Wolfpack's NCAA CF 2005 Power Ratings

[Wolfpack]

I've actually had this rating thing for a while, but never developed a program that could calculate for all the teams. Before, I just used an Excel spreadsheet and did it for just ACC teams against one another, but I've now developed a VB.Net app that can handle all the teams in Division I-A. The rating factors in who a team beats, where, and by how much (though the gain diminishes the larger the margin...IOW, there's more to gain by beating a team by 2 points versus 1 point, as opposed to 32 points and 31 points, though a 32-point win has more overall value than any smaller margin). The points accumulated each game are averaged out so that teams who play more or fewer games do not gain an advantage.

So, without further ado, here's the present Top 25 by my ranking system (the far right column is the average points per game by the rating system for each team):


1 Southern California 4-0 15.858
2 Virginia Tech 5-0 12.476
3 Penn State 5-0 9.858
4 Florida State 4-0 8.925
5 Georgia Tech 3-1 8.825
6 Alabama 5-0 8.574
7 Texas 4-0 8.472
8 Louisiana State 2-1 8.271
9 Notre Dame 4-1 8.171
10 Georgia 4-0 8.104
11 Miami (Florida) 3-1 7.617
12 Michigan State 4-1 7.461
13 Wisconsin 5-0 7.203
14 Colorado 3-1 6.726
15 UCLA 4-0 6.515
16 Florida 4-1 6.404
17 Ohio State 3-1 6.025
18 Minnesota 4-1 5.908
19 California 5-0 5.022
20 West Virginia 4-1 4.972
21 Arizona State 3-2 4.829
22 Texas Tech 4-0 4.683
23 Nebraska 4-0 4.595
24 Oregon 4-1 4.442
25 Michigan 3-2 4.226


It should be noted, this rating system does not predict any outcomes. It merely reflects at a given point in time the relative strength of one team versus another based on my criteria.

People may wonder why Texas is only 7th or why Georgia Tech is 5th. Well, the short answer is the fact that Texas has routed two teams at home (ULL and Rice) who have a combined record of 1-6. Home wins are not valued as much as road wins and beating up ULL and Rice do not net large amounts of points (in Rice's case, it's actually a little reduced because Rice is winless). This early in the year, their trashing of Missouri is more impressive than their win at Ohio State, but if Ohio State continues to win while Missouri struggles, then the Ohio State win will become more prominent as time goes along.

Georgia Tech, meanwhile, gets to be #5 because they won by 9 at Auburn (4-1) and they also beat UNC and UConn and both have .500 or better records. Their lone loss was a trashing at Virginia Tech, which is not as damaging as it could be right now. This is because Virginia Tech is unbeaten and they lost at Blacksburg. Had Virginia Tech had one loss and whipped them 51-7 in Atlanta, they would not be anywhere near this high in the ranking.

This is, as most computerized rankings are, going to be more accurate as more games are played and teams win or lose to add to or subtract from the efforts of other teams.

[wade moore]

As a Tennessee fan, I have to be curious.. how are they not on the list when they beat your #8 team and their only loss is to your #16 team?

I don't mean this as a "Defend yourself!" question, but as a "how does the math work?" question...

[MrBigglesworth]

PSU at 3? I like this system :)

[Huckleberry]

hxxp://www.hornfans.com/recruiting/huckleberry/

That link will take you to my ratings pages. I'm not opposed to shameless self-promotion. I'm always curious to see another approach to sports ratings.

[Wolfpack]

Tennessee's just off the list at #27. The reason they don't have a higher rank is because their winning margins I think are hurting them somewhat. Their most impressive win is a 3-point win at LSU. I've tried to diminish the impact margin has overall, but with all the close victories and losses, they haven't built up much cache in the computer yet. Florida's loss last week also hurt them as they now take the full credit for the loss in Gainesville. In this rating, it's most important to win, win comfortably, and do it on the road. That's where the big points lie. It's why USC is up top, because they've won by double-digit margins on the road in three of their four games.

[Wolfpack]

Penn State gets to be #3 because, well, they're unbeaten for one. For another, they've beaten opponents handily. That 30-point thrashing of Minnesota is especially strong, easily the best they have factoring into their rating right now.

[wade moore]

Tennessee's just off the list at #27. The reason they don't have a higher rank is because their winning margins I think are hurting them somewhat. Their most impressive win is a 3-point win at LSU. I've tried to diminish the impact margin has overall, but with all the close victories and losses, they haven't built up much cache in the computer yet. Florida's loss last week also hurt them as they now take the full credit for the loss in Gainesville. In this rating, it's most important to win, win comfortably, and do it on the road. That's where the big points lie. It's why USC is up top, because they've won by double-digit margins on the road in three of their four games.
Makes sense... I would expect that if UT continues to win, it will even out in the end...

