Week 11 CFP Rankings with Context

Numbers can be a good way of comparing football teams, but only if you know what numbers tell you and more importantly, what they don't. Much of what a number can tell you depends on context. At Buckeye Planet, we use differential statistics to compare teams because they have context built into them. Some quick examples before we get started. If you score twice as much as your opponents typically allow and pass for twice as many yards, then your Differential Scoring Offense and you Differential Passing Offense are both 2.0 (twice as good as average). If you allow half as many yards rushing as your opponents gain against everyone else and allow half as many points, then your Differential Rushing Defense and your Differential Scoring Defense are both 0.5 (also twice as good as average). For more complete explanations of the numbers herein and what the categories mean, see Our Glossary. See especially the entry on Differential Scoring Composite (DSC), which is an objective measure which we use to compare teams.

There are some surprises in the numbers this week. The first surprise this week is really only a surprise if you missed the Differential Statistical Analysis (DSA) rundown of the CFP Top 12 last week.

Team	DYpC ratio	DYpC Rank	CFP Rank
Ohio State	1.523	1	2
Michigan	1.396	3	3
Alabama	1.334	8	8
Oregon	1.327	9	12
Georgia	1.319	10	1
TCU	1.232	17	4
Penn State	1.210	19	11
Tennessee	1.207	21	5
USC	1.141	31	7
Utah	1.127	34	10
LSU	1.116	37	6
Clemson	1.076	45	9

As was explained last week, Ohio State's ranking in DYpC (2nd last week, 1st this week) does not mean that Ohio State doesn't have things to work on. Even though there is a good deal of context baked into these numbers, some important context is missing. Part of that could be explained by the number we call Rigidity, which is an indication of how well you hold up under the pressure of better competition. That number tells the same story this week as it did last week, but other numbers might tell the story better. For example, the Buckeyes gain a first down on 43.24% of running plays on 3rd down with 1 to 3 yards to go. If you're thinking, "that's bad", you're right. That's bad. How bad? 125th in the country bad.

It is worth noting, however, that Ohio State is not the only CFP-Top-Twelve team to rank lower than you might think in short yardage first downs. Everyone seems to think that Michigan is the king of running the ball when the other team knows it's coming, but they are only 17th in this metric (much better than 125th, granted, but surprisingly low for all that). Georgia is 36th, LSU is 42nd, TCU is 55th, Tennessee is 75th, and Alabama is 105th.

This calls into question the importance of short-yardage rushing. The entire Top Twelve fit into the top 45 in DYpC, yet half of the top 8 are ranked 55th or lower in short-yardage rushing. Additionally, 2 of the top 4 in short-yardage rushing are 4-6 teams, and number 6 is Nevada, who is 2-8. So while Ohio State's success rate on short yardage is a concern, and while it prevents us from proclaiming them the best rushing team in FBS, it should be kept in perspective.

Unlike the offensive numbers, one doesn't need to dig as deep to see a cause for concern for Ohio State fans when it comes to rushing defense.

Team	DYpC ratio	DYpC Rank	CFP Rank
Michigan	0.681	2	3
Georgia	0.698	3	1
Alabama	0.702	5	8
Tennessee	0.729	6	5
Louisiana State	0.805	14	6
Penn State	0.841	21	11
Ohio State	0.841	22	2
Texas Christian	0.873	28	4
Oregon	0.898	34	12
Clemson	0.945	46	9
Utah	0.948	48	10
Southern Cal	1.098	105	7

Ohio State's number changed significantly, as did several other teams. For Ohio State though, this might be a cause for concern. The Buckeyes allowed 150 yards on 40 carries to Indiana. That might not seem so bad, until you consider how bad Indiana is. This was the most rushing yards that Indiana gained against an FBS opponent all year. Their previous high against a B1G opponent was 67 yards. This left the Buckeyes with a DRD of 1.939 for the game and a DYpC of 1.648 for the game. These were BY FAR the Buckeyes worst rushing defense numbers on the year, and they follow a game against Northwestern when both numbers were over 1 in spite of Ohio State knowing that Northwestern wouldn't be able to throw the ball in that Nor'easter. The upshot of all of this is that, according to DSA, Ohio State's rushing defense appears to have fallen off the table, and it couldn't have come at much worse of a time. Here's hoping it gets shored up this week.

