Tip Top 25 in helmets, smaller
                                                    Home

Fixing the Final 2011 AP College Football Poll

1) Alabama 12-1
2) Louisiana State 13-1
3) Oklahoma State 12-1
4) Oregon 12-2
5) Arkansas 11-2
6) Southern Cal 10-2
7) Stanford 11-2
8) Boise State 12-1
9) South Carolina 11-2
10) Wisconsin 11-3
11) Michigan State 11-3
12) Michigan 11-2
13) Baylor 10-3
14) Texas Christian 11-2
15) Kansas State 10-3
16) Oklahoma 10-3
17) West Virginia 10-3
18) Houston 13-1
19) Georgia 10-4
20) Southern Miss 12-2
21) Virginia Tech 11-3
22) Clemson 10-4
23) Florida State 9-4
24) Nebraska 9-4
25) Cincinnati 10-3
Here is the final 2011 AP college football top 25. If you want to play along, you can access all of these teams' full schedules here: College Football Data Warehouse (amongst a number of other places). The fixed final AP top 25 follows the article below. 

But before I get to correcting the AP poll's errors, let me make one thing clear. This is not about what I would personally prefer to see in the rankings. I myself would probably rank Stanford ahead of Arkansas, Kansas State ahead of Baylor, and Clemson ahead of Southern Miss. But all those choices the AP poll made are fine-- logically valid choices. This is about fixing only those AP poll choices that are not logically valid or fair.

This year the AP poll voters did a very good job, and only had one small error in their entire top 16. As usual, however, the bottom 10 slots in the poll required a lot of fixing to make logical sense. The biggest problem for the AP poll was the ACC, which was the worst of the BCS conferences, yet nevertheless managed to place 3 teams in the top 25.

It is possible for the worst conference to place 3 teams in the top 25, if those teams separated themselves well from the rest of the conference, like Oregon, Stanford, and Southern Cal did (the PAC 12 was nearly as bad as the ACC this year). However, #22 Clemson and #23 Florida State were both beaten twice by ACC also-rans.

Before I get to all that, though, allow me a word or two on this year's so-called "national championship"...
2011 Iowa State-Oklahoma State football game

My painting of the game of the year-- not LSU-Alabama or LSU-Alabama Redux, but that fateful Friday night when Iowa State upset Oklahoma State in overtime, dealing the Cowboys their only loss and dooming all of us who do not live in Alabama to the worst "national championship" game in the BCS' short and too oft besmirched history. Today college football's power brokers are already talking about making substantial changes, and if this broken system gets fixed, we have Iowa State to thank.

2011 BCS Championship

You've probably read by now that Monday night's BCS championship game broke the record for least-watched BCS title game, receiving worse television ratings than previous lowpoints Nebraska-Miami 2002 and Oklahoma-LSU 2004, two games that, not coincidentally, also featured teams that should not have been there. Don't get me wrong-- Alabama was the best team of 2011, and I thought so even after LSU beat them the first time. So unlike Nebraska 2001 or Oklahoma 2003, the BCS did at least match the 2 best teams up against each other, as it was intended to do.

The problem is that pairing the 2 best teams in a matchup is not as important, in my opinion, as producing the fairest matchup, which this year would and should have been LSU-Oklahoma State. What we got instead was a relatively closed system, where only SEC teams were even allowed to compete for the so-called "national championship." The Big 12 actually finished with a better record against nonconference opponents than the SEC did this year, the first time anyone has topped the SEC in that regard since 2005, when the Big 12 beat them out with a terrific 31-5 record. This season the Big 12 did even better than that, finishing 33-5 (.868) to edge out the SEC's 47-8 (.855). But the Big 12 also played tougher schedules-- 61% of their nonconference opponents qualified for bowls, compared to 45% for the SEC. So the Big 12, top to bottom, was significantly better than the SEC, better than their straight records show. The Big 12 was 19-4 against bowl qualifiers (.826), the SEC 17-8 (.680).

Oklahoma State, of course, was the Big 12 champion, and at the end of the regular season, they had played and beaten more AP-rated teams and bowl-qualifiers than Alabama had. And that's just the illogical AP poll. Oklahoma State had beaten 6 teams that will make this year's fixed AP poll (as detailed below), Alabama just 2. That is why Oklahoma State deserved a shot at LSU. Alabama, on the other hand, had already gotten their shot against LSU, and at home. They lost. The best team doesn't always win, and Alabama's kicker failed them that night, much like Boise State's kicker has failed them the last 2 seasons. But unlike Boise State, Alabama got a mulligan.

