Home

Fixing the Final 2023 AP College Football Poll

1) Michigan 15-0
2) Washington 14-1
3) Texas 12-2
4) Georgia 13-1
5) Alabama 12-2
6) Oregon 12-2
Florida State 13-1
8) Missouri 11-2
9) Mississippi 11-2
10) Ohio State 11-2
11) Arizona 10-3
12) Louisiana State 10-3
13) Penn State 10-3
14) Notre Dame 10-3
15) Oklahoma 10-3
16) Oklahoma State 10-4
17) Tennessee 9-4
18) Kansas State 9-4
19) Louisville 10-4
20) Clemson 9-4
21) North Carolina State 9-4
22) Southern Methodist 11-3
23) Kansas 9-4
24) Iowa 10-4
25) Liberty 13-1

Here is the final AP poll college football top 25 for the 2023 season. The fixed final AP poll 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 be inclined to rank 11-2 Mississippi and 10-3 LSU ahead of 11-2 Missouri (more on this issue below), and I would put 10-3 Notre Dame ahead of 10-3 Arizona (also covered below), just as a couple of examples. But in these cases, the choices the AP poll made are logically viable. This is about fixing only those AP poll choices that are not logically valid or fair.

This has been an incredibly difficult AP poll for me to repair this year, and for that reason I am a few days later than normal getting this article out. Several issues took me hours of deliberation to make a decision on, and the bottom 2 slots of the top 25 required more than a full day for me to finally settle on the best two teams to fill them.

But enough moanin' and groanin'. Let's to get to work...

Michigan running back Blake Corum in the national championship game for the 2023 season

My ms-painting of Michigan running back Blake Corum, who tallied 134 of Michigan's 303 rushing yards in their 34-13 victory over Washington in the national championship game for the 2023 season.

Alabama > Georgia

Our first fix is an easy one that took no time at all to figure out. Here's what I wrote on this issue in an article I posted a week before the final AP poll was published: "Sometimes AP poll voters drop playoff losers behind teams that win lower-level bowl games, so I would not be surprised to see 12-2 Texas and 12-2 Alabama fall behind 13-1 Georgia in the final poll. Needless to say, such a move would be wrong. Texas defeated Alabama, who defeated Georgia in the SEC title game, and that is the order in which these teams should be ranked. Texas and Alabama both lost razor-close games to the top 2 teams in the country, and Georgia's 63-3 bowl "win" was a joke that came against a team that had 29 players opting out. Hopefully the sports writers show some sense here."

Unfortunately, the majority of sports writers did not show some sense here. Reality check: Alabama (#5) defeated Georgia (#4) 27-24 in the SEC title game, thereby winning the SEC and a slot in the playoffs, and there is nothing more to be said here. Anyone who rated Georgia higher than Alabama should not be allowed to vote anymore.

Switch 'em. Alabama moves to #4 and Georgia drops to #5.

Louisiana State > Arizona

This issue is even more of a slam dunk, and another one that I mentioned last week. #11 Arizona and #12 Louisiana State are both 10-3, which is sadly all that most AP voters look at. But a closer look reveals that Arizona played 1 top 10 opponent while LSU played 4, a rather gargantuan disparity. Arizona took 2 "upset" losses to lower-ranked teams, LSU none, and unlike Arizona, LSU defeated a higher-ranked team (#8 Missouri), so LSU's relevant record was effectively 3 games better than that of Arizona for the season. LSU should obviously be ranked higher, and it is not at all close.

LSU could be ranked even higher, and in fact I would rate them ahead of #8 Missouri (11-2), whom they defeated 49-39 on the road. I'll discuss that issue next. And Arizona could be ranked even lower. I would rate them behind 10-3 Notre Dame (discussed later), and they could even be logically placed behind 8-5 Southern Cal, who beat them (also discussed later in this article).

But for now, we'll just switch 'em. Louisiana State moves to #11 and Arizona drops to #12.