Although I am REALLY worried about UGA this weekend...

[cartman]

hxxp://www.hornfans.com/recruiting/huckleberry/

That link will take you to my ratings pages. I'm not opposed to shameless self-promotion. I'm always curious to see another approach to sports ratings.

Hey Huck,

Any reason you don't have the private high schools in Texas ranked? I'm guessing it is probably either a lack of quantity of teams, or lack of consistent info.

[Huckleberry]

Hey Huck,

Any reason you don't have the private high schools in Texas ranked? I'm guessing it is probably either a lack of quantity of teams, or lack of consistent info.
Lack of info. I still haven't found anywhere that will consistently post good information on their alignments and scores.

[Wolfpack]

Rankings after the games of the weekend of October 8


RK LW
1 1 Southern California 5-0 15.950
2 2 Virginia Tech 6-0 13.535
3 8 Louisiana State 3-1 13.141
4 3 Penn State 6-0 11.639
5 6 Alabama 5-0 11.457
6 10 Georgia 5-0 11.429
7 4 Florida State 5-0 10.091
8 7 Texas 5-0 10.024
9 9 Notre Dame 4-1 9.630
10 24 Oregon 5-1 9.601
11 20 West Virginia 5-1 8.739
12 14 Colorado 4-1 7.935
13 15 UCLA 5-0 7.805
14 18 Minnesota 5-1 7.788
15 11 Miami (Florida) 4-1 7.692
16 16 Florida 5-1 7.326
17 NR Texas Christian 5-1 7.151
18 12 Michigan State 4-1 6.728
19 22 Texas Tech 5-0 6.680
20 13 Wisconsin 5-1 6.507
21 5 Georgia Tech 3-2 6.257
22 NR Baylor 4-1 5.606
23 NR Boston College 5-1 5.577
24 17 Ohio State 3-2 5.324
25 NR Fresno State 3-1 5.064

Dropped out:
Cal (19)
Arizona State (21)
Nebraska (23)
Michigan (25)

As can be seen, there will be some jumping around week to week. USC and VT remain atop the list with solid home wins, while LSU gets a nice boost from winning big on the road at Vanderbilt (not often that one says something like that, but it's the case this year). In spite of winning at home, Penn State slips to fourth because a somewhat narrow home win over a team now 3-2 won't give a large boost. Alabama steps up to #5 in spite of not playing this weekend because 3 of the 5 teams they beat won this weekend, thus improving their score. Strength of schedule is somewhat of a factor in my ratings (basically a team gains the benefits of another team's wins when they beat them and is penalized for another team's losses if they lose to them, so it always behooves teams to beat good teams and never lose to bad teams, though at this point, it's still being sorted out what "good" and "bad" actually are).

[MrBigglesworth]

How exactly did LSU move up so much? They don't have any wins against teams in your top 25, and they lost a game to a team not even in your top five. I don't see where their high ranking is coming from. I just don't see a lot in their statistical profile.

[Wolfpack]

LSU got a very big bounce from winning big at Vanderbilt. Like I've noted before, I place emphasis on winning on the road by decent margins. Vanderbilt had four wins, which also kicked that point total up a bit. By the system's judgement, beating a 4-2 Vandy team in Nashville by 28 is 2.5 times better than beating a 3-3 Arizona State team in Tempe by 4. Now, if Vandy continues to slide from here on in, while ASU reels off some wins, that may change in the final accounting (actually, a quick check indicates that Vandy must end up with no more wins while ASU gets 8 wins to make the ASU win better overall between the two).

It is true that LSU's wins look totally unimpressive against who I have "ranked". However, it can't simply be viewed as rank vs. rank in this arrangement because the computer isn't considering ranking at all in the math, just wins (or losses), margin, and location. I suppose it is a limitation of the system in that it only considers one degree of separation in schedule strength (LSU to Vandy or LSU to ASU, rather than to LSU to Vandy to MTSU or LSU to ASU to USC).

[Young Drachma]

It's interesting and adds to the debate. Good stuff.