So that LSU-Alabama "game of the century" back in November? Meaningless. Might as well not have played it. Turns out it was not the game of the century, year, or even month. LSU is the SEC champion, but all that got them was an extra game they had to win against Georgia, whereas Alabama could sit back and chill, getting a full 6 weeks to prepare for their make-up test to be named "national champions" over their own conference champions. The whole thing is a ridiculous fiasco to everyone outside Alabama.

If there had been a fuller playoff, with Oklahoma State getting their shot, Alabama winning the national championship would not be a problem. As it is, though, we have been cheated by the BCS yet again, and this year's "championship" leaves most college football fans feeling hollow and indifferent to Alabama's "title."

Stanford

As I expected would happen, Stanford was dropped back behind Southern Cal in the final poll because the Cardinal lost their bowl game, and as I said in the linked commentary, that response to Stanford's bowl result is "wholly illogical." The AP poll had Stanford's bowl opponent, Oklahoma State, ranked #3 going into that bowl game, and Stanford lost in overtime. More than that, Stanford outgained OSU by 178 yards, and would have won had their kicker not missed 3 field goals, including an easy one at the end of regulation. I sure would like to hear an AP poll voter explain how Stanford could possibly have played that game better short of winning. Given that OSU was ranked higher than both Stanford and USC, it makes absolutely no sense to drop the Cardinal behind USC, who did not even play a bowl game.

Stanford beat Southern Cal head-to-head, and though it was close (56-48 in triple overtime), the game was played at USC. USC did beat #4 Oregon (who in turn beat Stanford), but the Trojans also lost 43-22 to 6-7 Arizona State. Stanford's only losses came to the #3 and #4 teams, both ranked higher than USC. Stanford also performed better, as just 1 of their 11 wins was by a touchdown or less (31-28 over rival Cal), while USC had 4 close wins, 3 of them over unranked opponents and 2 over losing opponents.

The AP poll instructs their voters to pay attention to head-to-head results, so if those voters cannot follow instructions, perhaps they should not be voting. I think you could divide the poor raters from the competent ones simply by looking at who correctly rated Stanford higher than Southern Cal, and who thoughtlessly dropped Stanford behind USC.

Switch 'em: Stanford to #6 and Southern Cal to #7. But congrats to the AP poll voters on keeping #10 Wisconsin ranked ahead of #11 Michigan State-- I didn't think they would. Obviously the voters are improving.

West Virginia and Georgia

#17 West Virginia (10-3) is the most overrated team in this year's poll. Their bowl result was amazing, a 70-33 annihilation of #22 Clemson, but Clemson is overrated as well (along with the other ranked ACC teams), and the Orange Bowl was just 1 game. In their other games, West Virginia lost to 2 unranked teams (49-23 at 5-7 Syracuse and 38-35 at home against 7-6 Louisville), and they struggled to beat 3 other unranked teams 37-31, 21-20, and 30-27. They also beat #25 Cincinnati 24-21, but Cincinnati is another overrated team, and the close score doesn't actually do much toward earning their #17 rating.

A parade of teams will be moving past West Virginia in this fixed AP poll, starting with #19 Georgia (10-4). Georgia's 4 losses came to #1 LSU, #8 Boise State, #9 South Carolina, and #11 Michigan State, all ranked well-higher than West Virginia. No upset losses. West Virginia, again, lost to 2 unrated teams, one in a rout, so Georgia is effectively 2 games better than the Mountaineers on the season. The teams were similar performance-wise, but if either team has the edge here, it is Georgia. Georgia had 3 close wins (touchdown or less), West Virginia 4, and Georgia was routed in 2 of 4 losses, WV in 2 of 3-- and of course, one of WV's routs came to a losing team, whereas both of Georgia's came to top 10 teams, and one wasn't much of a rout (14 point loss to Boise State).

West Virginia's one huge performance, winning by 37 over #22 Clemson, does not differentiate them from Georgia either, since Georgia beat #27 Auburn by 38.

#19 Georgia also has a better relevant record than #18 Houston (13-1), who was stomped by a lower-rated team, #20 Southern Miss, December 3rd.

So we'll move Georgia past both teams, to #17, dropping West Virginia and Houston back 1 spot each.

Houston

West Virginia vs. Houston is a tougher case. Houston has the better relevant record, as they took 1 upset loss, while WV took 2, and moreover, Houston's loss came to a ranked team, whereas WV was beaten by a pair of unranked teams. But while the Big East was not a particularly good conference, it was perhaps a mite tougher than CUSA, so one could argue that WV faced the tougher schedule. Furthermore, Houston did not post as spectacular a result as WV's 70-33 Orange Bowl win over Clemson.