Some might be thinking, "perhaps the Buckeyes just focused on stopping the pass the last two weeks." Well... obviously that wasn't the case vs Northwestern. But against Indiana the Buckeyes were facing a backup QB not known for his throwing arm and they STILL allowed a Differential Pass Efficiency greater than 1. It is conceivable that this was just another case of an opposing QB playing the best game of his life against Ohio State (see Penn State quarterbacks of the 21st century), but it is still a cause for concern. The Buckeyes fell from 16th to 25th in FBS in Differential Pass Efficiency defense this week. So while the rushing offense (leading the country in DYpC) is, not without reason, a leading topic of discussion on the message board of late; the defense should be considered a cause for concern. (Though as will be seen later, they have been keeping other teams out of the end zone fairly well.)

As with pass defense, pass offense has taken a down turn of late. But CJ is still on top of the world, as has been shown elsewhere. Rather than go over that same ground again, we'll simply observe here that CJ is the only quarterback in the country to have a Differential Passing Efficiency (DPE) above 1.5 (1.535), but that if the Northwestern Nor'easter is removed from consideration, CJ's DPE is 1.643. If we were to give Hendon Hooker a mulligan for the Georgia game (NOT played in a Nor'easter), his DPE improves to 1.549. Almost within a 1/10th of CJ, but that is still the biggest gap between adjacently ranked quarterbacks in FBS. However you slice it, it's CJ and then everyone else.

Having briefly reviewed rushing and passing, this leaves us with the most important stats, scoring offense and scoring defense. As usual, we use the differential version of these. If you're new to The Planet, refer to the Glossary link at the top of the post for a definition and explanation of DSO, DSD, DSC, and other DSA terms.

Several teams made big moves in DSC this week. Some of these came with a concomitant big move in CFP ranking, some didn't. Apart from their changes in DSC and (if applicable) CFP rankings (shown in the table below), the teams had something else in common:

Team	Wk 10 DSC	Wk 11 DSC	Wk 11 CFP	Wk 10 CFP	Wk 11 Result
Michigan	3	2	3	3	34-3
Kansas St.	11	7	19	15	31-3
Florida St.	15	9	23	19	38-3
Minnesota	19	13	NR	NR	31-3
Air Force	32	21	NR	NR	35-3
Boise St.	44	23	NR	NR	41-3

So apart from being the biggest movers in terms of DSC, and therefore in DSC ranking, they all allowed exactly 3 points in Week 11. They are all also ranked lower by the CFP than they are by DSC; in fact the last 3 are unranked in spite of being top 25 in DSC.

This illustrates a couple of things. 1) A good defensive outing will cause more movement in DSC than a good offensive outing (denominator more important than numerator), and 2) the CFP doesn't seem to notice defense.

By the way: Anyone who expected to see Florida State in the top 10 in DSC at this point in the season raise your hand. And Minnesota at 13?

Apart from those things, some of you may have noticed the team at the top of the previous table. Although their DSC ranking didn't move much, moving at all that close to the top is significant, especially if you remember where Ohio State was ranked last week.

