26[22]


Although LSU is 3rd and 5th in the subcomponents, they are high enough in both to retain the #1 spot in the overall power rating. South Florida is a close second and Kansas holds onto third. Basically what the Overall results are "saying" is, although USF is undefeated against a good schedule, in terms of dominance they haven't quite shown enough, barely beating Auburn and having close games against West Virginia and FAU. And while Kansas has blown out most of its opponents, their opposition hasn't been close to as good as LSU's. The Tigers on the other hand have won convincingly against excellent opposition and their sole loss doesn't affect that enough to knock them out of the top spot.

Again, Ohio State falls to fifth due to their opponents' losses, and since the teams below them had big wins. Ohio State won 48-3, but Kent State is rated 94th overall in Division I-AA. The complete ratings can be seen here.

Now a quick look at the and scoring average stats.

Offense
#   Team                         rec    rating  off    def
1.  Hawaii                      (7-0)    7.55  46.52  38.97
2.  Oklahoma                    (6-1)   30.95  46.13  15.18
3.  Kentucky                    (6-1)   18.96  44.94  25.97
4.  Florida                     (4-2)   26.57  44.27  17.70
5.  Oregon                      (5-1)   26.34  43.82  17.47
6.  Kansas                      (6-0)   30.98  43.35  12.36
7.  Missouri                    (5-1)   22.00  43.24  21.23
8.  West Virginia               (5-1)   28.35  42.84  14.49
9.  LSU                         (6-1)   30.57  42.13  11.56
10. Kansas St                   (4-2)   22.50  41.38  18.87

Kentucky gets a boost from scoring on LSU's tough defense, but also gets an artificial boost due to the overtime scoring, which gets factored in the same as normal score. It's too much work to factor that out of the results, which would be ideal. Defense:

Defense
#   Team                         rec    rating  off    def
1.  Ohio State                  (7-0)   26.37  32.05   5.68
2.  Auburn                      (5-2)   20.58  30.19   9.60
3.  LSU                         (6-1)   30.57  42.13  11.56
4.  South Florida               (6-0)   28.91  41.23  12.31
5.  Kansas                      (6-0)   30.98  43.35  12.36
6.  Virginia Tech               (6-1)   10.79  25.14  14.34
7.  West Virginia               (5-1)   28.35  42.84  14.49
8.  Georgia Tech                (4-3)    9.10  23.98  14.88
9.  South Carolina              (6-1)   15.06  30.16  15.09
10. Oklahoma                    (6-1)   30.95  46.13  15.18

Ohio State stays on top of the defensive charts, and Auburn moves into #2 after holding Arkansas to 7 points. Former top ten Boise State falls to #54 after their 69-67 win last night, which probably caused a massive shift among every team related to them. Again, though the game was high scoring before the 4 overtimes, the overtime scoring inflated the averages. The score margin was barely affected by all of this, of course, which is why overtime doesn't have a huge impact on the Points power ratings.

BCS Computers Bullish on South Florida

(I made that headline up myself, and I'm not even going to do a web search to see if anyone else came up with it.)

Of the eight computer systems the BCS uses, at least six have South Florida #1, enough to make them the overall #1 in the computer composite with an average of 1.00 (the high and low are thrown out before the average).

This jibes with the Win/Loss component of my power rating:

SportsRatings Power Rating Win/Loss component vs. BCS computer composite:

rnk Team                         rec    SR W/L  BCS comp rank
1.  South Florida               (6-0)    8.06      1
2.  Arizona St                  (7-0)    7.63     t5
3.  Boston College              (7-0)    7.02      7
4.  Ohio State                  (7-0)    6.82     t5
5.  LSU                         (6-1)    6.81      2
6.  Kansas                      (6-0)    5.96     t8
7.  South Carolina              (6-1)    5.71      3
8.  Kentucky                    (6-1)    5.63      4
9.  Hawaii                      (7-0)    4.90    t25
10. Oklahoma                    (6-1)    4.39     11
11. California                  (5-1)    3.67    t14
12. Oregon                      (5-1)    3.62     13
13. Virginia                    (6-1)    3.57    t14
14. Virginia Tech               (6-1)    3.42     t8
15. Missouri                    (5-1)    3.25     16
16. West Virginia               (5-1)    3.07     10
17. Texas Tech                  (6-1)    2.98     --
18. Cincinnati                  (6-1)    2.86     22
19. Auburn                      (5-2)    2.64     12
20. Michigan                    (5-2)    2.57     24
21. Georgia                     (5-2)    2.26     19
22. Connecticut                 (5-1)    2.22     --
23. Southern Cal                (5-1)    2.10     23
24. Tennessee                   (4-2)    2.07     17T
25. Wyoming                     (4-2)    2.04     --

The computer ratings the BCS uses are not allowed to take into consideration the score of the game; most systems that don't use score margin tend to be roughly similar. So why is South Florida #1?

They've played and defeated some good teams: Auburn and West Virginia, both of whom show up in the BCS computer composite top 12.  Ohio State, Boston College, and Arizona State all fall below the LSU-Kentucky-South Carolina axis, due to their lack of big wins. Kansas is further down and Hawaii all the way at #25. Virginia Tech is in the top ten since their only loss it to LSU; West Virginia, likewise, only lost to South Florida. This is the logic of an algorithm that has only wins and losses to go by.

That LSU is still #2 reflects their strength of schedule. That Oklahoma is #11 according to the computers is understandable when you take victory margin out of the equation.

Florida (17), Kansas State (20), and Wake Forest (20) are the only teams in the BCS computer composite not in my Win-Loss rating (and they rank 27, 26 and 30 in mine).

If I were able to find an up-to-date list of each BCS computer's rankings, I could see which one is closest to my method, but some of the results aren't online yet. Next time. My final power rating for this week won't be available until after the Boise State game.