Houston also had as many weak performances (close wins) against unrated teams as West Virginia did. However, all 3 of Houston's weak performances came in September. After that, they routed every team but Southern Miss. West Virginia, on the other hand, beat a pair of losing teams by 1 and 3 points in their last 2 regular season games. And Southern Methodist's 28-6 spanking of Pitt in the BBVA Compass Bowl casts some doubt on the Big East being much better than CUSA, if at all. Houston beat SMU 37-7, and WV beat Pitt 21-20.

In the end, the biggest difference between these teams, by far, is those losses they took. There is just no excuse for WV's pair of upsets to unrated, and nowhere-near-rated, teams, and so no good reason for them to be rated higher than Houston, who took just 1 upset, and to the #20 team. WV's 70-33 bowl win over 10-4 Clemson looks more impressive than Houston's 30-14 bowl win over 9-4 Penn State, but those are just style points-- both teams won their bowls handily, and both beat 4-loss teams. Clemson finished ranked, and Penn State did not, but that choice is arbitrary. PSU's 4 losses all came to ranked teams, while Clemson was upset by a pair of unrated teams (Georgia Tech and North Carolina State), so PSU had the better relevant record.

Move Houston to #18, dropping West Virginia to #19.

Auburn

I've been talking about 8-5 Auburn for weeks, and though they did finally make a big jump in the last AP poll, they still finished unranked at #27 (they had not appeared on any voter's ballot in the previous poll). All 5 of their losses came to rated teams-- #1 Alabama, #2 LSU, #5 Arkansas, now-#17 Georgia, and #22 Clemson. They also defeated #9 South Carolina, so at the very least, one would think that they belong at #23, behind Clemson. However, they should actually be rated higher than that.

I criticized AP writers for not heeding head-to-head results in regards to Stanford-Southern Cal above, so why am I advocating that 8-5 Auburn should be ranked ahead of 10-4 Clemson, given that Clemson beat Auburn 38-24 and finished with a better record to boot? Simple-- Clemson has the better straight record, but they do not have the better relevant record, as Auburn played a massively tougher schedule. Clemson took 2 losses to unranked teams, Auburn none, so that's an effective 2-game lead for Auburn right there. Clemson's win over Auburn only makes up for one of them, leaving Auburn 1 game ahead. And Auburn's win over South Carolina, who stomped Clemson, puts them back to an effective 2 games better than Clemson.

Yes, Clemson beat Auburn, but Southern Miss beat Houston, and AP poll voters still correctly rated Houston higher, since USM took 2 upset losses, and Houston took no other loss, leaving Houston effectively 1 game better. Auburn's case may seem more confusing because they had 5 losses, but Clemson did not play a schedule nearly as tough. Take Alabama, LSU, and Arkansas off of Auburn's schedule, replace them with unrated opponents, and Auburn's record likely transforms to 11-2, and we would not even be discussing this, because they would already be rated far ahead of Clemson. That is the "magic" of schedule toughness.

Clemson did not themselves beat any teams as good as Alabama, LSU, Arkansas, or Georgia, so those 4 Auburn losses are irrelevant when comparing Auburn to Clemson. And while Auburn got stomped in all 4 of those losses, Clemson got stomped in the only 2 games they played against higher-rated opponents, 34-13 at South Carolina and 70-33 against West Virginia in the Orange Bowl. And again-- Auburn beat South Carolina, 16-13 on the road. Auburn beat 8-5 Virginia 43-24 in the Chick-fil-a Bowl, and Virginia is comparable to the 2 unrated teams that beat Clemson, 8-5 Georgia Tech and 8-5 North Carolina State.

So Auburn should be rated higher. How much higher? Well, they have a better relevant record than every team ahead of Clemson up to Georgia, who beat Auburn 45-7, so let's cut to the chase and compare them to Houston, who sits right behind Georgia. Houston and Auburn each took 1 upset loss, Houston 49-28 to #20 Southern Miss and Auburn 38-24 at #22 Clemson, so they're even there. Auburn's other 4 losses all came to teams currently ranked higher than Houston, so those losses are fairly irrelevant to this comparison. It is worth noting that Auburn got stomped in those losses, but it is hard to compare when Houston did not even play any such teams.