[Wolfpack]

Rankings after the games of the weekend of October 15


RK LW
1 1 Southern California 6-0 18.561
2 2 Virginia Tech 6-0 16.059
3 6 Georgia 6-0 15.501
4 8 Texas 6-0 13.533
5 3 Louisiana State 4-1 12.596
6 6 Alabama 6-0 11.832
7 4 Penn State 6-1 11.456
8 13 UCLA 6-0 10.156
9 10 Oregon 6-1 10.110
10 11 West Virginia 6-1 9.582
11 20 Wisconsin 6-1 9.280
12 9 Notre Dame 4-2 9.188
13 19 Texas Tech 6-0 9.013
14 7 Florida State 5-1 8.960
15 24 Ohio State 4-2 8.077
16 15 Miami (Florida) 5-1 7.746
17 17 Texas Christian 6-1 7.506
18 23 Boston College 6-1 6.939
19 12 Colorado 4-2 6.902
20 NR Nebraska 5-1 6.863
21 21 Georgia Tech 4-2 6.818
22 NR Oklahoma 3-3 6.603
23 25 Fresno State 4-1 6.474
24 14 Minnesota 5-2 6.318
25 16 Florida 5-2 5.982

Dropped out:
Michigan State (18)
Baylor (22)


The top two remain unchanged, but there's fluctuation after that as Georgia and Texas are finally beginning to reap the reward of being unbeaten. Both teams also absorbed good schedule strength by beating other good teams.

Interestingly, the big mover is Ohio State, who went up nine places. They absorbed the strength from beating Michigan State and have been fairly fortunate in that one of their two losses is to Texas by just three points. Texas's unbeaten record actually cancels out the effect of the home loss in the math being employed.

Wisconsin also moved up nine with the stunning win at Minnesota and they'd be even better off if they hadn't coughed up the game at Northwestern. But then, Badger fans already knew that.

[Wolfpack]

Rankings after the games of October 22


RK LW
1 1 Southern California 7-0 19.967
2 4 Texas 7-0 18.619
3 2 Virginia Tech 7-0 18.447
4 3 Georgia 7-0 14.618
5 7 Penn State 7-1 13.197
6 6 Alabama 7-0 12.923
7 15 Ohio State 5-2 12.040
8 8 UCLA 7-0 11.700
9 9 Oregon 7-1 11.480
10 5 Louisiana State 5-1 11.413
11 12 Notre Dame 5-2 10.785
12 10 West Virginia 6-1 10.705
13 17 Texas Christian 7-1 10.693
14 11 Wisconsin 7-1 9.811
15 16 Miami (Florida) 5-1 9.916
16 14 Florida State 6-1 8.996
17 NR Missouri 5-2 8.814
18 19 Colorado 5-2 8.425
19 18 Boston College 6-1 8.089
20 22 Oklahoma 4-3 7.625
21 13 Texas Tech 6-1 7.430
22 24 Minnesota 5-2 7.422
23 23 Fresno State 5-1 7.392
24 NR Texas-El Paso 5-1 7.245
25 NR Northwestern 5-2 7.113

Dropped out:
Nebraska (20)
Georgia Tech (21)
Florida (25)


Texas finally gets the big win against the good opponent and jumps over everybody in front of them except USC to get the #2 spot. It also helps that their win at Ohio State is now beginning to carry a bit more weight as the Buckeyes continue to win with the passing weeks.

Some will see Oklahoma and wonder why they are hanging around at 4-3. Simple, really. Their strength-of-schedule (as calculated by me) is the toughest at this point. All the teams they've lost to (Texas, UCLA, and, yes, Tulsa) have kept winning, thus limiting the damage they took from those losses.

[timmynausea]

Oklahoma beat Tulsa. They lost to TCU.

[Wolfpack]

Ah, thanks. My goof. Wrong "T" team. So much the better for Oklahoma then. It means that TCU isn't losing, thus not harming the Sooners, while Tulsa keeps winning, helping the Sooners.

[JeeberD]

I dig your rankings... :)

[Wolfpack]

I was expecting you'd show up here sooner or later... :)

[JeeberD]

Oh, but we're UTEP, not Texas-El Paso. You didn't list UCLA as California-Los Angeles, did you?

[Wolfpack]

Yeah, well...it's something I picked up doing because all the historic data I look at tends to use "Texas-El Paso" rather than UTEP. UTEP's not alone. I use "Nevada-Las Vegas" as well. UCLA is about the only exception to this rule. I wonder if they should just officially dub the school UCLA since it is hardly, if ever, called, "The University of California at Los Angeles".