SportsRatings Power Rating - week 6

I have new pages up for the Power Rating (with Points and Win-Loss sub-ratings) and for Score Margin stats (broken down into offense and defense).

lw  #  Team                         rec   Overall  Pts   W/L
1   1. LSU                         (6-0)   44.26    1     1
4   2. Ohio State                  (6-0)   35.84    4     4
6   3. Kansas                      (5-0)   33.67    3     9
2   4. Oklahoma                    (5-1)   33.46    2    14
12  5. Missouri                    (5-0)   32.77    7     2
5   6. West Virginia               (5-1)   31.14    5    17
7   7. Cincinnati                  (6-0)   30.03    9     8
3   8. South Florida               (5-0)   29.95   10     3
9   9. Florida                     (4-2)   28.68    6    31
10 10. Oregon                      (4-1)   28.03    8    16
7  11. California                  (5-0)   26.41   11     7
14 12. Arizona St                  (6-0)   26.16   12     5
18 13. Boston College              (6-0)   22.57   18     6
17 14. Illinois                    (5-1)   21.85   14    10
16 15. Connecticut                 (5-0)   20.62   15    12
19 16. Auburn                      (4-2)   20.62   13    27
21 17. South Carolina              (5-1)   20.16   19    11
   18. Boise St                    (4-1)   18.68   16    26
13 19. Kentucky                    (5-1)   18.19   20    18
   20. Kansas St                   (3-2)   17.51   17    44
   21. Purdue                      (5-1)   16.13   24    24
   22. Penn State                  (4-2)   16.00   21    48
11 23. Southern Cal                (4-1)   15.89   23    29
23 24. Arkansas                    (3-2)   15.56   22    54
25 25. Colorado                    (4-2)   14.85   25    28

LSU is now #1 in both sub-ratings, and #1 by a wide margin in the overall Power Rating. Ohio State,  Kansas, and Missouri move up after impressive wins, while Oklahoma and South Florida fall after narrow victories. Kansas State and Purdue move up even after losing, for staying (relatively) close in score that they gain points. The Points column tells the story; all the teams in the overall top 25 are in the top 25 in the Points sub-rating, but their order gets shuffled by the addition of the Win-Loss rating:

lw  #  Team                         rec    rating
1   1. LSU                         (6-0)   11.21
9   2. Missouri                    (5-0)    7.58
3   3. South Florida               (5-0)    7.28
10  4. Ohio State                  (6-0)    7.14
5   5. Arizona St                  (6-0)    7.07
8   6. Boston College              (6-0)    6.38
2   7. California                  (5-0)    6.31
12  8. Cincinnati                  (6-0)    5.60
16  9. Kansas                      (5-0)    4.73
18 10. Illinois                    (5-1)    4.45
15 11. South Carolina              (5-1)    4.42
11 12. Connecticut                 (5-0)    4.26
14 13. Hawaii                      (6-0)    4.18
17 14. Oklahoma                    (5-1)    3.18
   15. Virginia Tech               (5-1)    3.14
21 16. Oregon                      (4-1)    3.04
25 17. West Virginia               (5-1)    2.97
6  18. Kentucky                    (5-1)    2.83
   19. Florida St                  (4-1)    2.76
   20. Wyoming                     (4-1)    2.73
4  21. Wisconsin                   (5-1)    2.68
   22. Virginia                    (5-1)    2.64
   23. Texas A&M                   (5-1)    2.60
13 24. Purdue                      (5-1)    2.54
   25. Indiana                     (5-1)    2.52

The Win-Loss rating more closely resembles the national polls, with teams moving up for winning and down for losing, as opposed to the Points rating in which a team may falls for a disappointing win and rise for a close loss to a good team. As you can see, Missouri and Ohio State jump for wins over winning teams, while South Florida doesn't budge. Kansas, Illinois, and South Carolina beat winning teams and all jumped over Hawaii, who beat a team with a losing record.

Illinois and South Carolina are the top-rated teams with a loss for good reason: both have lost to an undefeated team, and have defeated other teams who have only one loss. In fact, Illinois has beaten both 5-1 Wisconsin and 5-1 Indiana. Connecticut and Hawaii have weak enough schedules that they don't make the top ten. Hawaii in particular has a very low rating for a 6-0 team; they do, however, have the #1 scoring offense in the country, even adjusted for their weak competition:

#   Team                         rec    rating  off    def
1.  Hawaii                      (6-0)   10.36  47.35  36.99
2.  Oklahoma                    (5-1)   30.27  45.53  15.26
3.  Oregon                      (4-1)   24.98  45.19  20.21
4.  Florida                     (4-2)   26.67  44.01  17.34
5.  West Virginia               (5-1)   28.17  43.18  15.01
6.  Missouri                    (5-0)   25.18  43.13  17.95
7.  Cincinnati                  (6-0)   24.42  42.25  17.82
8.  Texas Tech                  (5-1)   11.92  42.23  30.31
9.  LSU                         (6-0)   33.04  41.94   8.90
10. California                  (5-0)   20.09  41.75  21.65

and Ohio State has the most dominant defense: the average team can expect to score only 3 points against them on a neutral field.

#   Team                         rec    rating  off    def
1.  Ohio State                  (6-0)   28.69  31.48   2.78
2.  LSU                         (6-0)   33.04  41.94   8.90
3.  Kansas                      (5-0)   28.93  41.51  12.58
4.  South Florida               (5-0)   22.66  35.90  13.23
5.  Auburn                      (4-2)   18.30  31.54  13.23
6.  Boise St                    (4-1)   16.23  29.62  13.39
7.  Georgia Tech                (3-3)    9.70  23.32  13.61
8.  Virginia Tech               (5-1)    8.52  23.35  14.82
9.  West Virginia               (5-1)   28.17  43.18  15.01
10. Oklahoma                    (5-1)   30.27  45.53  15.26

A surprise here is Boise State. They're more known for their offense, but their adjusted offense numbers are not far above average; their defense, though, rates high. Only one team has scored over 20 points on them, whereas last year it happened six times.