The big difference here is that Auburn beat #9 South Carolina, whereas Houston beat no ranked opponent at all. That's what makes Auburn effectively 1 game better than Houston. Also, Auburn's only close win came in their 42-38 opener over Utah State, whereas Houston had 3 such results in September (38-34 over UCLA, 35-34 at Louisiana Tech, and 49-42 at UTEP).

Move Auburn into the top 25, all the way up to #18, dropping Houston and everyone behind them 1 spot each. 10-3 Cincinnati falls out of the top 25, but they did not belong in the first place, with 2 losses to unrated teams, both in routs, and no wins over rated opponents. They also performed poorly, beating 4 unrated opponents by a touchdown or less, and all after mid-October, including an unimpressive 31-24 bowl win over 6-7 Vanderbilt. All of that says they were closer to a #35 team than a #25 team. As for Auburn, they once again further their case for being the most consistently underrated team in AP poll history.

Nebraska

9-4 Nebraska, now sitting at #25, has the same relevant record as Auburn, with 1 upset win and 1 upset loss. However, Nebraska's upset loss came to an unrated team (6-7 Northwestern), while Auburn's came to a rated team (now-#23 Clemson), and Nebraska posted 2 close wins over unrated teams to Auburn's 1. Nebraska also did not perform well in their bowl game, losing to South Carolina 30-13 (whom Auburn beat on the road 16-13).

So while Nebraska, like Auburn, was effectively 1 game better than now-#19 Houston, there is more reason to nevertheless rate Houston higher. I would rate Nebraska higher than Houston myself, but Houston was much more impressive in their bowl game, beating 9-4 Penn State 30-14 (Nebraska beat them 17-14), and Houston's upset loss at least came to a ranked opponent. Nebraska also performed poorly from November on, whereas Houston performed spectacularly well from October on, save for the one loss to Southern Miss. We can give this one to the AP poll voters.

West Virginia, on the other hand, has no business being rated higher than Nebraska. They did perform much better in their bowl game, but that is it. No other argument. And beating Clemson 70-33 is not actually more impressive than Nebraska beating #11 Michigan State 24-3 anyway. Nebraska's 1 upset loss to an unrated opponent is trumped by West Virginia's 2, and WV performed poorly to boot, posting twice as many close wins as Nebraska did.

Move Nebraska up to #20, dropping West Virginia and everyone behind them 1 spot each.

Missouri, Texas, and Texas A&M

As I've been saying since late November, 8-5 Missouri, 8-5 Texas, and 7-6 Texas A&M are all underrated. In the "Others Receiving Votes" section, Missouri is #29 and Texas #30, yet Texas defeated #26 Brigham Young head-to-head. Obviously that ranking is wrong, but that is the least of it. As a group, these 3 teams are effectively 2 games better than now-#21 West Virginia.

Let me reiterate that the Big 12 was this season's best conference, going 33-5 against nonconference opponents and 6-2 in bowl games-- and 61% of those opponents qualified for bowl games. That 33-5 record is the best posted by any conference since 1997, when the SEC went 37-5, and it is the 2nd best record by any conference in the last 50 years. Missouri, Texas, and Texas A&M may have taken 16 losses in total, but 15 of those losses came to ranked teams or to each other, the one exception being Missouri's overtime loss at 6-7 Arizona State. West Virginia, however, took 2 losses to unrated opponents, including a 49-23 embarrassment  at 5-7 Syracuse. Missouri and Texas had half as many close wins as WV did, and Texas A&M had just 1 close win.

Look closely at 7-6 Texas A&M's record. Their 6 losses came to #3 Oklahoma State, #5 Arkansas, #15 Kansas State, #16 Oklahoma, Missouri, and Texas. The first 4 are all rated higher than West Virginia, and the other 2 should be, as both have better relevant records and performed better as well. West Virginia's 1 big win, 70-33 over now-#24 Clemson, is simply not as impressive as Texas A&M's 55-28 win over #13 Baylor, who is of course ranked higher than WV. Texas A&M lost to the #3 team by just 1 point and the #5 team by 4 points-- both performances are that of a top 10 team. They also lost to #15 KSU in overtime, to Missouri in overtime, and to Texas by 2 points. They took no upset losses, while WV took 2, and they had just 1 close win to WV's 4.

Move all 3 teams in ahead of West Virginia: Missouri to #21, Texas to #22, and Texas A&M to #23. That drops West Virginia to #24, Southern Miss to #25, and 3 ACC teams fall out of the top 25-- 11-3 Virginia Tech, 10-4 Clemson, and 9-4 Florida State.