Missouri, Ohio State, Mississippi, and Louisiana State

11-2 Mississippi (#9) defeated 10-3 Louisiana State (now #11) 55-49, LSU defeated 11-2 Missouri (#8) 49-39 on the road, and Missouri defeated 11-2 Ohio State (#10) 14-3 in the Cotton Bowl, and I would definitely rate these teams in that order. However, as I wrote on this issue last week, "I think there is a case for Missouri to be ranked higher than Mississippi and LSU, if 11-2 Ohio State (whom Missouri defeated in the Cotton Bowl) is also ranked ahead of both Mississippi and LSU."

As they are ranked now, it doesn't make sense to rate 11-2 Ohio State ahead of 10-3 Louisiana State. That's because all 3 of LSU's losses came to teams ranked higher than Ohio State, and LSU defeated Missouri 49-39 on the road, while Ohio State lost to Missouri 14-3 in the Cotton Bowl. Those results give LSU a relevant record that is effectively 1 game better than that of Ohio State for the season. However, if Ohio State is ranked ahead of Mississippi, then LSU does have a loss to a team ranked behind Ohio State.

And Ohio State can be viably ranked ahead of Mississippi, because the two teams have the same relevant record, and Ohio State clearly performed better than Ole Miss did for the season. Mississippi lost by more than a touchdown twice, Ohio State once, and Ohio State lost to #1 Michigan by only 6 points on the road. And Mississippi posted 3 close wins over unranked teams (4-8 Arkansas, 6-7 Auburn, and 7-6 Texas A&M), while no unranked team came within a touchdown of Ohio State. And if you want to try and claim that mid-level SEC teams are just far tougher opponents than mid-level Big 10 teams, the facts beg to differ. Ohio State beat Maryland by 20, and Maryland beat Auburn by 18 in their bowl game. Ohio State beat Wisconsin by 2 touchdowns on the road, and LSU only beat Wisconsin by 4 in their bowl game. And Ohio State beat Rutgers by 19 on the road, Rutgers beat Miami (Florida) by 7 in their bowl game, and Miami beat Texas A&M by 15 points.

So to sum up, these teams can be ranked Mississippi > LSU > Missouri > Ohio State, which I think is best, or they can be ranked Missouri > Ohio State > Mississippi > LSU. And since what the AP poll has now is far closer to the latter, we'll go with that. All we have to do is move Ohio State ahead of Mississippi, which isn't much of a move, given that Mississippi was only 24 poll points ahead of the Buckeyes anyway. The idea here is that Missouri's bowl win over Ohio State is so significant that it overrides their "upset" loss to LSU. And that idea doesn't logically work if OSU is ranked behind Ole Miss.

I have one last comment on this issue. While my preference, again, would be to rank these teams according to their simple and clear victory chain, Mississippi > LSU > Missouri > Ohio State, one merit to the AP poll's preference is the fact that Missouri is the only rated opponent LSU defeated, and so easier to dismiss as a fluke (Missouri, Mississippi, and Ohio State all defeated at least 2 rated opponents). Also, while LSU beat Missouri by 10 points, the game was actually much closer than that score would indicate, as LSU returned a desperation pass for a touchdown with 34 seconds left. And again, Ohio State clearly outperformed Mississippi on the season, so this fix we're settling on here has a couple of merits going for it.

We have another simple switch here, Ohio State up to #9 and Mississippi down to #10.

Penn State > Arizona

This is yet another issue I addressed last week. Arizona (now #12) and Penn State (#13) may both be 10-3, but similar to the LSU vs. Arizona case discussed above, these teams do not have the same relevant record. Arizona took 2 "upset" losses to lower-ranked teams, Penn State none (all 3 of PSU's losses came to top 10 teams), so Penn State was effectively 2 games better than Arizona for the season. Arizona's bowl win over Oklahoma is far better than any Penn State win, but that doesn't come close to making up for 2 upset losses. And Penn State clearly performed better than Arizona did too, as all of their wins came by more than a touchdown, while Arizona posted 2 close wins (by 1 and 3 points) over losing teams (3-9 Stanford and 4-8 Colorado).

So Arizona should be dropped back behind Penn State. As I alluded to earlier, Arizona could be dropped even further, all the way back behind 8-5 Southern Cal, who beat them this season. Southern Cal will be ranked in the fixed poll (as covered below) thanks to their 42-28 bowl win over 10-4 Louisville. USC has 2 upset losses (to Utah and UCLA), Arizona 1 (Mississippi State), so USC's win over Arizona makes them even, with USC holding the head-to-head tiebreaker. However, USC's win came at home, in triple overtime, so it was not at all decisive, and Arizona finished the season vastly stronger than USC did, winning their last 7, while USC was losing 5 of their last 7. So although I think that you could logically rank Arizona behind USC, I think the AP poll is correct not to do so.