SportsRating Week 5 College Football Power Rating

Last week's #1 Oklahoma lost, so naturally they aren't #1 any more. Unsurprisingly, LSU takes over. Surprisingly, the Sooners only fall to #2.

lw  #  Team                     rec    rating  PR   W/L
2   1. LSU                     (5-0)   40.33    2    1
1   2. Oklahoma                (4-1)   36.02    1   17
24  3. South Florida           (4-0)   32.70    4    3
9   4. Ohio State              (5-0)   30.52    6   10
10  5. West Virginia           (4-1)   30.16    3   25
6   6. Kansas                  (4-0)   29.36    5   16
11  7. California              (5-0)   28.73   10    2
7   8. Cincinnati              (5-0)   27.74    9   12
3   9. Florida                 (4-1)   27.14    7   23
4  10. Oregon                  (4-1)   26.80    8   21
8  11. Southern Cal            (4-0)   26.29   11    7
12 12. Missouri                (4-0)   25.84   12    9
13 13. Kentucky                (5-0)   25.52   13    6
19 14. Arizona St              (5-0)   24.76   14    5
5  15. Hawaii                  (5-0)   22.95   15   14
   16. Connecticut             (5-0)   21.01   19   11
20 17. Illinois                (4-1)   20.99   16   18
23 18. Boston College          (5-0)   20.43   23    8
   19. Auburn                  (3-2)   20.09   17   30
15 20. Georgia                 (4-1)   19.66   18   20
16 21. South Carolina          (4-1)   18.38   22   15
14 22. Alabama                 (3-2)   17.88   21   47
18 23. Arkansas                (2-2)   17.42   20   67
   24. Florida St              (3-1)   15.97   25   28
   25. Colorado                (3-2)   15.85   24   40

The Sooners still rank #1 in the Points rating, but fall to 17th in the Win/Loss component. Mine isn't the only power rating (or any kind of rating, for that matter) having difficulty reconciling Oklahoma's performance Saturday with the rest of the season. Jeff Sagarin's USA Today power rating has Oklahoma fourth, #1 based on pure score margins, but a whopping 38th in the ELO Chess (score unimportant) rating.

South Florida leaps to #3 for beating West Virgina. The Mountaineers themselves actually move up this week, due to the fact that all their previous opponents won this week, with Maryland upsetting Rutgers. That was enough to make up for the loss, since relative strengths change a lot in the first half of the season.

Colorado even moves up to #25. Are they overrated just because of beating Oklahoma? Perhaps. But consider that their losses came to 1-loss Florida State and undefeated Arizona State and they aren't looking too bad anymore.

Win-Loss Component
lw  #  Team                         rec    rating
1   1. LSU                         (5-0)    8.00
15  2. California                  (5-0)    6.04
22  3. South Florida               (4-0)    5.83
9   4. Wisconsin                   (5-0)    5.79
11  5. Arizona St                  (5-0)    5.37
13  6. Kentucky                    (5-0)    5.35
10  7. Southern Cal                (4-0)    5.31
12  8. Boston College              (5-0)    5.25
4   9. Missouri                    (4-0)    4.96
7  10. Ohio State                  (5-0)    4.75
17 11. Connecticut                 (5-0)    4.26
18 12. Cincinnati                  (5-0)    4.14
20 13. Purdue                      (5-0)    3.94
16 14. Hawaii                      (5-0)    3.86
25 15. South Carolina              (4-1)    3.18
21 16. Kansas                      (4-0)    3.12
3  17. Oklahoma                    (4-1)    2.85
   18. Illinois                    (4-1)    2.76
   19. Nebraska                    (4-1)    2.59
24 20. Georgia                     (4-1)    2.58
5  21. Oregon                      (4-1)    2.54
6  22. Texas                       (4-1)    2.47
2  23. Florida                     (4-1)    2.46
8  24. Michigan St                 (4-1)    2.45
19 25. West Virginia               (4-1)    2.35

LSU remains #1 on the "politically correct" no-victory-margin Win-Loss component, followed by California, South Florida, and Wisconsin, all of whom beat previously undefeated teams last week. The rest of the undefeated teams move up to fill the void left by losing teams. Missouri didn't play, so they basically stood still and were passed. Although they did gain a small amount, probably from Illinois' victory over Penn State.

Hawaii barely nudges up despite winning; their schedule is starting to catch up with them. Kansas was idle and is passed by 1-loss South Carolina. Oklahoma falls all the way from #3 to #17, and spots 21 thru 25 are filled with plummeting teams that lost on Saturday. Worst case is Florida, who fall from #2 to #23 for losing to Auburn. Interestingly, Auburn rates higher on this list than Colorado, so one would think Oklahoma might rank below the Gators. You could guess that Florida's former opponents did better on Saturday than Oklahoma's, and that's partly true: both had 2 former opponents that lost, but Oklahoma's (North Texas and Utah State) didn't have any wins to begin with, so Oklahoma wasn't getting additional points from them last week anyway.