Clemson, Virginia Tech, and Florida State

I know that the AP poll likes to represent every BCS conference in the top 25, and I do my best to work with them toward that goal, but sometimes a conference just does not merit having a top 25 team. The ACC went 32-16 against nonconference teams, which edges out the PAC 12's 23-12, but the PAC 12 played tougher opponents, so the ACC was actually the weakest of the BCS conferences this season. The ACC was 7-13 against bowl qualifiers (.350) and 25-3 against losers (.893), while the PAC 12 was 6-10 against bowl qualifiers (.375) and 17-2 against losers (.895), edging out the ACC in both categories. But let's look at the particular 3 ACC teams in question.

First of all, ACC champion Clemson was originally rated #22, behind #21 Virginia Tech, and it need hardly be said that this rating was wrong, as proven by Clemson's 2 wins over Virginia Tech, both very decisive-- 22-3 and 38-10. Terrible rating right there. Clemson, of course, was thoroughly embarrassed by now-#24 West Virginia 70-33 in the Orange Bowl. They lost to 2 unrated teams, 31-17 at Georgia Tech and 37-13 at North Carolina State. They beat Wofford by 8. None of that says that they are a top 25 caliber team, but they did beat now-#18 Auburn 38-24, 9-4 Florida State 35-30, and they posted those 2 routs of 11-3 Virginia Tech, so I would rate them #25, in place of Southern Miss, myself. However, the AP poll rated USM higher to begin with, and that is a viable option, as Clemson performed dismally over their last 6 games, so USM keeps their #25 spot and Clemson remains at #26.

Virginia Tech parlayed their easy schedule into a nice-looking 11-3 record, but they were stomped by Clemson twice, so it makes no sense to see those results as accidents, and they only played 1 currently rated opponent. That was a positive result, an overtime loss to #12 Michigan in the Sugar Bowl, but Tech also beat unrated opponents by 7, 3, 4, and 3 points, along with taking those routs at the hands of Clemson, and 6 bad results outweigh 1 good one.

Florida State, like Clemson, took 2 losses to unrated opponents (35-30 at Wake Forest and 14-13 to Virginia at home), and though they performed a bit better than Clemson, they lost to Clemson, and like Virginia Tech they did not beat any rated teams. Even their 18-14 bowl win over 7-6 Notre Dame does them few favors, given that Dame did not even get a single poll point.

Fixed AP Top 25

11-3 Virginia Tech, 10-4 Clemson, 9-4 Florida State, and 10-3 Cincinnati fall out of this fixed AP poll top 25, replaced by 8-5 Auburn, 8-5 Missouri, 8-5 Texas, and 7-6 Texas A&M. Your typical AP poll voter would likely look at this and say, "Are you nuts? The teams you're tossing out were 40-14, and the teams you're putting in were only 31-21!" However, what matters is this: discounting their games against each other, the 4 teams falling out had 6 losses to unrated opponents and 1 win over a rated team, while the 4 teams that replace them had 2 losses to unrated opponents, one of those being #26 Clemson, and they had 2 wins over rated teams. That is a rout for the replacement teams. Their straight records only look worse because they faced far, far tougher schedules. Does any AP poll voter really believe that he would be rating Cincinnati right now if the Bearcats had played the schedule faced by Auburn, Missouri, Texas, or Texas A&M? I hope not.

I do think Clemson should be #25 in place of Southern Miss, but the AP poll made that bed when they rated USM higher. I shake my head at it, but it's a logically viable choice.

Anyway, here it is, now logically coherent, your fixed final AP poll for 2011.


1) Alabama 12-1 --
2) Louisiana State 13-1 --
3) Oklahoma State 12-1 --
4) Oregon 12-2 --
5) Arkansas 11-2 --
6) Stanford 11-2 +1
7) Southern Cal 10-2 -1
8) Boise State 12-1 --
9) South Carolina 11-2 --
10) Wisconsin 11-3 --
11) Michigan State 11-3 --
12) Michigan 11-2 --
13) Baylor 10-3 --
14) Texas Christian 11-2 --
15) Kansas State 10-3 --
16) Oklahoma 10-3 --
17) Georgia 10-4 +2
18) Auburn 8-5 IN
19) Houston 13-1 -1
20) Nebraska 9-4 +4
21) Missouri 8-5 IN
22) Texas 8-5 IN
23) Texas A&M 7-6 IN
24) West Virginia 10-3 -7
25) Southern Miss 12-2 -5

OUT: #21 Virginia Tech 11-3
#22 Clemson 10-4
#23 Florida State 9-4
#25 Cincinnati 10-3

Home