[JeeberD]

I think that UTEP petitioned the NCAA a few years ago to be able to use UTEP as their official name rather than The University of Texas at El Paso and it was granted. Most media sources seem to list us as UTEP now, but a few (such as the Dallas Morning News :mad: ) still list us as Texas-El Paso.

And I'm not trying to give you a hard time or anything, I just really dislike that format...

[ISiddiqui]

Very cool... it seems as the year has gone on, the power rankings are 'fitting' more in line with the polls and other (personal) ranks.

[Wolfpack]

Rankings after the games of October 29


RK LW
1 2 Texas 8-0 21.496
2 3 Virginia Tech 8-0 20.943
3 1 Southern California 8-0 20.281
4 7 Ohio State 6-2 15.426
5 6 Alabama 8-0 13.747
6 5 Penn State 8-1 13.651
7 9 Oregon 7-1 13.171
8 4 Georgia 7-1 12.975
9 8 UCLA 8-0 12.501
10 11 Notre Dame 5-2 12.077
11 10 Louisiana State 6-1 11.716
12 12 West Virginia 6-1 11.614
13 13 Texas Christian 8-1 11.525
14 21 Texas Tech 7-1 10.826
15 20 Oklahoma 5-3 10.758
16 23 Fresno State 6-1 10.265
17 14 Wisconsin 8-1 10.250
18 18 Colorado 6-2 9.937
19 16 Florida State 7-1 9.890
20 15 Miami (Florida) 6-1 9.864
21 NR Michigan 6-3 9.109
22 NR Florida 6-2 9.026
23 NR Louisville 5-2 8.856
24 19 Boston College 6-2 8.527
25 24 Texas-El Paso 6-1 7.141

Dropped Out:
Missouri (17)
Minnesota (22)
Northwestern (25)


Texas is finally at #1. That's not too big a surprise. The surprise is VaTech also leapfrogging USC and moving up to #2. Right now, the schedule strength is dragging on USC. However, that will change as their closing stretch features current #16 Fresno State and #9 UCLA. In the same way, Texas's SOS is pretty good, but will get eroded because they draw Baylor, Kansas, and A&M to finish out. Baylor is the highest of those three at #59 right now. Virginia Tech also faces a slightly weaker close, though they do get Miami next week. Both UNC and UVa are in the 50s, rankings-wise.

People may also wonder why Ohio State is now ahead of Penn State, a team they lost to. Ohio State continues to benefit from Texas. That is their only other loss, also by three points. They're also getting a little bit of a push from Miami of Ohio, who have finally gotten on a winning track. They've also played almost exclusively teams at or over .500 overall in the Big 10 (though Indiana is sinking fast right now). They will get a mild downgrade in strength playing Illinois coming up, though. Penn State, on the other had, is getting held back increasingly by a weak schedule. They unfortunately have had to absorb Big 10 bottom-feeders Illinois and Purdue in the last two weeks. In addition, their OOC slate is uninspiring, with #47 South Florida being the best of those.

Beyond that, Oregon has quietly, (almost unnoticed, even) crept up to #7. UCLA's living dangerously allows them to win, but they aren't building up in the MOV aspects because of it, which is what's probably keeping them from moving up despite being unbeaten. Oklahoma also continues the steady climb back to respectability.

Also, no one can accuse me of ACC homerism. The next best team after VT is FSU at #19. FSU and Miami continue backsliding even while winning. Again, it is SOS. Duke and Temple are just killing Miami's ratings, while The Citadel and Syracuse are doing great harm to FSU's. (I deal out severe punishment for playing D-IAA teams. Essentially, they're treated as winless teams for rating purposes.)

[Wolfpack]

Rankings after the games of November 5

RK LW
1 1 Texas 9-0 25.102
2 3 Southern California 9-0 22.169
3 2 Virginia Tech 8-1 19.529
4 20 Miami (Florida) 7-1 17.547
5 6 Penn State 9-1 16.960
6 4 Ohio State 7-2 15.774
7 5 Alabama 9-0 15.408
8 7 Oregon 8-1 14.938
9 8 Georgia 7-1 13.681
10 13 Texas Christian 9-1 12.883
11 10 Notre Dame 6-2 12.851
12 14 Texas Tech 8-1 12.846
13 12 West Virginia 7-1 11.948
14 18 Colorado 7-2 11.482
15 15 Oklahoma 5-3 11.388
16 11 Louisiana State 7-1 11.039
17 23 Louisville 6-2 10.691
18 17 Wisconsin 8-2 10.470
19 21 Michigan 6-3 10.199
20 16 Fresno State 7-1 10.009
21 22 Florida 7-2 9.645
22 NR South Florida 4-3 9.013
23 9 UCLA 8-1 8.709
24 NR Georgia Tech 6-2 8.624
25 24 Boston College 6-3 8.486

