There are a lot of ways to rank football teams. Sometimes very different approaches produce the same results.
Last week I introduced what I thought was a fun update to the long-lost BCS rankings of yore -- a combination of poll averages and computer ratings (of both the power rating and résumé ranking varieties) -- as an alternate way of both looking at the College Football Playoff landscape and evaluating outliers in the CFP committee's rankings. Then, on Tuesday night, the committee produced no major outliers. The top nine teams as picked by the committee and using my formula were exactly the same.
This week, we walk through the process once again. How will the committee react to a couple of top Pac-12 teams losing? Does LSU's status take a knock after an unimpressive slog over Arkansas? (If so, does USC benefit?) What matters more, TCU's road win over a ranked Texas team or Tennessee's search for style points against Missouri? Let's rank teams BCS style, then see where the committee disagrees.
This week's résumé SP+ rankings
We once again start with résumés. While my SP+ ratings, like ESPN's FPI, are designed to be predictive and forward facing, I created what I call "résumé SP+" in an attempt to look backward at how a team has played against its opponents.
Résumé SP+ is a look at two things: (1) how the average SP+ top-five team would be projected to perform against a given team's schedule -- in terms of scoring margin (which I cap at 50 points) instead of straight wins and losses -- and (2) how the team's scoring margin compares to that projection. Throw in a seven-point penalty for every loss a team has suffered (because losses matter on the résumé!), and we have what the CFP rankings would look like if SP+ and only SP+ were in charge.
This week's résumé SP+ top 15:
1. Georgia (10-0)
2. Ohio State (10-0): 1.1 points behind
3. Michigan (10-0): 8.1 points behind
4. Tennessee (9-1): 10.9 points behind
5. TCU (10-0): 14.8 points behind
6. Alabama (8-2): 20.8 points behind
7. USC (9-1): 24.8 points behind
8. Penn State (8-2): 28.0 points behind
9. Clemson (9-1): 28.2 points behind
10. Utah (8-2): 28.7 points behind
11. LSU (8-2): 28.7 points behind
12. Oregon (8-2): 31.0 points behind
13. Ole Miss (8-2): 31.4 points behind
14. Kansas State (7-3): 32.8 points behind
15. North Carolina (9-1): 33.0 points behind
The top six remains the same as last week, though both Tennessee and TCU closed the gap a bit -- TCU via a solid road win over a Texas team the computers love and Tennessee via some late-round haymakers. Up just 28-24 on Missouri early in the third quarter, the Volunteers scored 38 points in the game's final 24 minutes, even going deep in the final two minutes and using a timeout to assure one last touchdown in a 66-24 win. I doubt that sways the CFP committee at all, but it did add 0.7 points to Tennessee's rating here.
This week's biggest jumps: Penn State and Kansas State each moved up six spots thanks to dominant wins over decent teams -- PSU beat Maryland (now 41st in SP+) 30-0, while K-State beat Baylor (25th) 31-3 on the road. North Carolina moved up two spots after a win over Wake Forest, but despite one loss, the Tar Heels' résumé clearly lacks. They have yet to face a team in the SP+ top 30, the best team they played (No. 36 Notre Dame) beat them by 13, and their only top-50 wins are over No. 39 Wake Forest and No. 49 Pitt.
On the flip side, Oregon and Ole Miss each moved down five spots after losses to Washington and Alabama, respectively, and UCLA plummeted from 10th to 18th following a jolting upset loss to Arizona.
What does The Formula have to say?
First, allow me to acknowledge a massive missed opportunity: When I went through this redesigned BCS process last week, I simply named the formula at hand "The Formula." As one of our editors and folks on Twitter pointed out, I was missing a wide-open opportunity to call it the Bill Connelly System -- or BCS, for short. What a complete lack of awareness and opportunism on my part. I must live with this mistake forever.
Anyway, The Formula is indeed a BCS-esque approach, using an average of the two most prominent college football polls (the AP poll and the coaches poll) for two-thirds of the rating and an average of four computer ratings -- FPI and SP+ (both because they are two of the most accurate ratings available and they are very familiar to me) for team quality, strength of record and résumé SP+ (ditto) for proper résumé evaluation -- for the remaining one-third. (You can find further detail in last week's piece.)
This approach basically nailed the top of last week's CFP rankings, albeit with a few outliers further down.
The Formula, last week:
The committee still liked Clemson more than either the computers or my eyeballs and, perhaps because of that, ranked both Notre Dame (which served Clemson its only loss the week before) and Florida State (which suffered a tight loss to the Tigers a few weeks ago) a bit higher as well. But the committee and The Formula agreed about the top teams.
What has changed after last weekend's games?
The Formula, this week:
Falling out of the top 25: NC State (from 17th to 27th), Illinois (19th to 29th), Liberty (22nd to 54th) and Kentucky (25th to 43rd).
Moving in: Florida State (from 26th to 20th), Cincinnati (27th to 21st), Oklahoma State (30th to 24th) and Oregon State (36th to 25th).
