Home

Comments on the AP College Football Poll: December 4, 2016

1) Alabama 13-0
2) Ohio State 11-1
3) Clemson 12-1
4) Washington 12-1
5) Penn State 11-2
6) Michigan 10-2
7) Oklahoma 10-2
8) Wisconsin 10-3
9) Southern Cal 9-3
10) Florida State 9-3
11) Colorado 10-3
12) Western Michigan 13-0
13) Oklahoma State 9-3
14) West Virginia 10-2
15) Louisville 9-3
16) Stanford 9-3
17) Auburn 8-4
18) Virginia Tech 9-4
19) Louisiana State 7-4
20) Florida 8-4
21) Iowa 8-4
22) Pittsburgh 8-4
23) Temple 10-3
24) Nebraska 9-3
25) South Florida 10-2

Others Receiving Votes
Houston 9-3
Utah 8-4
Boise State 10-2
Washington State 8-4
Texas A&M 8-4
Air Force 9-3
Navy 9-3
Tennessee 8-4
San Diego State 10-3
Western Kentucky 10-3
Miami (Florida) 8-4
Georgia Tech 8-4
Kansas State 8-4

If you've been wondering why I haven't been fixing the AP college football poll every week, it's because I don't bother to fix the AP poll's top 25 until after the final edition is released in January. When there are no more games to be played, there is no longer any wiggle room for ignoring head-to-head results and season-long performance. But before that, there can be plenty of room for wiggling.

For example, 10-2 Michigan is currently ranked #6, one spot behind 11-2 Penn State, but Michigan defeated Penn State by the very decisive score of 49-10, and should therefore be rated ahead of PSU (more on this issue below). However, if Michigan loses to #10 Florida State in the Orange Bowl, then the AP poll voters will have been proven correct, so we'll wait and see what happens in the bowl games before fixing the AP poll.

But while I will not fix the AP poll until the final one in January, I will go ahead and point out problems with the current poll that would be outright errors if the season had ended this last weekend, and no more games were to be played. This can also give voters (in the AP or in any other poll) an idea of potential errors to avoid going forward.

Playoff Committee Rankings

Two years ago, the playoff committee top 25 was superior to the AP poll. Last year, the committee's top 25 was practically the same as the AP poll. This year, the committee's top 25 has degraded further, and it is now significantly worse than the AP poll's top 25. Why this is, I cannot say. Theoretically, the committee's rankings should be better because they spend a lot more time on it than AP poll voters presumably do, and they actually debate with each other face to face, which should further sharpen their rankings.

I will address the differences between the committee's top 25 and the AP poll's top 25 at the end of this article. However, because the playoff committee will not be ranking teams in a final post-bowl top 25, I will not be fully fixing any of their rankings like I will the final AP poll top 25 in January.

Now let's get to commenting on the problems with the latest AP poll...

Michigan vs. Penn State

As I predicted would happen last week, the winner of the Big 10 title game, 11-2 Penn State, has passed up 10-2 Michigan in this week's AP poll. And as I said last week, this is not a correct rating. Michigan already defeated Penn State 49-10, and it doesn't get any clearer than that. They should be rated higher. Furthermore, it makes no sense at all to move PSU ahead of Michigan as a result of PSU beating Wisconsin, given that Michigan has already beaten Wisconsin themselves.

Of course, that is not the real reason AP poll voters moved PSU ahead of Michigan. They did it because PSU is now the Big 10 champion. Clearly the playoff committee's supposed emphasis on conference titles has bled into the minds of AP poll voters, and it is no wonder, given that sportswriters and broadcasters have been hammering this issue all week long. So let me address the issue here.

The Proper Valuation of Conference Titles When Ranking Teams

My position is that conference titles should mean nothing when rating teams. Zip. Zero. Nada. A conference title is its own reward, with a nice trophy that its winner should be proud of, but that trophy does not itself make its owner any better of a team.

And that is the important point: logically speaking, being named a conference champion does not itself make a team any better. If Clemson was an independent team that played exactly the same schedule that they did, it would not make them any worse a team than they are now with an ACC title. And if Pittsburgh was in the Big 10, then Penn State would not be conference champion, and would certainly be ranked behind Michigan. Yet Penn State is exactly the same team in terms of power level whether Pittsburgh is in the Big 10 or not, and whether PSU is named conference champion or not.

Penn State won the Big 10 because Pittsburgh is not in the conference, but the fact is, they still lost that game, and when rating PSU against Michigan (or Ohio State) in a national top 25, that loss counts, and it is the reason PSU should be ranked behind both teams.

In the end, it is as simple as that: a conference race is decided by a subset of games played by teams in that conference, while a national top 25 is based on all of the games played by all of the teams.

Oklahoma

10-2 Oklahoma is ranked #7, 10-3 Wisconsin #8. But don't be fooled by those straight records: Wisconsin's 3 losses all came to teams ranked higher than Oklahoma, whereas Oklahoma took an upset loss to an unranked team, so Wisconsin's relevant record is effectively a game better than that of Oklahoma. Not only that, but all 3 of Wisconsin's losses were close, including an overtime loss to Ohio State. And Ohio State beat Oklahoma 45-24!