Dropped out:
Florida State (19)
Texas-El Paso (25)


A few surprises this week. The big one is that in spite of losing by 20 at home, VT is still 3rd. Miami jumped a long way up from where they were, but it wasn't enough to get over the Hokies. One of the inherent flaws in computer rankings. Then again, pollsters are known to have trouble doing the same thing in the aggregate, as well. The big thing that continues to drag on Miami is that Temple and Duke are a combined 1-19. VT also has Duke weighed in, but their weakest OOC opponent is Ohio at 4-5. Now, it could very well be as the season finishes out and Miami and VT both win out, Miami can actually pass the Hokies because they will now reap the benefit every time VT wins a game.

As an interesting side note, no matter what components I choose to use (margins, no margins, location matters or doesn't) VT still finishes ahead of Miami. It's not just me, either. Massey's website has a conglomerate ranking of 80-odd computer rankings and the consensus still has VT rated ahead of Miami.

Another curiosity is South Florida, which jumped in at #22 this week. They're only 4-3, though. So, how do they get in here? Again, it's the schedule. They've got one crap game in Florida A&M, which for rating purposes has a record of 0-7. But, they beat Louisville and beat Rutgers, both of whom have 6 wins, and they beat Rutgers at Rutgers, another boost. In addition their remaining OOC opponents were Penn State and Miami (both losses, but a combined 16-2 and both were on the road, so the damage is limited there) and a win against a now bowl-eligible Central Florida. Their third loss was to Pittsburgh, who seems to have gotten things righted a bit and now stand at 4-5 after having a horrid start to the year.

However, USF's stay in the Top 25 should be short-lived. They play 1-7 Syracuse next week which should be another drag on their rating, win or lose. They might be able to get back into the Top 25 after that, though, as they play .500 teams Cincinnati and UConn and then get a big boost schedule-wise with a season-ending game against West Virginia.

[Butter]

Very nice job.... always interesting.

[MikeVick7]

What's keeping LSU so low?

[Wolfpack]

The drag on LSU right now is schedule strength. Their only loss is to Tennessee, but Tennesee keeps losing and with every game they lose it is really hurting LSU. Time was, it didn't inflict too much pain (look upthread at prior rankings to see the Tigers at a much higher station than now). In addition, they've got Vanderbilt (a 4-5 team that hasn't won since they played them), Mississippi State (2-7), North Texas (2-6), and a I-AA game against Appalachian State (0-8 for rating purposes). That's a combined 8-26 they aren't getting a lot of benefit from. They do have Florida and they do have Auburn counting for them, but right now, that's it. They also have Arizona State, who was expected to be a good team. However, the Sun Devils are scraping along at 5-4, hardly the expected mark of a team that was supposed to challenge USC in the Pac-10.

Looking ahead, they do have the big tilt with Alabama, which, if they win, will get them a lot of points (on the road against an unbeaten opponent). The schedule sags a bit again after that as they close with Mississippi and Arkansas (combined record of 5-11).

[MikeVick7]

Ahh, yeah that makes a lotta sense. I haven't been too high on them this year either.

[JeeberD]

BOOOOOOOOOOOO!!!!! ;)

[Wolfpack]

I thought about putting in an apology, Jeebs. However, looking at UTEP's middling excuse for a schedule, I've got nothing to be sorry about. ;)

[JeeberD]

It's true. Texas Southern next week?!?!? Ugh...

[st.cronin]

Hey now, Texas Southern is a football power, having produced such NFL greats as Michael Strahan. I think.

[JonInMiddleGA]

(I deal out severe punishment for playing D-IAA teams. Essentially, they're treated as winless teams for rating purposes.)

Just out of curiosity, and since you mentioned trying variations on the formulas, have you looked at the effects if you simply ignore the I-AA games on the schedule? i.e. don't treat them like winless teams, just don't consider them at all.

No biggie, just idly curious.