Here's a look at the Points-only component:

#  Team                         rec    rating  off    def
1.  Oklahoma                   (4-1)   33.17  50.27  17.10
2.  LSU                        (5-0)   32.32  42.26   9.93
3.  West Virginia              (4-1)   27.80  42.56  14.75
4.  South Florida              (4-0)   26.87  36.92  10.04
5.  Kansas                     (4-0)   26.24  40.63  14.39
6.  Ohio State                 (5-0)   25.76  32.18   6.41
7.  Florida                    (4-1)   24.67  43.33  18.65
8.  Oregon                     (4-1)   24.25  47.74  23.48
9.  Cincinnati                 (5-0)   23.60  41.32  17.72
10. California                 (5-0)   22.68  39.95  17.26
11. Southern Cal               (4-0)   20.98  37.34  16.36
12. Missouri                   (4-0)   20.88  45.51  24.63
13. Kentucky                   (5-0)   20.17  42.99  22.81
14. Arizona St                 (5-0)   19.39  38.42  19.03
15. Hawaii                     (5-0)   19.09  49.61  30.52
16. Illinois                   (4-1)   18.23  34.21  15.98
17. Auburn                     (3-2)   17.91  30.20  12.28
18. Georgia                    (4-1)   17.08  33.38  16.30
19. Connecticut                (5-0)   16.75  36.14  19.39
20. Arkansas                   (2-2)   16.66  40.95  24.28
21. Alabama                    (3-2)   16.59  35.44  18.84
22. South Carolina             (4-1)   15.19  30.15  14.96
23. Boston College             (5-0)   15.18  41.19  26.01
24. Colorado                   (3-2)   14.15  29.90  15.75
25. Florida St                 (3-1)   13.68  24.09  10.41

The offensive averages fell quite a bit this week: Hawaii, Oregon, Oklahoma, and Florida all lose about 5 points. Ohio State now has the #1 scoring defense in the country, adjusted for competition.

SportsRatings College Football Power Rating - Week 4

Boy do I have a butt-load of data for y'all. At some point I'm going to make a page that contains the current Power Rating and all its subcomponents, which will consist of (in some linkable, readable format): The overall Power Rating, the Points component, the Win-Loss component, and the offense and defense subcomponent of the Points subcomponent. Confused by now? Actually the offense and defense lists are more a "statistics" thing, so I might not cover them extensively in Power Rating posts. Additionally, I'm working on a Yards-based rating predictive system for both passing and rushing yards, which I've never seen done before.

I like how the Power Rating is shaping up, as it should, in Week IV. A lot of movement from last week but that will slow down.

lw  #  Team                    rec    rating  PTS  W/L
9   1. Oklahoma               (4-0)   47.02    1    3
3   2. LSU                    (4-0)   43.58    2    1
2   3. Florida                (4-0)   39.41    3    2
4   4. Oregon                 (4-0)   35.03    4    5
22  5. Hawaii                 (4-0)   32.67    5   16
8   6. Kansas                 (4-0)   31.95    6   21
1   7. Cincinnati             (4-0)   30.33    7   18
12  8. Southern Cal           (3-0)   29.44    9   10
19  9. Ohio State             (4-0)   29.30   10    7
   10. West Virginia          (4-0)   29.13    8   19
15 11. California             (4-0)   27.12   11   15
   12. Missouri               (4-0)   26.31   14    4
11 13. Kentucky               (4-0)   26.09   13   13
6  14. Alabama                (3-1)   24.83   12   31
17 15. Georgia                (3-1)   22.48   15   24
18 16. South Carolina         (3-1)   21.16   17   25
7  17. Rutgers                (3-0)   20.39   18   23
5  18. Arkansas               (1-2)   20.26   16   76
   19. Arizona St             (4-0)   19.88   20   11
   20. Illinois               (3-1)   19.44   19   26
   21. Wisconsin              (4-0)   19.14   22    9
16 22. Purdue                 (4-0)   18.32   21   20
14 23. Boston College         (4-0)   18.08   23   12
   24. South Florida          (3-0)   16.20   25   22
   25. Texas                  (4-0)   16.17   28    6

Oklahoma, LSU and Florida at #1, 2, and 3. Pretty reasonable, I'd say.

USC is a bit underrated but has only 3 games to average, so any flukes in game results are magnified. Nebraska's 4th quarter touchdowns, plus their subsequent near-loss to Ball State, reduced the value of the Trojans' win over them.

Offensive machines Hawaii and Oregon round out the top 5. Hawaii's 2 blowouts over Division I-AA opponents carry weight at this point and inflate their rating; as the season progresses, those wins will become nearly negligible. Oregon, however, has a win over Michigan that will likely only grow in impressiveness. But like Hawaii they don't rank well defensively, and they will be put to the test against California next week.

The second 5 includes USC, Ohio State, and West Virginia, but also Kansas and Cincinnati (last week's top team!). Both the Jayhawks and Bearcats are beating up on subpar competition, as reflected by their Win-Loss rankings of #21 and #18 respectively. Cincinnati's win over Oregon State is probably the slight difference in schedule strength. It will be interesting to watch these teams compete in conference play.

There's a glut of SEC teams in the second ten, from #13 to #16 we have Kentucky, Alabama, Georgia, and South Carolina. After complaining about the AP having Georgia ahead of South Carolina, it shows up in my own Power Rating (stupid computer! :). But this logjam makes sense; all these games were very close, with SC beating Georgia by 4, Georgia beating 'Bama by 3, Kentucky and Alabama beating #18 Arkansas narrowly. The SEC is going to be even more of a slog this year than usual. Arkansas is the only highly rated team with a losing record, due to their close games with the other rated schools, and a now-somewhat-impressive win over Troy. Their ranking based solely on winning and losing is #76, but when margin of loss/victory is used they rate much better.

A handful of Big Ten teams land at #20 thru #22, with #21 Wisconsin rated much lower (and probably, much more accurately) than they are in the national polls. Texas at #25 also seems to match their results so far more than a top ten rank.