Furthermore, Wisconsin has beaten 3 ranked teams, Oklahoma 2, and Wisconsin has posted 1 close win (touchdown or less) over an unranked opponent, Oklahoma 3. So this is not even close: Wisconsin should be rated higher.

Frankly, I find it rather dubious that Oklahoma should even be rated higher than #9 Southern Cal and #11 Colorado.

Colorado and Florida State

10-3 Colorado falls 1 spot behind 9-3 Florida State this week, and I think this ranking is wrong. All 3 of Colorado's losses are to teams ranked higher than Florida State, while FSU has taken 2 losses to teams ranked behind Colorado, so Colorado is effectively 2 games better than FSU for the season and should be ranked higher.

Oklahoma State and West Virginia

Here is an interesting issue: 9-3 Oklahoma State is ranked #13, 10-2 West Virginia #14, and what makes this interesting is that the AP poll voters appear to be disregarding OSU's loss to Central Michigan. You may recall that early in the season, Central Michigan was incorrectly given one untimed down at the end of the game, with which they completed a miracle touchdown pass to win. I can certainly understand disregarding that loss, but my position is, and always has been, that losses caused because one team was "robbed" by the officials are still losses. This is certainly the position of the NCAA, who will always regard OSU's game with CMU as an official loss.

In the end, OSU could have avoided being "robbed" by establishing more than a mere 27-24 lead over Central Michigan by that point in the game.

Still, as I said, I can understand why AP poll voters would make this choice, even if I disagree.

Houston and Louisville

I'm just going to basically repeat what I wrote last week:

Remember way back when 9-3 Houston slaughtered 9-3 Louisville 36-10? Less than 3 weeks ago? Apparently too far back in time for your average sportswriter to remember, but I tell you, that actually happened. So why is 9-3 Louisville ranked #15 and 9-3 Houston not ranked at all?

Maybe it's because Louisville owns an impressive upset win over #10 Florida State? Well, except that Houston tops that with an even more impressive upset win over #7 Oklahoma. So maybe it's because Louisville plays a tougher schedule, being in a major conference and all? Well, except that FSU is the only rated team Louisville has beaten, and more than that, FSU is the only team Louisville has beaten that even has a winning record! Houston has beaten 3 teams with winning records, 2 of them in the AP poll's own top 25, and another one that could be (9-3 Tulsa).

The relative rating of these 2 teams is simply atrocious. Houston should be rated ahead of Louisville, whom, again, they just beat 36-10 less than 3 weeks ago.

South Florida and Navy

Not only should 9-3 Louisville (#15) be ranked behind 9-3 Houston (#26), but both should be ranked behind 10-2 South Florida (#25) and 9-3 Navy (now way back at #32). It's a simple head-to-head chain: 10-2 USF beat 9-3 Navy, who beat 9-3 Houston, who beat 9-3 Louisville.

USF, Navy, and Houston are all American Athletic Conference teams, and all are underrated due to that conference's dim reputation in the minds of AP poll voters. But compare these teams to many of the teams ranked ahead of them:

#17 Auburn (8-4) has 2 losses to unranked teams, while #25 USF (10-2) has none and 9-3 Navy has just 1. Houston has 2 (Navy would be a 3rd, but they should be rated), but they also own 2 wins over teams ranked higher than Auburn.

#18 Virginia Tech has 3 losses to unranked teams (more on them below).

#20 Florida has 2 losses to unranked teams.

#21 Iowa has 2 losses to unranked teams, and one is an FCS team!

I think that 10-2 South Florida, 9-3 Navy, and 9-3 Houston all belong in the middle of the top 25. Of course, if one of them loses their bowl game, that will no longer be the case.

South Florida and Temple

Here is another conference title issue. 10-3 Temple is now ranked at #23, while 10-2 South Florida is ranked #25, and Temple's suddenly higher rating is obviously due to their having just won the American Athletic Conference title. It should be noted that Temple did beat USF earlier this season, 46-30 at home. However, Temple took 2 upset losses, to Memphis and Army, while USF took none, which is why USF finished a game better than Temple on the season despite the head-to-head result.

South Florida should be rated higher (though Temple should remain ranked higher than Navy, of course).

Stanford

Covered last week: 9-3 Stanford is underrated at #16. They have one upset loss, but remember that 9-3 Florida State (#10) and 9-3 Oklahoma State (#13) each have 2 upset losses. Stanford also owns an upset win over #9 Southern Cal, so Stanford is effectively 2 games better than FSU and OSU, and needless to say they should be rated higher than both. They belong ranked right behind Colorado (who, as covered above, should also be ranked ahead of FSU).

Virginia Tech

Covered last week: 9-4 Virginia Tech, ranked #18, is still the most overrated team in the AP poll. They do own a win over #22 Pittsburgh, but unfortunately they also own 3 losses to unranked opponents. Virginia Tech should be dropped from the top 25, and yes, that means that they should not even be rated higher than Pittsburgh.