[Wolfpack]

Rankings after the games of November 12

RK LW
1 2 Southern California 10-0 28.226
2 1 Texas 10-0 27.830
3 3 Virginia Tech 8-1 21.482
4 4 Miami (Florida) 8-1 19.889
5 5 Penn State 9-1 18.261
6 6 Ohio State 8-2 18.175
7 11 Notre Dame 7-2 16.854
8 8 Oregon 9-1 15.861
9 13 West Virginia 8-1 15.236
10 7 Alabama 9-1 15.016
11 10 Texas Christian 10-1 14.172
12 17 Louisville 7-2 14.004
13 16 Louisiana State 8-1 13.683
14 9 Georgia 7-2 13.029
15 15 Oklahoma 6-3 12.436
16 20 Fresno State 8-1 12.418
17 19 Michigan 7-3 10.919
18 NR Auburn 8-2 10.786
19 23 UCLA 9-1 10.762
20 12 Texas Tech 8-2 10.631
21 14 Colorado 7-3 10.578
22 25 Boston College 7-3 10.522
23 NR Minnesota 7-3 9.928
24 22 South Florida 5-3 9.793
25 NR Toledo 7-2 9.243

Dropped out:
Wisconsin (18)
Florida (21)
Georgia Tech (24)


*A pretty quiet week overall. The top six barely changed at all, the only difference being a flip-flop as USC moved in front of Texas again (remember, Texas's schedule strength is expected to worsen while USC's will improve).

Even the losses by Alabama and Georgia didn't disturb things a whole lot because the margins were narrow. Alabama only dropped three places based on the OT loss at home to LSU, which, incidentally, isn't a bad team to lose to. Georgia slid five places to 14 after losing to Auburn by a point. Auburn finally pops into the poll at #18, though. They've been hanging around just off the fringe for most of the season. LSU, meanwhile, doesn't get much bounce at all from the Alabama win, again because of the narrow margin involved and because it's getting countered by the weaker overall schedule (same principle is in play WRT to Miami and Virginia Tech right now, though that is a rather extraordinary case of a 20-point win being countered by having Duke and Temple pulling on the average the other way).

The biggest shufflings occurred near the bottom as both Colorado and Texas Tech tumbled from perches just outside the top-10. Wisconsin, Florida, and Georgia Tech are bounced out by losses, which actually allows South Florida to cling to the bottom of the Top 25 and introduces a new mid-major, Toledo, to the list.

To JIMGA, I haven't had a chance to program and therefore project what the outcome would be if I removed I-AA games from the schedule. If I had to guess, the results shouldn't make a lot of difference as just about all teams will benefit from not having a I-AA team weighing the average down either winning or losing. It would strengthen those teams that have played I-AA teams while weakening those who played those who had (because they'd have one less win in most cases to count into the system). Therefore, some shuffling would be expected, but the top teams probably wouldn't change a whole lot since most played a full D-IA schedule.

[Wolfpack]

Dola...

Going through my rankings a bit more, this is actually the only portion of the system (the full one with partial margin, home field, and record) that actually ends up with USC on top. Every other instance (taking some components but not all three or using a different weighting on the margin) has Texas on top. Only other ranking system I have that has USC on top is my variation on ELO. Interesting.

[Wolfpack]

Rankings after the games of November 19

RK LW
1 1 Southern California 11-0 30.209
2 2 Texas 10-0 29.554
3 3 Virginia Tech 9-1 24.901
4 5 Penn State 10-1 20.653
5 6 Ohio State 9-2 20.244
6 4 Miami (Florida) 8-2 18.569
7 8 Oregon 10-1 16.954
8 7 Notre Dame 8-2 16.273
9 11 Texas Christian 10-1 15.959
10 9 West Virginia 8-1 15.489
11 13 Louisiana State 9-1 15.370
12 14 Georgia 8-2 14.650
13 12 Louisville 7-2 14.514
14 22 Boston College 8-3 13.624
15 10 Alabama 9-2 13.393
16 18 Auburn 9-2 13.009
17 19 UCLA 9-1 12.134
18 15 Oklahoma 6-4 11.859
19 16 Fresno State 8-2 11.823
20 20 Texas Tech 9-2 11.762
21 21 Colorado 7-3 10.768
22 17 Michigan 7-4 10.645
23 NR Georgia Tech 7-3 10.271
24 NR Iowa 7-4 10.231
25 24 South Florida 6-3 10.162

Dropped out:
Minnesota (23)
Toledo (25)


Yes, I know, it's been a couple of weeks, but in the interest of full disclosure, here is what the numbers looked like before last weekend. (Was out of town all last week, so no chance to update until now)

[ISiddiqui]

I'm thinking South Florida drops out after this week's games are counted :D.