The Win-Loss component isn't too significant yet, but it gives an interesting picture of which teams have defeated other good team so far:

Win-Loss Component:
#   Team                         rec    rating
1.  LSU                         (4-0)    6.92
2.  Florida                     (4-0)    5.19
3.  Oklahoma                    (4-0)    5.08
4.  Missouri                    (4-0)    4.69
5.  Oregon                      (4-0)    4.67
6.  Texas                       (4-0)    4.49
7.  Ohio State                  (4-0)    4.49
8.  Michigan St                 (4-0)    4.30
9.  Wisconsin                   (4-0)    4.25
10. Southern Cal                (3-0)    4.18
11. Arizona St                  (4-0)    4.17
12. Boston College              (4-0)    4.13
13. Kentucky                    (4-0)    3.94
14. Clemson                     (4-0)    3.92
15. California                  (4-0)    3.90
16. Hawaii                      (4-0)    3.62
17. Connecticut                 (4-0)    3.61
18. Cincinnati                  (4-0)    3.59
19. West Virginia               (4-0)    3.59
20. Purdue                      (4-0)    3.43
21. Kansas                      (4-0)    3.25
22. South Florida               (3-0)    2.98
23. Rutgers                     (3-0)    2.69
24. Georgia                     (3-1)    2.28
25. South Carolina              (3-1)    2.21

No major surprises here, except perhaps that Missouri's schedule is rated so high. The only difference in ratings among teams that are 4-0 is reflected by schedule difficulty. Thus we see LSU, who has beaten Virginia Tech and South Carolina, well ahead of the pack ratings wise. Points are awarded cumulatively, so the teams that are only 3-0 at this point will score lower, but USC still makes the top ten. All of the remaining undefeated teams are ranked, with Georgia and South Carolina scoring best among teams with a loss. Ratings for the Win-Loss component grow as the season progresses, making it more significant to the overall ratings, while the Points component stays roughly static (victory margins don't change much).

The Points component so closely reflects the overall ratings that it isn't worth taking up space with, but the Offense and Defense rankings are interesting. Here are the top ten teams in the country, ranked by offense and defense:

Offense
#   Team                         rec    rating
1.  Hawaii                      (4-0)   57.09
2.  Oregon                      (4-0)   55.35
3.  Oklahoma                    (4-0)   54.98
4.  Florida                     (4-0)   51.93
5.  Texas Tech                  (3-1)   47.24
6.  LSU                         (4-0)   46.57
7.  Clemson                     (4-0)   46.28
8.  Missouri                    (4-0)   46.13
9.  West Virginia               (4-0)   45.01
10. Arkansas                    (1-2)   44.07

Defense
#   Team                         rec    rating
1.  Florida St                  (2-1)    7.61
2.  Michigan                    (2-2)    9.67
3.  LSU                         (4-0)    9.91
4.  South Carolina              (3-1)   10.46
5.  Ohio State                  (4-0)   10.89
6.  Iowa                        (2-2)   11.11
7.  Rutgers                     (3-0)   11.27
8.  Cincinnati                  (4-0)   12.31
9.  Miami FL                    (3-1)   12.56
10. Kansas                      (4-0)   12.65

These results are close to simple scoring offense and scoring defense lists, but they are corrected for strength of schedule. For example, LSU is only 16th in scoring offense, but corrected for the tougher-than-average defense they've faced, they rank #6. Hawaii again is slightly overrated since 2 of its opponents have no comparison basis and factor too heavily at this point. In both these lists, the rating corresponds to play against an average opponent, so for offense, that's the amount of points the team is expected to score in an average game. For defense, it's how many points an average opponent should score against them.

On the defensive list, Florida State, which is only 24th in pure scoring defense, ranks best when corrected for the offensive strength of their opponents: they nearly shut out Colorado, and held Clemson to its lowest point total. Unfortunately for the Seminoles, their offense is ranked 116th out of 120 teams, making them perhaps the most schizmatic team in college football history.

Michigan is surprisingly number 2. Didn't they get shelled by both Appalachian State and Oregon? Yes, but in the next 2 weeks they shut out Notre Dame and held high-powered Penn State to 9 points. Also, Oregon's 39 was the Ducks' lowest output thus far. Weighing that, the algorith essentially disregards the first game as an anomaly. With such divergent games, Michigan's rating will be interesting to watch.

Another team that scores better here than in the pure averages is Miami, which, due to the bomb Oklahoma dropped on them, is only 39th nationally in scoring defense. But they come in at #10 when you correct for the fact that Oklahoma has one of the best offenses in the country. Miami's other games bear this rating out, as they held Texas A&M scoreless for 3 quarters and gave up only 12 points to their other 2 opponents.

And once again, as a post-script, I have to report on how Notre Dame ranks. It's not pretty, either, for you Fighting Irish fans. Overall the Irish rank 117th out of 120 teams, 111th in the Points rating and 114th in the Win-Loss rating. Their offense and defense both rank 105th, so if you're trying to decide which platoon to blame for their woes, it's a judgement call.

 

Power Rating - week 3

Week three and the Power Rating is starting to mesh. Still a few oddities but also some interesting results...

Combined Power Rating
lw   #  team                     rec   rating
21   1. Cincinnati              (3-0)   58.65
14   2. Florida                 (3-0)   54.72
5    3. LSU                     (3-0)   49.17
6    4. Oregon                  (3-0)   46.79
25   5. Arkansas                (1-1)   41.50
7    6. Alabama                 (3-0)   40.90
     7. Rutgers                 (3-0)   35.16
3    8. Kansas                  (3-0)   34.42
     9. Oklahoma                (3-0)   31.61
    10. Oregon St               (2-1)   30.75
15  11. Kentucky                (3-0)   30.57
    12. Southern Cal            (2-0)   28.68
    13. Troy                    (1-2)   27.88
    14. Boston College          (3-0)   26.41
    15. California              (3-0)   25.26
2   16. Purdue                  (3-0)   24.65
    17. Georgia                 (2-1)   24.63
11  18. South Carolina          (3-0)   24.02
4   19. Ohio State              (3-0)   23.91
16  20. Penn State              (3-0)   23.62
13  21. Michigan St             (3-0)   23.41
17  22. Hawaii                  (3-0)   22.08
    23. Air Force               (3-0)   20.87
    24. Louisville              (2-1)   19.40
    25. Utah                    (1-2)   18.68

Check out Cincinnati at #1...the Bearcats have blown out three opponents. Now, perhaps only one of them was any good, but you can say the same about Florida and LSU. It's still too early for an unbiased power rating to know that blowing out Virginia Tech or Tennessee should be rated better than crushing Oregon State (and who knows, it's possible we'll find out Oregon State is better than either of those teams).