Team	DSO ratio	DSO Rank	DSD ratio	DSD Rank	DSC power	DSC Rank	CFP Rank
Georgia	1.542	6	0.411	1	3.755	1	1
Michigan	1.607	3	0.489	2	3.285	2	3
Ohio State	1.997	2	0.610	8	3.275	3	2
Tennessee	2.054	1	0.737	22	2.786	4	5
Alabama	1.586	4	0.615	9	2.577	5	8
Penn State	1.447	9	0.638	11	2.267	6	11
Kansas State	1.191	35	0.622	10	1.916	7	15
Texas	1.237	27	0.659	13	1.877	8	NR
Florida State	1.378	11	0.742	24	1.857	9	19
Utah	1.313	17	0.709	18	1.851	10	10
UCF	1.317	15	0.718	19	1.836	11	20
Louisiana State	1.175	37	0.675	14	1.740	12	6
Minnesota	1.005	64	0.598	7	1.680	13	NR
Illinois	0.970	66	0.581	6	1.669	14	NR
Clemson	1.353	13	0.830	32	1.629	15	9
Iowa	0.889	80	0.549	3	1.621	16	NR
Louisville	1.129	45	0.725	21	1.557	17	NR
Southern Cal	1.463	8	0.940	49	1.556	18	7
Texas Christian	1.279	21	0.842	33	1.520	19	4
Notre Dame	1.145	41	0.764	27	1.499	20	18
Wisconsin	1.305	18	0.896	43	1.456	24	NR
Oregon	1.365	12	0.998	59	1.367	29	12

Before we get to the elephant in the room, let's first address the small matter of Tennessee passing the Buckeyes in Differential Scoring Offense. This was based entirely on the fact that Tennessee scored 10 more points than Ohio State did in week 11. The teams both scored 28 points in the first half, so the difference was entirely in the 2nd half, when Ohio State was coasting. So Tennessee, who needs style points, kept scoring and Ohio State, who does not need style points, did not. *shocking*



With that out of the way... yes, TTUN passed Ohio State in DSC this week. And yes, DSC already has a lot of context built into it. But in this case, there is a significant bit of context that numbers won't tell you about.

The Cornhuskers were without their starting QB for the second straight week. This left Sophomore Logan Smothers and Freshman Chubba Purdy as the Nebraska signal callers. While numbers can't tell you that the starter was missing, they can give you a hint as to the size of the drop off.

Casey Thompson's lowest passing efficiency on the year was 128.02; Nebraska's lowest DPE while he was the starter was 0.905. Neither Purdy nor Smothers has managed a PE within 10 points of Thompson's lowest number on the year, and their collective DPE was 0.654 against Minnesota and 0.799 against TTUN.

So while DSA typically provides relevant context by comparing how well you did compared to your opponent's usual performance, DSA compared what Michigan's defense did vs Chubba and Logan with what Nebraska's other opponents did against a much better QB (and Minnesota did better against those backups than Michigan did). The upshot is: In this case the context that DSA added was misleading.

Yes, holding Nebraska to 3 points greatly improved Michigan's DSC. As we saw above, that worked for half a dozen teams this past weekend. The difference is that Michigan's DSC-bump was 100% fraudulent, much like the DSC-hit that the Buckeyes took in the Northwestern Nor'easter was misleading in the other direction.

It's all about context. When you look behind the numbers and find the right context, The Ohio State Buckeyes compare favorably to everyone this side of Athens. And there is a lot of time for the Ohio State defense, still in its first year of a new system, to continue to improve between now and then.

 

It is worth noting, however, that Ohio State is not the only CFP-Top-Twelve team to rank lower than you might think in short yardage first downs. Everyone seems to think that Michigan is the king of running the ball when the other team knows it's coming, but they are only 17th in this metric (much better than 125th, granted, but surprisingly low for all that). Georgia is 36th, LSU is 42nd, TCU is 55th, Tennessee is 75th, and Alabama is 105th.

This calls into question the importance of short-yardage rushing. The entire Top Twelve fit into the top 45 in DYpC, yet half of the top 8 are ranked 55th or lower in short-yardage rushing. Additionally, 2 of the top 4 in short-yardage rushing are 4-6 teams, and number 6 is Nevada, who is 2-8. So while Ohio State's success rate on short yardage is a concern, and while it prevents us from proclaiming them the best rushing team in FBS, it should be kept in perspective.

One of the best points (above) I've ever seen made on a football message board. When people watch games and see their team get stuffed on a run on 3rd and short they assume it means a great deal, whereas the meaning they're ascribing to it is driven primarily by the emotion they're experiencing during it. "Now we have to punt and I feel like shit, so that play must've meant a lot."