[Wolfpack]

Rankings after the games of November 26


RK LW
1 1 Southern California 11-0 31.075
2 2 Texas 11-0 30.817
3 3 Virginia Tech 10-1 25.411
4 4 Penn State 10-1 20.985
5 5 Ohio State 9-2 20.570
6 7 Oregon 10-1 18.728
7 6 Miami (Florida) 9-2 18.472
8 12 Georgia 9-2 18.151
9 8 Notre Dame 9-2 17.478
10 10 West Virginia 9-1 17.390
11 9 Texas Christian 10-1 17.097
12 11 Louisiana State 10-1 15.813
13 15 Alabama 9-2 15.491
14 13 Louisville 8-2 13.490
15 18 Oklahoma 7-4 13.295
16 14 Boston College 8-3 13.266
17 16 Auburn 9-2 13.034
18 20 Texas Tech 9-2 12.670
19 17 UCLA 9-1 12.349
20 NR Minnesota 7-4 11.918
21 19 Fresno State 8-3 11.691
22 NR Florida 8-3 11.393
23 22 Michigan 7-4 10.952
24 NR Wisconsin 9-3 10.658
25 NR Boise State 9-3 10.617

Dropped out:
Colorado (21)
Georgia Tech (23)
Iowa (24)
South Florida (25)


The top 5 remain unchanged from last week, but there is some shuffling in places below that. If anyone is curious as to why Minnesota dropped out last week and then suddenly jumped back in again and moved up to #20 this week, it is because I discovered an error in my schedule input that had them beating Tulane and not Tulsa earlier in the year, which makes a big difference for their strength of schedule and consequently for the points they've accumulated this year. I went back over the schedule (easy to do at this point since just about everyone has played 11 games) and found a couple of other small errors, but nothing affecting the rankings a great deal otherwise. At any rate, this ranking is now reflective of accurate data for all games played to date.

[Wolfpack]

Out of curiosity, here's how the conferences stack up compared to one another. It's simply an average of the individual teams in the conference:


1 Big 12 6.427
2 Big 10 6.054
3 Pac 10 4.863
4 ACC 4.659
5 SEC 3.686
6 Big East 2.126
7 Mountain West -0.015
8 Conference USA -2.377
9 MAC -3.562
10 WAC -5.558
11 Sun Belt -9.787


The Big 12 currently rates highest, but a closer check reveals that if you drop Texas, the average rating for the Big 12 is cut in half. Similarly, dropping USC from the Pac-10's average reduces it to below the Big East. Then again, this is a bit of cherry-picking. You could add and remove teams from other conferences with somewhat similar effects.

[Wolfpack]

Rankings after the games of December 3, 2005

RK LW
1 1 Southern California 12-0 32.550
2 2 Texas 12-0 31.043
3 3 Virginia Tech 10-2 23.105
4 4 Penn State 10-1 20.985
5 5 Ohio State 9-2 20.570
6 8 Georgia 10-2 19.637
7 10 West Virginia 10-1 18.940
8 6 Oregon 10-1 18.728
9 7 Miami (Florida) 9-2 18.472
10 9 Notre Dame 9-2 17.809
11 11 Texas Christian 10-1 17.097
12 13 Alabama 9-2 15.157
13 14 Louisville 9-2 14.560
14 12 Louisiana State 10-2 13.896
15 15 Oklahoma 7-4 13.289
16 18 Texas Tech 9-2 13.114
17 17 Auburn 9-2 13.049
18 16 Boston College 8-3 12.939
19 20 Minnesota 7-4 12.650
20 22 Florida 8-3 12.134
21 24 Wisconsin 9-3 11.232
22 25 Boise State 9-3 11.196
23 19 UCLA 9-2 11.024
24 23 Michigan 7-4 10.952
25 NR Iowa 7-4 10.437

Dropped out:
Fresno State (21)


The final regular season numbers are now posted. As expected, USC and Texas are #1 and #2, but in spite of losing the ACC championship game to FSU, Virginia Tech still doggedly clings to #3. Because VT's strength has been in victory margins, it's awfully difficult for them to slide, especially when they lost by five points on a neutral field (treated as a road loss for rating purposes). If margins are not considered, VT moves down the list.

Penn Stae and Ohio State hold their ground at #4 and #5, mainly because they were not affected by what happened over the weekend (IOW, nobody they played played anybody this weekend).

Georgia and West Virginia leapfrog Oregon, Miami, and Notre Dame to move up to #6 and #7, respectively.

After that, aside from the LSU and UCLA losses directly dropping those teams a few places, there was some shuffling around as teams benefitted or were penalized by beating or losing to teams that played this weekend. (BC, for example, took a hit because their 20-point loss to VT got doubled with VT's second loss of the year).