But, again, Penn State clearly belongs ahead of Arizona, so we have a 4th simple switch in a row here. Move Penn State up to #12 and drop Arizona to #13.

Notre Dame and Oklahoma

Here is another issue I have addressed before. In that article I wrote, "10-3 Oklahoma and 10-3 Notre Dame have each taken two upset losses to lower-ranked teams, but Oklahoma made up for one of theirs with a huge upset win over 12-2 Texas, who will finish ranked in the top 5. That win gives Oklahoma a relevant record that is effectively one game better than that of Notre Dame, and Oklahoma should therefore be rated higher than Notre Dame."

However, as I look closer at these teams, I think there is perhaps some case for Notre Dame (#14) to be ranked higher than Oklahoma (#15), as Notre Dame performed better, and they have a much deeper collection of impressive wins (Texas is the only rated team Oklahoma defeated, while Notre Dame defeated 3 teams that will finish ranked in the fixed AP poll, and a 4th that will finish just outside the top 25).

The problem is, at the moment, 10-3 Notre Dame is ranked right behind 10-3 Arizona, and that just doesn't work. I would definitely rate Notre Dame ahead of Arizona myself. Both teams took 2 upset losses to lower-rated teams, but both of the lower-rated teams Notre Dame lost to are ranked in the top 25, while Arizona took a loss to 5-7 Mississippi State. And Notre Dame clearly outperformed Arizona, a fact made all the more clear by the fact that Notre Dame did massively better against 3 common opponents. Arizona edged Stanford 21-20, while Notre Dame crushed Stanford 56-23; Arizona edged Oregon State 27-24 at home, Notre Dame stomped OSU 40-8 in the Sun Bowl; Arizona lost to Southern Cal 43-41 in triple overtime, and Notre Dame beat USC 48-21.

So there are 2 logically viable ways to rate these 2 teams, Arizona > Oklahoma > Notre Dame, or Notre Dame > Arizona > Oklahoma. I think the latter is much better, and that, again, would be my choice, but this is about what AP poll voters prefer (within logical reason). Arizona had 861 poll points, Notre Dame 745, and Oklahoma 691. Since Notre Dame is much closer to Oklahoma in the AP poll, the viable option that is closest to what the AP poll has is Arizona > Oklahoma > Notre Dame. So that is what we will go with here. Again, Oklahoma can be fairly ranked ahead of Notre Dame due to their huge win over #3 Texas (which gives them a better relevant record than that of Notre Dame), and that being the case, Arizona's 38-24 win over Oklahoma in the Alamo Bowl is rendered all the more significant, and the thing that fairly buoys them above Notre Dame as well.

We have a 5th simple switch in a row! Move Oklahoma up to #14 and drop Notre Dame to #15.

Kansas State > Tennessee

9-4 Tennessee (#17) and 9-4 Kansas State (#18) each took one upset loss to an unranked team, so the teams have the same relevant record as well as straight record. But Kansas State clearly accomplished more and performed better than Tennessee did this season, and I can't see a good reason for Tennessee to be rated higher.

The one thing in Tennessee's favor is the fact that their upset loss came in game 3 (to 5-7 Florida), and Kansas State's came in game 12 (7-6 Iowa State), giving KSU's upset a bit more emphasis. But this one factor is greatly outweighed by Kansas State's advantages. Tennessee did not defeat a team that will finish ranked in the fixed AP poll (10-4 Iowa, whom they defeated in the Citrus Bowl, is worthless and should not be ranked, as covered later in this article). Kansas State defeated 2 ranked teams (9-4 North Carolina State and 9-4 Kansas). All 4 of Tennessee's losses came by more than a touchdown, including a 29-16 loss to 5-7 Florida. All 4 of KSU's losses came by a touchdown or less, including an overtime loss at #3 Texas. Tennessee posted 2 close wins (touchdown or less) over unranked opponents, while KSU beat all of their unranked opponents by more than a touchdown. And for the cherry on top, Tennessee lost at #8 Missouri by 29 points, and KSU lost at Missouri by 3 points.