Of course it's too early to declare Cincinnati a national title prospect. For now these results should be taken to mean: keep an eye on Cincinnati. They may be very good, and they certainly should be a factor in the Big East race. West Virginia, Louisville, and Rutgers just found a new stumbling block on their schedules.

Arkansas' rise to #5, ahead of Alabama, demonstrates how a 'pure points' power rating values games. Since the bulk of the rating at this point is determined by that component, here are the top 25 rated teams by that system:

Pure Points Component (HFA = 3.22)
#   team                     rec   rating
1.  Cincinnati              (3-0)   55.14
2.  Florida                 (3-0)   50.99
3.  LSU                     (3-0)   44.94
4.  Oregon                  (3-0)   43.64
5.  Arkansas                (1-1)   40.90
6.  Alabama                 (3-0)   37.29
7.  Rutgers                 (3-0)   32.39
8.  Kansas                  (3-0)   31.85
9.  Oregon St               (2-1)   29.64
10. Oklahoma                (3-0)   28.59
11. Troy                    (1-2)   27.50
12. Kentucky                (3-0)   26.71
13. Southern Cal            (2-0)   26.00
14. Georgia                 (2-1)   23.59
15. Boston College          (3-0)   22.92
16. California              (3-0)   22.27
17. Purdue                  (3-0)   22.09
18. Penn State              (3-0)   21.05
19. South Carolina          (3-0)   20.87
20. Ohio State              (3-0)   20.18
21. Michigan St             (3-0)   19.64
22. Hawaii                  (3-0)   19.28
23. Louisville              (2-1)   18.53
24. Utah                    (1-2)   18.14
25. Houston                 (1-1)   18.14

Arkansas lost at Alabama 41-38. But notice the home field advantage of 3.22. As far as the computer is concerned, that game demonstrated that Arkansas is slightly better than Alabama. It doesn't hurt the Razorbacks that first-week opponent Troy had a big win over Oklahoma State (and lost only to Arkansas and Florida). This puts Troy higher than their 1-2 record would seem to warrant, but the algorithm is unconcerned with the W and L column. This particularly holds true for Utah, who beat UCLA by 38, and had fairly close losses to Oregon State and Air Force.

The Win-Loss component is less radical, but offers a few surprises:

Win-Loss Component
#   team                      rec   rating
1.  LSU                      (3-0)    4.23
2.  Kentucky                 (3-0)    3.85
3.  Michigan St              (3-0)    3.77
4.  Ohio State               (3-0)    3.73
5.  Florida                  (3-0)    3.73
6.  Alabama                  (3-0)    3.61
7.  Texas                    (3-0)    3.59
8.  Cincinnati               (3-0)    3.51
9.  Boston College           (3-0)    3.49
10. Wisconsin                (3-0)    3.34
11. Missouri                 (3-0)    3.30
12. Clemson                  (3-0)    3.20
13. Air Force                (3-0)    3.14
14. Oregon                   (3-0)    3.14
15. South Carolina           (3-0)    3.14
16. West Virginia            (3-0)    3.01
17. Oklahoma                 (3-0)    3.01
18. California               (3-0)    2.99
19. Texas Tech               (3-0)    2.98
20. Hawaii                   (3-0)    2.80
21. Rutgers                  (3-0)    2.77
22. Connecticut              (3-0)    2.71
23. Southern Cal             (2-0)    2.68
24. Indiana                  (3-0)    2.61
25. Kansas                   (3-0)    2.56
26. Penn State               (3-0)    2.56
27. Texas A&M                (3-0)    2.56
28. Purdue                   (3-0)    2.56
29. Arizona St               (3-0)    2.55

What does this list tell us? LSU has, at this point, played the toughest schedule of any 3-0 team, according to these results. And how is schedule strength determined? By these very same ratings, of course! In other words, only wins and losses are counted. Good teams have lots of wins and few losses. Teams are ranked and re-ranked until the order converges.

Unsurprisingly, the top teams are all 3-0  (#29 Arizona State is the lowest ranked 3-0 team). USC is the lone 2-0 team in the top 25. Having more wins does yield a higher rating—this component is additive, not averaged like the Pure Points component. This fact means less as the season continues, but right now having 3 wins is a big advantage over having only 2. Again, for diversity between the two components, it's a good thing to have one that is additive and one that is averaged.

Compare Kansas' position in the two components. Being number 8 in Pure Points is a reflection of how big their margins of victory were; their #25 rank in Win/Loss reflects the fact that their opponents are a combined 2-7 (and two of their opponents played each other or that might be even worse).

*  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  * 

Next week is normally when the ratings start to really make sense. Though schedules are watered down under the 12-game scheme, week IV tends to be when the poorer teams riding a Division I-AA blowout start to fall by the wayside, and strength of schedule can be more accurately determined.

And by the way, in case you were wondering: Notre Dame comes in at #98 out of 119 teams so far: 96th in Pure Points and a dismal 116th in the Win/Loss component.

My various rating systems

I update three different rating systems on this site, which may (should?) cause some confusion. So this is a short explanation of each.