As a further exercise of the rankings and ratings, here's the bowl season and how my rankings compare the teams that are playing one another:

New Orleans
Southern Mississippi (6-5, 4.724--#42) vs. Arkansas State (6-5, -6.390--#87)

GMAC
UTEP (8-3, 1.302--#55) vs. Toledo (8-3, 5.649--#38)

Las Vegas
BYU (6-5, 1.606--#51) vs. Cal (7-4, 4.907--#40)

Poinsettia
Colorado State (6-5, 1.426--#54) vs. Navy (7-4, -1.349--#74)

Fort Worth
Kansas (6-5, 1.593--#52) vs. Houston (6-5, 1.124--#56)

Hawaii
Nevada (8-3, 0.898--#61) vs. UCF (8-4, 4.470--#45)

Motor City
Memphis (6-5, -0.602--#69) vs. Akron (7-5, -2.316--#76)

Champs Sports
Clemson (7-4, 8.589--#28) vs. Colorado (7-5, 6.244--#33)

Insight
Arizona State (6-5, 4.717--#44) vs. Rutgers (7-4, 1.562--#53)

MPC Computers
Boise State (9-3, 11.196--#22) vs. Boston College (8-3, 12.939--#18)

Alamo
Michigan (7-4, 10.952--#24) vs. Nebraska (7-4, 6.048--#35)

Emerald
Georgia Tech (7-4, 8.411--#29) vs. Utah (6-5, -0.839--#72)

Holiday
Oregon (10-1, 18.728--#8) vs. Oklahoma (7-4, 13.289--#15)

Music City
Minnesota (7-4, 12.650--#20) vs. Virginia (6-5, 1.035--#58)

Sun
Northwestern (7-4, 5.755--#36) vs. UCLA (9-2, 11.024--#23)

Independence
South Carolina (7-4, 6.167--#34) vs. Missouri (6-5, 4.816--#41)

Peach
Miami (9-2, 18.472--#9) vs. LSU (10-2, 13.896--#14)

Meineke Car Car
South Florida (6-5, 7.493--#31) vs. NC State (6-5, 0.481--#63)

Liberty
Tulsa (8-4, 8.962--#27) vs. Fresno State (8-4, 9.190--#26)

Houston
TCU (10-1, 17.097--#11) vs. Iowa State (7-4, 7.347--#31)

Cotton
Texas Tech (9-2, 13.113--#16) vs. Alabama (9-2, 15.157--#12)

Outback
Iowa (7-4, 10.436--#25) vs. Florida (8-3, 12.134--#20)

Gator
Louisville (9-2, 14.560--#13) vs. Virginia Tech (10-2, 23.105--#3)

Capital One
Wisconsin (9-3, 11.231--#21) vs. Auburn (9-2, 13.049--#17)

Fiesta
Notre Dame (9-2, 17.809--#10) vs. Ohio State (9-2, 20.571--#5)

Sugar
West Virginia (10-1, 18.728--#7) vs. Georgia (10-2, 19.637--#6)

Orange
Penn State (10-1, 20.985--#4) vs. Florida State (8-4, 5.707--#37)

Rose
USC (12-0, 32.550--#1) vs. Texas (12-0, 31.043--#2)

[MrBigglesworth]

I love your rankings and I am a biased PSU fan, but do the rankings take margin of loss into account? Because VaTech got pasted by Miami.

[Wolfpack]

They do take margin into account. However, VT has a lot of large victory margins (GT, BC, etc) which does quite a bit to mitigate that 20-point loss to Miami. It's the nature of the beast with computer rankings. If you take into account margin, then it must carry equal weight regardless of who beat who. A 20-point win is a 20-point win regardless of whether it's Miami over VT or VT over BC. What can be controlled is how much that margin is made a part of the ranking. In my ranking, it's one of three parts I consider, the others being wins and losses and location (home or away). If you throw out scoring margin, SC, Texas, Penn State, and TCU are the top 5 and VT drops to #6.

[sooner333]

So are you saying that strength of schedule aside, that a 20 point win over Kent is as much as a 20 point win over Miami...then the part where Kent sucks and Miami is good comes into play in the SOS? Just wondering.

[Wolfpack]

Yeah, pretty much. It's a bit simple in terms of SOS (one degree of separation, so it's just who you play and not who you played and who they played). The other component that gets considered is whether the winner was at home or on the road (road wins get more weight).