This is an easy one, and gives us our 6th simple switch in a row. Move Kansas State up to #17 and drop Tennessee to #18.

Southern Cal > Louisville > North Carolina State > Clemson

I've been saying that 10-4 Louisville (#19) is overrated since they lost to Kentucky in November, and due to their 42-28 loss to 8-5 Southern Cal (#30 in the Others Receiving Votes section) in the Holiday Bowl, I wrote last week that they should be rated behind USC. Moreover, last week I said that since 8-5 Oregon State defeated 8-5 Utah this year, and Utah beat Southern Cal, these teams should be rated as such: Oregon State > Utah > USC > Louisville. However, after looking at the situation more closely, I think it is acceptable to rate USC higher than Utah despite the head-to-head result. That is because Utah took their big upset loss (to 8-5 Northwestern) in their bowl game, giving it more weight. As such, you could fairly see USC as having the better relevant record by the margin of that extra significance. Since the AP poll did rate USC higher than Utah (USC #30, Utah #32 in Others Receiving Votes section of the AP poll), we'll just go ahead and accept that.

Southern Cal vs. Louisville

So let's compare 8-5 Southern Cal (#30) to 10-4 Louisville (#19). Southern Cal has 2 upset losses (to 8-5 Utah and 8-5 UCLA), and Louisville also has 2 upset losses (3-9 Pittsburgh and 7-6 Kentucky). USC made up for one of their upset losses with an upset win over 10-3 Arizona (now #13), and Louisville made up for one of theirs with an upset win over 10-3 Notre Dame (#15). So the only real difference between these teams is the fact that USC defeated Louisville 42-28 in the Holiday Bowl. The only reason USC has 1 more loss than Louisville is because they played 2 top 10 teams (Oregon and Washington), and Louisville only played 1 (Florida State). USC's head-to-head win was dominating and came in a bowl game, giving it extra weight, and there is simply no legitimate doubt that USC should be rated ahead of Louisville.

Since 10-4 Louisville won 13-10 at 9-4 North Carolina State, and NC State won 24-17 over 9-4 Clemson, we have the following victory chain: Southern Cal > Louisville > NC State > Clemson. The AP poll did place Louisville ahead of NC State and Clemson, but unfortunately the poll put Clemson at #20 and NC State at #21. Since Clemson did have a fantastic finish, and was certainly a better team over their last 5 games than they had been early, and since Clemson did perform better than NC State did this season, I spent a lot of time looking very closely at this case.

North Carolina State vs. Clemson

Clemson and NC State each took an upset loss to unranked Duke, but Clemson's came in their opener, while NC State's came in game 7. Clemson took both of their upset losses in their first 7 games, but as stated, they were very impressive in their last 5 games, including an upset win over 10-3 Notre Dame (now rated #15). If you place a little more emphasis on the late season and a little less on the early games, Clemson looks a little better. Clemson also performed better than NC State did, posting 2 close wins (touchdown or less) over unrated opponents, and losing just 1 game by more than a touchdown. NC State had 3 close wins over unrated teams, and they lost 3 games by more than a touchdown, so they posted twice as many weak performances as Clemson did. Finally, NC State's head-to-head win over Clemson was a close home win, 24-17. Clemson outgained them 364 yards to 202. So it was not at all a decisive outcome.

Due to all of that, I think there is some case for rating Clemson higher than North Carolina State. However, doing so would open a Pandora's box of other changes that would end up having a huge impact on the top 25. Louisville did not play Clemson, and as shown above, Louisville must be rated below Southern Cal, so if NC State is rated below Clemson, then Louisville must be as well. And if Louisville is dropped below Clemson, then 7-6 Kentucky, who beat 10-4 Louisville 38-31 on the road in their regular season finale, should also be rated ahead of Louisville.