SportsRatings College Football Top 25

This is a subjective ranking of the top 25 teams in the nation. Generally speaking this would be how I would vote in the AP poll, based on which teams I think are the best and/or have accomplished the most during the season so far. That said, I do have a few rules (I should call them 'guidelines' since there can be exceptions) I try to follow, being:

- If Team A beats Team B and has an equal or better record, Team A should always be ahead of Team B, and

- New teams enter the Top 25 only by beating a Top 25 team, but

- Before any team with N+2 losses can stay in the Top 25, all teams with N losses should already be in the Top 25.

- I can break any of these rules if I strongly feel it is warranted, since this is a subjective ranking. The rules are there to keep it from being too subjective.

SportsRatings College Football Power Rating

This is a computerized, therefore algorithm-based rating system of my creation. Therefore it is fully objective, that is, there is no chance that a team I like better than another, or who has a better football reputation than another, or who I happened to watch on Saturday, gets any favoritism. Another upside of power ratings is all available data can be taken into consideration. The downside of any power rating is of course, only the available data can be used. In my case, this includes the scores of all Division I-A games, and location. No injury analysis, comparison of offensive lines and safeties, the Quarterback's mindset, etc.

To comlicate matters I use two different algorithms and combine them to make the final rating. One is a "pure points" or predictive element, the other a win/loss analysis. Jeff Sagarin's ratings use a similar two-component method; my points system is similar to his, while my win/loss element is quite different. Over the season I'll see how they compare as I've never combined the two rankings into one before.

The reason I use 2 ratings, and I suspect Sagarin came to the same conclusion earlier, is that neither alone sufficiently captures a team's quality. Pure points methods give the best indication of how "powerful" a team is, but if these teams lose, there isn't much penalty. Likewise, to rank by win/loss alone and ignore point spread gives a good indication of who has done well, but doesn't give as good of a predictive element, and teams that get lucky and win close games tend to be ranked too high. By combining the two it rewards teams for winning the close games, and also rewards teams that blow out opponents, which I think makes sense.

The win/loss component has a rating that generally increases over the season, gaining importance in the combined, final rating as weeks go by. This reflects the fact that, the later it is in the season, the more you can tell by looking at a team's record and who they played. Early on, until about the middle of the season, looking at scores can tell you a lot more than the record. Over time this evens out. I will start posting both elements of the rating this week.

Lineal Top 25

Also called "Linear" or "Daisy-Chain" sometimes, a Lineal ranking is one where the winner of a game replaces the loser in the rankings. Like in boxing, when the title holder is defeated, the winner becomes the title holder. The previous title holder falls to #2. The works the same for the entire top 25; if the #10 team is defeated by the #20 team, #20 jumps to #10 and 10 falls to 11.

I have a longer explanation of this from a few weeks ago. I don't put my stamp on this "ranking system" because it's not original to me, it's a simple idea that's been around a long time, and it's going to be the same for anyone who calculates it. The end result is more for amusement than a serious attempt at ranking the best teams.

SportsRatings college football power rating - week two

Craziness!

That's what week 2 power ratings give you. Because some of the most common, points-based systems, converge at weirdness. With 2 games (or 1) for each team to go by, the push and pull of the algorithm takes the teams' ratings to extremes sometimes. The results, to be kind, are...unexpected.

And the 2nd component, the win-based one, doesn't matter much right now. In fact, the top teams in that listing have just over 2 points, whereas the points system has teams in the 40s and 50s. So Week Two is a hodgepodge of nuttiness, be warned!

Note that I don't withhold results for the first few weeks; nor do I pad results with last year's rankings to make them seem "logical." This is the raw data. Week Three will make more sense, even yield some insights that might be missed due to impartiality and not having time to analyze each game. Week three will have it's oddities, too, but they will be far fewer, and farther between than this week. Here we go:

lw  tw  team                        record    rating
11   1. Georgia Tech                ( 2- 0)   59.01
16   2. Purdue                      ( 2- 0)   55.86
15   3. Kansas                      ( 2- 0)   52.89
25   4. Ohio U.                     ( 2- 0)   44.92
3    5. LSU                         ( 2- 0)   37.46
     6. Oregon                      ( 2- 0)   34.01
13   7. Alabama                     ( 2- 0)   30.74
18   8. Nebraska                    ( 2- 0)   29.12
     9. Wisconsin                   ( 2- 0)   29.09
10  10. Connecticut                 ( 2- 0)   28.30
    11. South Carolina              ( 2- 0)   27.72
11  12. Washington                  ( 2- 0)   27.37
22  13. Michigan St                 ( 2- 0)   27.27
13  14. Florida                     ( 2- 0)   27.20
20  15. Kentucky                    ( 2- 0)   26.70
4   16. Penn State                  ( 2- 0)   26.20
5   17. Hawaii                      ( 2- 0)   25.74
    18. Bowling Green               ( 1- 1)   25.43
    19. Central Michigan            ( 1- 1)   23.81
7   20. Cincinnati                  ( 2- 0)   20.28
    21. Texas                       ( 2- 0)   20.03
    22. Louisiana-Lafayette         ( 0- 2)   19.74
    23. Miami OH                    ( 1- 1)   18.20
    24. Missouri                    ( 2- 0)   17.95
    25. Arkansas                    ( 1- 0)   17.95


Now that's some mess, right? Not at the very top, necessarily. Georgia Tech actually could be the best team in the country, based on their two blowout wins so far. Remember, strength of schedule hasn't coalesced into anything meaningful yet. Beating up on Notre Dame and Samford still has objective value at this point, as long as Georgia Tech beat them up worse than their other opponents did. Likewise Purdue and Kansas blew out a couple of opponents each so far.

Then it gets weird. If you do the math, you can see how Ohio U, which beat Louisiana Lafayette and Gardner-Webb, is ranked so high. Strangely, it's because of South Carolina, who beat L-Lafayette in addition to Georgia. The Georgia win pushes SC's score up; Ohio's road win over LL ranks higher than SC's, which pushes Ohio up. Since both of LL's opponents rank high, LL's score goes up. Louisiana Lafayette lands in the top 25, at #22, due to this effect.