This may seem confusing, since Virginia Tech is 9-4, Pitt is 8-4, and VT beat them 39-36. So let's go over this game-by-game. Yes, VT beat Pitt, so that gives them a 1-game lead over Pitt. Pitt took another 2 losses to unrated teams, but VT took 3 losses to unrated teams, so that makes VT and Pitt even, with VT having the head-to-head tiebreaker. Pitt's other loss came to #11 Oklahoma State, but since OSU is ranked higher than VT, that loss is irrelevant when comparing VT to Pitt. Pitt has 2 upset wins, over #3 Clemson and #5 Penn State, very impressive, and since both teams are ranked higher than VT, that leaves Pitt effectively 2 games better than VT on the season. Pitt should therefore be ranked higher.

But Virginia Tech should be dropped further than that: they belong behind 8-4 Tennessee (outside the top 25 at #33), who beat them 45-24. I'd call that score decisive.

Auburn, Louisiana State, and Florida

The AP poll has 8-4 Auburn at #17, 7-4 Louisiana State at #19, and 8-4 Florida at #20. I think these teams are all overrated (I touch on this issue in the "South Florida and Navy" section above), but what I want to discuss here is the relative order of these 3 teams. The AP poll's order might be viable, but I think that Florida > LSU > Auburn is a better order for these teams. This difference boils down to the fact that Florida's 16-10 win over LSU came on the road and late in the season (November 26), whereas Auburn's 18-13 win over LSU came at home and early in the season (September 24).

Boise State

Covered last week: 10-2 Boise State didn't quite make the AP poll's top 25, but they're at #28, well within striking distance, and there are a number of teams that should be ranked ahead of them, and therefore closer to making the final top 25 should some slots open up after the bowl games. First of all, we have 9-3 Air Force (#31), who defeated Boise State 27-20 in their finale. However, Air Force has taken 3 upset losses, so they shouldn't be up for a final top 25 slot either.

Luckily, there are even better options than Air Force. Texas A&M (8-4, #30) and Tennessee (8-4, #33) should be rated ahead of Boise State, and so should a pair of fellow "little big teams": 9-3 Tulsa and 8-4 Memphis. Speaking of which...

Tulsa and Memphis

Covered last week: the American Athletic Conference is having a terrific season so far! I've already covered 4 AAC teams that I think are underrated (#23 Temple, #25 South Florida, #26 Houston, and #32 Navy), and here we have another couple of underrated AAC teams, giving us half a dozen in all. I don't know if 9-3 Tulsa and 8-4 Memphis, neither of whom collected any AP poll points, are top 25 material, but I can tell you that both are severely underrated right now.

Certainly they belong ahead of teams like Boise State (as mentioned above) and Air Force (9-3, #31). Air Force, after all, has taken a huge 3 losses to unranked teams. 9-3 Tulsa has only taken 2 such losses, and 1 of those is to Houston, who is ranked ahead of Air Force.

8-4 Memphis has taken 3 losses to unranked teams, just like Air Force, but one of those is to Tulsa, who should be ranked ahead of Air Force (as discussed above). Memphis also owns 2 wins over teams that are ranked higher than Air Force, #23 Temple and #26 Houston.

North Carolina

Covered last week: 8-4 North Carolina should not be ranked at all, so the AP poll got that right, but a couple of voters gave 8-4 Miami (Florida) and 8-4 Georgia Tech a few poll points, putting those teams at #36 and #37, and UNC, who got no poll points at all, at least belongs ahead of those 2 teams. UNC beat Miami 20-13 on the road, and they beat Georgia Tech 48-20 at home.

Comments on the Playoff Committee's Top 25

As I previously indicated, the playoff committee's rankings have gotten worse, relative to the AP poll, each year, and this year their top 25 is not even close to as good as the AP poll's top 25. I see 14 differences between the playoff committee's top 25 and the AP poll top 25, and in my opinion, the AP poll made the better choice 11 times to 3! As I said, not even close. I'm not going to go over all of these differences, but here are some highlights:

Colorado vs. Florida State

The playoff committee correctly rated Colorado ahead of Florida State, unlike the AP poll (this issue covered earlier in article). This is the only significant difference that the committee was right about.

Clemson vs. Ohio State

The committee has Clemson #2 and Ohio State #3. I agree with the AP poll that Ohio State should be ranked higher. Using the committee's own rankings, OSU took their loss to #5 on the road, Clemson to #23 at home. OSU has beaten #6, #7, #8, and a pair of unranked 9-3 teams, while Clemson has beaten #11, #13, #14, and #22. And 4 of Ohio State's wins have been close (touchdown or less), Clemson 6. Sorry committee, but Ohio State should be ranked higher, and it is not close.

Oklahoma State vs. Western Michigan

The committee has Oklahoma State ranked #12 and Western Michigan #15, while the AP poll has WMU #12 and OSU #13. I compared these 2 teams in some detail last week. The committee is very wrong here.

South Florida vs. Navy

The committee has Navy ranked but not South Florida, while the AP poll has USF ranked but not Navy. USF is 10-2, Navy 9-3, and USF beat Navy, so the committee fell asleep on this one.

Home