I had been calling for Kentucky to be rated ahead of Louisville since Kentucky beat them. However, after Clemson defeated Kentucky in their bowl game, it no longer became necessary to drop Louisville behind Kentucky, as long as Louisville and NC State remained ranked ahead of Clemson. You can rate these teams Louisville > NC State > Clemson > Kentucky, or you can rate them Clemson > Kentucky > Louisville > NC State. 7-6 Kentucky is currently ranked in a tie for #37-39 in the Others Receiving Votes section of the AP poll, so moving them up ahead of Louisville and NC State into the top 25 would represent a massive change.

Our Two Options

So we have two options. Option A is to go with the AP poll and keep NC State ranked behind Clemson. If we do that, the bottom of our top 25 looks something like this: Clemson, USC, Kentucky, Louisville, NC State. In addition, Kansas, who is currently rated #23, would move up. They would likely move past NC State, and possibly rise all the way past Clemson to #18 (In this scenario, Clemson would have 3 upset losses instead of 2). Option B is to heed NC State's head-to-head win and rank them above Clemson. If we do that, the bottom of our top 25 looks something like this: USC, Louisville, NC State, Clemson. Since option B is far closer to what the AP poll has now, we will go with that one. All it requires is moving #30 USC up above #19 Louisville, then having Clemson and NC State switch places. Kentucky will remain unranked, and, as covered later in this article, I think that Kansas can remain ranked behind Clemson (Clemson only has 2 upset losses in this scenario).

Where to Rank Them

So we're going with 8-5 Southern Cal (#30) > 10-4 Louisville (#19) > 9-4 North Carolina State (#21) > 9-4 Clemson (#20). Where do we put them? Well, the average rank of these 4 teams is 22.5, which would place them right behind #22 Southern Methodist. However, SMU is about to be tossed out of the top 25 entirely (covered next), so we can just keep these teams where the 3 ACC teams already are, and just move USC up above Louisville, and drop Clemson behind NC State.

That gives us the following ratings shift: Southern Cal #19, Louisville #20, North Carolina State #21, and Clemson #22. This fix ended up being relatively simple, but needless to say, it took me a very long time to settle on it.

Southern Methodist

As I've previously stated (just above), and as I've been saying for weeks, 11-3 Southern Methodist (#22) should not be rated, and in fact they should not even be close to rated.

I had hoped that SMU's 23-14 bowl loss to 7-6 Boston College would put some sense into the poll voters, but no such luck. SMU should be rated behind BC, which is to say, well below the top 25. In fact, as I've been saying since November, SMU should be rated behind 5-7 Texas Christian, who beat them 34-17! SMU went 0-3 against major conference teams, 2 of them unrated, so the fact that they went 11-0 against a bunch of unrated minor conference teams is simply not impressive or relevant. The best teams they beat are 11-3 Tulane, who lost 41-20 to 7-6 Virginia Tech in their bowl game (notice a trend here?), and 10-3 Memphis, who lost to Tulane.

Toss 'em out.

Kansas

9-4 Kansas (#23) took one upset loss (to 7-6 Texas Tech), but they made up for it with a big upset win over 10-3 Oklahoma (now #14). That gives Kansas a better relevant record than that of all 5 teams currently rated in front of them: 9-4 Clemson (2 upset losses and 1 upset win), 9-4 North Carolina State (1 upset loss), 10-4 Louisville (2 upset losses and 1 upset win), 8-5 Southern Cal (2 upset losses and 1 upset win), and 9-4 Tennessee (1 upset loss). I myself would rate Kansas ahead of all these teams, putting them right behind 9-4 Kansas State (now #17), who only beat them by 4 points.

However, I think a case can be made for keeping Kansas where they are, right behind Clemson. Kansas' only win of any note was that upset win over Oklahoma, making it easier to dismiss it as a fluke. Clemson defeated at least 3 good teams (10-3 Notre Dame, 8-5 North Carolina, and 7-6 Kentucky). Also, as discussed previously, Clemson was far stronger down the stretch, winning their last 5 games, and their 2 upset losses came in their first 7 games, one of them in their opener, giving it less weight. This is a pretty close case, but I think that Clemson being rated ahead of Kansas is acceptable.

Note that, as I previously indicated, Kansas compares far better against North Carolina State, so if NC State were rated behind Clemson, as the AP poll originally wanted, then Kansas would be moving up, and in that case, Clemson would have 3 upset losses instead of 2, and I think that would be the end of their case for being rated higher than Kansas.