This kind of thing doesn't happen in later ratings when there is more information to go by (that is to say, it still happens, but it's only one game of many, and doesn't completely dictate several teams' rankings). It's one reason the BCS doesn't start until the middle of the season—many of the computer polls use information from the previous season for a while, until things start to "make sense."

LSU is #5, followed by several teams that are in the AP poll or have done well so far this season: Oregon, Alabama, Nebraska, Wisconsin. Connecticut comes in at #10, which makes me check to see exactly what they've done so far: beat Duke 45-14, and Maine, 38-0.

South Carolina, Washington, Michigan State, Florida, Kentucky, Penn State; all these teams are 2-0 and won their games by comfortable margins. Hawaii gets by on one blowout that was big enough for 2 games (their second victim, Louisiana Tech, whom they beat by one point, is down near #100). They were #5 last week, so the heartless rating dropped them 12 spots.

More statistical flukery lets in Bowling Green and Central Michigan, for losing to teams with bloated ratings. Miami, Ohio, is somehow related to this mess, but I'd have to step through the math to figure out how, and I don't have the patience to do that even with only 2 games per team at this point. Cincinnati actually falls, a lot, for crushing Oregon State. And quite kindly, Texas, who beat TCU, jumps to #21, and brings along the team they barely beat last week, Arkansas State. (Note, after letting the computer combine the ratings instead of me doing it by hand, Arkansas State lost a couple spots :)

Next week will be more sane. And the win-loss component will start to shake out an order; right now I don't list it here because there are so many groupings, basically ties. Also, I plan to do some interesting statistical things with run/pass yardage, which should give accurate rankings of teams in these areas, anaylses I haven't seen done anywhere else. So stay tuned.

Power Rating, week one

The early weeks of any power rating are either largely worthless or partially fraudulent. Using some form of "seeding" from last year's results is a common strategy to make the early weeks relevant. On the other hand if no such method is used, the first week's results are pointless; week 2 is interesting but sometimes spins out of control; and week 3 begins to coalesce but is still hampered by oddities.

After that, a good power rating has increasing relevance in making sense of the large amount of games played, mainly to give an objective comparison of the strength of all of the teams.

I boldly launch my power rating with no holdover data in the first week. What fun is a power rating if you withhold it for weeks until it has to chance to make sense of things? As my power rating includes 2 elements, one a points-based algorithm and another wins-based, the former at the points sees only one game, the latter only 2 groupings of teams - winners and losers. Still it's interesting to see what things look like with only that information to go on.

Points rule the day early on; as the season progresses, wins begin to take over in importance, moreso the further toward the top you get. So here's the first week - essentially an ordering of who won by the most points, tempered a bit by a chunky home field advantage (over 8 points right now!)

 1. Oklahoma                    ( 1- 0)   31.39
 2. Louisville                  ( 1- 0)   28.39
 3. LSU                         ( 1- 0)   27.60
 4. Penn State                  ( 1- 0)   26.39
 5. Hawaii                      ( 1- 0)   25.39
 6. Texas Tech                  ( 1- 0)   25.10
 7. Cincinnati                  ( 1- 0)   24.89
 8. Boise St                    ( 1- 0)   21.39
 9. Indiana                     ( 1- 0)   20.89
10. Connecticut                 ( 1- 0)   20.60
11. Georgia Tech                ( 1- 0)   20.10
tie Washington                  ( 1- 0)   20.10
13. Alabama                     ( 1- 0)   19.89
tie Florida                     ( 1- 0)   19.89
15. Kansas                      ( 1- 0)   19.39
16. UCLA                        ( 1- 0)   19.10
tie Purdue                      ( 1- 0)   19.10
18. Nebraska                    ( 1- 0)   17.89
tie Arizona St                  ( 1- 0)   17.89
20. Kentucky                    ( 1- 0)   16.89
21. West Virginia               ( 1- 0)   15.89
22. Michigan St                 ( 1- 0)   15.39
23. Rutgers                     ( 1- 0)   14.39
24. Tulsa                       ( 1- 0)   14.10
25. Ohio State                  ( 1- 0)   12.89
26. Air Force                   ( 1- 0)   12.39
tie Texas A&M                   ( 1- 0)   12.39
28. Miami FL                    ( 1- 0)   10.89
tie Southern Cal                ( 1- 0)   10.89
30. Navy                        ( 1- 0)   10.60
31. Northwestern                ( 1- 0)   10.39
tie TCU                         ( 1- 0)   10.39
33. Kent St                     ( 1- 0)    9.60
34. Pittsburgh                  ( 1- 0)    8.89
tie Vanderbilt                  ( 1- 0)    8.89
36. Arkansas Tech               ( 1- 0)    8.60
tie UNLV                        ( 1- 0)    8.60
38. North Carolina              ( 1- 0)    8.39
39. Southern Miss               ( 1- 0)    7.89
40. Iowa                        ( 1- 0)    7.5
41. New Mexico St               ( 1- 0)    7.39
42. Wisconsin                   ( 1- 0)    7.39
tie Louisiana Tech              ( 1- 0)    7.39
tie Oregon                      ( 1- 0)    7.39
tie Fresno St                   ( 1- 0)    7.39
tie Georgia                     ( 1- 0)    7.39
47. Wyoming                     ( 1- 0)    6.89
tie Arkansas                    ( 1- 0)    6.89
49. Nicholls St                 ( 1- 0)    6.10
tie Mississippi                 ( 1- 0)    6.10
tie Appalachian St              ( 1- 0)    6.10
tie Central Florida             ( 1- 0)    6.10

Starting next week (if it's relevant) I'll break up and separately list the two components.