The degree of significance of any facet of the game is determined by its correlation with winning. If rushing success rate on 3rd and short correlates relatively weakly with winning, as is alluded to in the above post, then one should feel about that like they do about the yardage their team is averaging on punts, for example. Rushing success on 3rd and short may be more important than that, per se, but nobody should be running around believing a team can't possibly be successful if it's relatively weak in that area.

The hopes and dreams for the 2022 OSU football team shouldn't be riding on how they run the ball on 3rd and short.

 

Ohio State's number changed significantly, as did several other teams. For Ohio State though, this might be a cause for concern. The Buckeyes allowed 150 yards on 40 carries to Indiana. That might not seem so bad, until you consider how bad Indiana is. This was the most rushing yards that Indiana gained against an FBS opponent all year. Their previous high against a B1G opponent was 67 yards. This left the Buckeyes with a DRD of 1.939 for the game and a DYpC of 1.648 for the game. These were BY FAR the Buckeyes worst rushing defense numbers on the year, and they follow a game against Northwestern when both numbers were over 1 in spite of Ohio State knowing that Northwestern wouldn't be able to throw the ball in that Nor'easter. The upshot of all of this is that, according to DSA, Ohio State's rushing defense appears to have fallen off the table, and it couldn't have come at much worse of a time. Here's hoping it gets shored up this week.

Some more context....

With sacks removed (as they should be, but are not in CFB), Indiana actually rushed 36 times for 179 yards, which makes the Buckeyes' numbers even worse.

But some more context....

Here is the length of every rush by Indiana: -2; -2; -1; -1; -1; 0; 0; 1; 1; 1; 2; 2; 2; 2; 3; 3; 3; 3; 3; 4; 4; 4; 4; 4; 4; 5; 6; 6; 7; 7; 8; 10; 11; 16; 16; 44

Factor out the clear anomaly (the 44-yard run on a trick play to begin the 3rd quarter), and the Buckeye defense gave up 135 yards on 35 carries, which isn't so bad.

Now factor out the QB runs from Dexter Williams II, the Hoosiers' reserve QB for whom the defense did not game plan: -1; 2; 3; 4; 4; 4; 5; 7; 7; 8; 16; 16

So the Buckeye defense allowed 23 carries for 60 yards on "traditional" runs, with 16 of those 23 runs resulting in a "win" for the defense (3 yards or less allowed): -2; -2; -1; -1; 0; 0; 1; 1; 1; 2; 2; 2; 3; 3; 3; 3; 4; 4; 4; 6; 6; 10; 11

The Buckeyes clearly have problems containing QBs who can run, and they are still susceptible to the occasional trick play, but it's not like the defense is getting consistently gashed up the middle like last year.

 

Some more context....

With sacks removed (as they should be, but are not in CFB), Indiana actually rushed 36 times for 179 yards, which makes the Buckeyes' numbers even worse.

But some more context....

Here is the length of every rush by Indiana: -2; -2; -1; -1; -1; 0; 0; 1; 1; 1; 2; 2; 2; 2; 3; 3; 3; 3; 3; 4; 4; 4; 4; 4; 4; 5; 6; 6; 7; 7; 8; 10; 11; 16; 16; 44

Factor out the clear anomaly (the 44-yard run on a trick play to begin the 3rd quarter), and the Buckeye defense gave up 135 yards on 35 carries, which isn't so bad.

Now factor out the QB runs from Dexter Williams II, the Hoosiers' reserve QB for whom the defense did not game plan: -1; 2; 3; 4; 4; 4; 5; 7; 7; 8; 16; 16

So the Buckeye defense allowed 23 carries for 60 yards on "traditional" runs, with 16 of those 23 runs resulting in a "win" for the defense (3 yards or less allowed): -2; -2; -1; -1; 0; 0; 1; 1; 1; 2; 2; 2; 3; 3; 3; 3; 4; 4; 4; 6; 6; 10; 11

The Buckeyes clearly have problems containing QBs who can run, and they are still susceptible to the occasional trick play, but it's not like the defense is getting consistently gashed up the middle like last year.

I miss GPAs