No change.

Iowa, Liberty, and Oregon State

10-4 Iowa currently sits at #24, 13-1 Liberty at #25, and 8-5 Oregon State just outside the top 25 at #26. Iowa has an upset loss and Liberty does not, so Liberty has a better relevant record than Iowa does (as well as a better straight record, of course). Still, at first glance I thought that rating Iowa higher was fine, since they play in a major conference, and beat a decent opponent, while Liberty beat no one of any value whatsoever. Liberty's only game against a major conference team was their embarrassing 45-6 loss to Oregon in the Fiesta Bowl. And then I thought that rating Liberty ahead of Oregon State was fine, since Oregon State had an upset loss, and Liberty therefore had a better relevant record.

But can you spot the flaw in logic here? It doesn't make consistent sense to say that being in a major conference trumps relevant record in the case of Iowa vs. Liberty, but then say the opposite in the case of Oregon State vs. Liberty. Either relevant record matters more, and Liberty should be rated higher than both teams (each of whom took an upset loss to an unranked team), or being in a major conference and beating a nearly-rated opponent matters more, and Liberty should be rated lower than both teams. I think you can make a case for either option. Iowa had 150 poll points, Liberty had 139, and Oregon State had 96. Since Liberty was closer to Iowa than they were to Oregon State, we will go with the option of rating Liberty ahead of both teams.

Of course, either way, Iowa had no chance of finishing ranked. They are the worst performing team in the top 25. Their 3 losses to rated teams came by scores of 31-0, 26-0, and 35-0. The 35-0 loss was particularly pathetic, coming to #18 Tennessee in their bowl game. They also posted 5 close wins (touchdown or less) over unranked teams, including wins by 2 and 3 points over losing teams in their last 2 regular season games. In fact, when you look closely at how consistently poorly Iowa played, it seems pretty silly to consider ranking them higher than Liberty just because they play in a major conference. The fact is, we just don't know how Liberty would fare against a full schedule of mediocre and bad teams from major conferences. But we do know that Iowa lost to 6-7 Minnesota, who only beat Bowling Green by 6 points in their bowl game, and Liberty beat Bowling Green by 10 points.

But the real point here is that 8-5 Oregon State outperformed 10-4 Iowa by a country mile, and they should definitely be rated higher than Iowa. Both teams took an upset loss (OSU to 5-7 Washington State, and Iowa, again, to 6-7 Minnesota). Oregon State has 1 more loss than Iowa does because they played 4 rated opponents and Iowa played 3. And while Iowa was shut out and stomped by all 3 of their rated opponents, OSU lost to now-#13 Arizona by 3 points and to #2 Washington by 2 points. And only 1 of Oregon State's wins came by a touchdown or less (compared to 5 for Iowa). So the performance advantage here is overwhelming. Oregon State also defeated 2 teams (8-5 Utah and 8-5 UCLA) in the "Others Receiving Votes" section of the AP poll, Iowa just 1 (8-5 Northwestern).

Iowa has been overrated all year. Time to show 'em the door. Liberty moves up to #24 and Oregon State to #25. But not for long...

Duke

I've been writing articles commenting on the AP poll rankings since November, and at no point did Duke (8-5) come up as an option for a top 25 slot, even in the last article I posted before the final AP poll was published. I didn't even think about Duke as an option until I started working on fixing the final AP poll, and saw that 8-5 Duke was ranked #36 in the Others Receiving Votes section of the AP poll.

So 8-5 Duke meriting a slot in the fixed AP poll top 25 took me by surprise. But they have a better relevant record than do all the teams ranked ahead of them up to #24 Liberty, and that put them on my radar.

Let's compare 8-5 Duke to our current #25, 8-5 Oregon State. Duke took 2 upset losses to unranked teams (8-5 North Carolina and 3-9 Virginia), but they made up for both with a pair of dominating upset wins over ranked teams, 28-7 over now-#22 Clemson and 24-3 over #21 North Carolina State. Oregon State took one upset loss to an unranked team (5-7 Washington State), but they did not defeat any ranked teams to make up for it. So Duke was effectively one game better than Oregon State for the season as a whole. Both teams may be 8-5, but Duke played 5 rated teams and Oregon State played 4, so they are not equal.

The only potential argument for Oregon State is the fact that Duke was far stronger in their first 6 games, when they got both of their big wins over ranked opponents, than they were over their last 7 games, when they took both of their losses to unranked teams. They lost to 3-9 Virginia in game 11, and that is a very bad loss, worse than any by Oregon State. If earlier games get less emphasis and later games get more emphasis, then Duke's case gets weaker. However, I look at these timing issues as tiebreakers, and remember that Duke and Oregon State are not really even. If they had the same relevant record, then I think these timing issues would matter, but Duke is effectively 1 game better than Oregon State for the season, and therefore I just don't think these timing issues are enough to let the AP poll keep Oregon State rated higher than Duke.

Furthermore, Duke's finish was not entirely bad, as they looked strong over their last 2 games, capped by a 17-10 bowl win over 11-3 Troy, a team that is still ranked in the Others Receiving Votes section of the AP poll (higher than Duke, which, needless to say, is idiotic). And Oregon State's last 2 games were very poor showings, capped by an embarrassing 40-8 loss to now-#15 Notre Dame in the Sun Bowl. Duke, by the way, lost to Notre Dame by just 7 points.

So I just don't see any good reason for Oregon State to be rated higher than Duke. As such, we'll go ahead and move Duke ahead of Oregon State. But we could go further than that...

Duke vs. Clemson

8-5 Duke defeated 9-4 Clemson (now #22) 28-7, and it would actually be logically viable to rate Duke higher than Clemson. Duke has 2 upset losses and Clemson has 1. Duke's win over Clemson makes them even, with Duke having the head-to-head tiebreaker. Each defeated 1 higher-rated team (Duke beat North Carolina State, who beat Clemson, and Clemson beat Notre Dame, who beat Duke). But these are 2 teams that went in opposite directions this season. As covered above, Duke was very strong in their first 6 games, when they collected both of their wins against rated teams (their win against Clemson came in their opener), and they were very poor in their final 7 games, when they took both of their losses to unranked teams (losing to 3-9 Virginia in week 11). Clemson, on the other hand, was at their worst in their first 7 games, when they lost to Duke and to unranked Miami (Florida), and they were very strong in their final 5 games, when they beat now-#15 Notre Dame and 3 other bowl qualifying teams.

Above, I said that I consider these timing issues to be tiebreakers, and so I didn't think that they made up for Duke being effectively 1 game better than Oregon State for the season. But in this case, Duke and Clemson have the same relevant record (Duke may have 5 losses and Clemson 4, but Duke played 5 ranked teams and Clemson played 3 plus Duke), and so in this case I do think that the timing issues are a legitimate reason to rank Clemson higher. Also, Oregon State didn't have a strong finish themselves. In this case, the timing issues are much more compelling because on top of Duke being so strong early and so poor late, Clemson was very weak early and very strong late. And to cap it all off, the game most at issue here, Duke's win over Clemson, came in the opener, giving it the least emphasis possible.

As I say, you could viably rate Duke ahead of Clemson. But I think it is also fine to rate Clemson higher than Duke, as the AP poll did.

Duke vs. Liberty

So we've moved 8-5 Duke ahead of #25 Oregon State. How about #24 Liberty (13-1)? Liberty, again, beat absolutely no one of any real value. We let the AP poll keep Liberty ranked higher than Oregon State because OSU took an upset loss and Liberty did not, leaving Liberty effectively 1 game better than Oregon State for the season. However, Duke made up for their 2 upset losses with 2 upset wins over teams ranked higher than Liberty, so Duke and Liberty have the same relevant record.

This was a very tough issue for me, because Duke accomplished so much more than Liberty did. In addition to their 2 upset wins over ranked opponents, Duke also beat 8-5 Northwestern and 11-3 Troy, who are both in the Others Receiving Votes section of the AP poll. I spent a lot of time looking at this case, but in the end I decided that the timing issues discussed above are enough to let the AP poll keep Liberty ranked higher than Duke. Again, because Duke and Liberty are effectively even, the fact that Duke was terrific the first half of the season and terrible the last half matters more, and I think it is enough to make the difference.

Particularly persuasive is Duke's loss to 3-9 Virginia in week 11. Virginia lost to James Madison, who took a pair of "upset" losses to minor conference teams, and as weak as Liberty's schedule was, we can safely say that they were better than that.

Duke sticks at #25 and Liberty remains just ahead of them at #24.

Duke vs. North Carolina

Some of you may be wondering why 8-5 Duke belongs rated ahead of 8-5 North Carolina, who beat Duke 47-45 in double overtime, so I thought I would take a moment to address this issue. As you can probably guess, although both teams are 8-5, their records are not actually equal. That's because Duke played 5 ranked teams and North Carolina played 2 plus Duke. That's a huge schedule disparity. Both teams took an ugly upset loss to 3-9 Virginia, but NC also took an upset loss to 7-6 Georgia Tech. NC's head-to-head win over Duke makes up for that, but Duke defeated Clemson and North Carolina State, both of whom NC lost to, and so in the end, Duke was effectively 2 games better than North Carolina for the season.

Duke #25!

I've spent days on this issue now, and at this point I have looked at many more teams and rating possibilities than I've come close to covering in this already-long article. But in the end, Duke gets slot #25.

Kentucky

Before I wrap this up, I offer a few words on 7-6 Kentucky. Kentucky took one upset loss, 17-14 to 5-7 South Carolina, but they made up for it with a 38-31 upset win at 10-4 Louisville (now #20). They lost a very close game 38-35 to now-#22 Clemson in the Gator Bowl, and they could viably be ranked right behind Clemson at #23. It would also be viable to rate Kentucky ahead of Louisville, if Louisville were to be ranked behind Clemson (Clemson > Kentucky > Louisville > North Carolina State being the victory chain).

However, 7-6 Kentucky has the same relevant record as 8-5 Duke (now #25), and the AP poll rated Duke just a hair above Kentucky in the Others Receiving Votes section of the AP poll, so Kentucky will have to settle for #26 at best (though Oregon State actually has a good argument against them). It's just as well, because Louisville is the only decent team Kentucky defeated, and in fact Louisville is the only winning team of any kind that they beat. In addition to Duke's 2 upset wins over rated teams, they defeated a pair of nearly-rated teams, 8-5 Northwestern and 11-3 Troy.

Fixed AP Poll Top 25

Wow, finally done! It's been years since I've had such a difficult time fixing an AP poll. The funny thing is, for all the sweat and toil I've put into this, almost all of the changes to this year's poll are small (mostly cases where 2 teams switch places). The only major changes are Southern Cal moving up from #30 to #19 and Duke moving up from #36 to #25, and those don't seem so major because they occur near the bottom of the top 25.

Two teams fall out of this fixed AP poll top 25, 11-3 Southern Methodist (#22) and 10-4 Iowa (#24). They are replaced by the aforementioned 8-5 Southern Cal (now #19) and 8-5 Duke (#25). The 2 teams that fall out totaled 3 losses to unranked teams, and they did not defeat any team that the AP poll had rated. The 2 teams that replace them totaled 4 losses to unranked teams, one game worse, but they collected a far superior 4 wins against AP-rated teams (2 each). That's a big win for the replacements, and thus for the fixed AP poll over the original.

And here it is, your fixed final AP poll for the 2023 season, now logically coherent.

1) Michigan 15-0	--
2) Washington 14-1	--
3) Texas 12-2	--
4) Alabama 12-2	+1
5) Georgia 13-1	-1
6) Oregon 12-2 Florida State 13-1	-- --
8) Missouri 11-2	--
9) Ohio State 11-2	+1
10) Mississippi 11-2	-1
11) Louisiana State 10-3	+1
12) Penn State 10-3	+1
13) Arizona 10-3	-2
14) Oklahoma 10-3	+1
15) Notre Dame 10-3	-1
16) Oklahoma State 10-4	--
17) Kansas State 9-4	+1
18) Tennessee 9-4	-1
19) Southern Cal 8-5	IN
20) Louisville 10-4	-1
21) North Carolina State 9-4	--
22) Clemson 9-4	-2
23) Kansas 9-4	--
24) Liberty 13-1	+1
25) Duke 8-5	IN

OUT: #22 Southern Methodist 11-3
#24 Iowa 10-4

Fixed AP Polls
Home