The top five are unchanged, though while the averages for the top three teams remained almost exactly the same, TCU and Tennessee each closed the gap a hair. The polls are very much propping up the rankings of both USC (which would be 10th if this was just a computer average) and LSU (which would be 11th), while the computers, unfooled by a lofty 9-1 record, are dragging UNC down a couple of spots.
This week's most interesting teams
With the top five teams all winning, four by comfortable margins over unranked foes, it's fair to assume we won't see much movement at the top of this week's CFP rankings when they are revealed Tuesday night (9 ET on ESPN). TCU's résumé got a boost from its road win over Texas -- though if the Longhorns fall out of the CFP top 25 because of it, that might mitigate the boost a bit, which is quite silly when you think about it -- but all indications suggested the Horned Frogs were far enough back of the top three that the win probably won't change their ranking.
Still, we're always searching for signs of how the committee views certain teams and certain factors, so here are a couple of things I'm curious about for Tuesday night.
Does USC pass LSU? While linebacker Harold Perkins Jr. was unbelievable Saturday in Fayetteville, the rest of his LSU team was not. The Tigers' offense barely made the trip from Baton Rouge, and LSU needed some late stops to get past Arkansas and its backup quarterbacks 13-10. As was the case last week, the Tigers' lead over USC, per The Formula, is minuscule. Would the committee members bump the Trojans ahead as an acknowledgment of the Tigers' troubles?
My guess is they won't; among other things, USC's Week 11 victim was a lowly, one-win Colorado team, which probably won't move the needle much. If they did, however, it would be a useful data point, serving as a reminder that, with UCLA, Notre Dame and a potential Pac-12 championship game opponent still to come on the schedule, Lincoln Riley's Trojans remain viable in the playoff race and could still move up quite a bit in the coming weeks. (They obviously could move up even if they remain behind LSU this week, but moving up to No. 6 would be an extremely encouraging development.)
How far do Oregon and UCLA fall? In Week 11, the Pac-12 went from having three one-loss CFP contenders to having just one thanks to Oregon and UCLA both losing as double-digit favorites. The Formula was pretty unforgiving: The Ducks dropped from sixth to 12th (despite Washington moving up to 15th with the win), while the Bruins dropped from 11th to 16th.
The losses both hurt the Pac-12's CFP odds and proved the conference's depth -- it now has five teams ranked between seventh and 15th per The Formula. Might Oregon in particular find a softer landing spot due to both Washington's quality and the conference's strength? This could again say at least slightly positive things about USC's potential path into the top four in the coming weeks.
Did the committee get it right in 2014?
Everything is pretty straightforward at the moment in the 2022 CFP race: The top four teams are all unbeaten, and the No. 5 team (assuming Tennessee remains No. 5) lost only to the No. 1 team. Ohio State and Michigan play each other in a couple of weeks, and if higher-ranked teams win all the major remaining games, we'll end up with a CFP consisting of Georgia, the Ohio State-Michigan winner, TCU and either Tennessee, the Ohio State-Michigan loser, USC or Clemson.
There are almost always surprises, of course, but while we wait to be shocked, the picture is pretty clear. In the future, however, things could get blurrier. We don't yet know when the CFP will expand to 12 teams -- it might be 2026, and it might be as soon as 2024 -- but we know that it will be expanding. (If we end up with some wild "three SEC teams get into the CFP" scenario this year, expansion might happen by next fall. I'm sort of joking, but only sort of.)
While the current committee process is used for picking four playoff teams and some bowl pairings, it will also be picking byes and home games and attempting to separate teams with more flawed résumés for important playoff spots in the 12-team future.
What can The Formula tell us about how the committee might handle this future job? To peek at an answer, I revisited maybe the trickiest season of the CFP era: 2014. That year featured, among other things, only one unbeaten team (a very flawed Florida State), an almost impossible-to-separate trio of one-loss Ohio State, Baylor and TCU teams, and a veritable truckload of three-loss teams. How did the committee perform as compared to The Formula?
Once again, it did pretty well. Here's how The Formula would have ranked teams at the end of the 2014 regular season:
1. Alabama (12-1): 0.972 Formula average (No. 1 in CFP rankings)
2. Oregon (12-1): 0.956 (No. 2)
3. Florida State (13-0): 0.948 (No. 3)
4. Ohio State (12-1): 0.887 (No. 4)
5. TCU (11-1): 0.881 (No. 6)
6. Baylor (11-1): 0.880 (No. 5)
7. Michigan State (10-2): 0.806 (No. 8)
8. Mississippi State (10-2): 0.786 (No. 7)
9. Ole Miss (9-3): 0.724 (No. 9)
10. Kansas State (9-3): 0.693 (No. 11)
11. Georgia Tech (10-3): 0.689 (No. 12)
12. Arizona (10-3): 0.663 (No. 10)
13. Georgia (9-3): 0.645 (No. 13)
14. UCLA (9-3): 0.588 (No. 14)
15. Missouri (10-3): 0.565 (No. 16)
There are some minor differences -- Baylor and TCU switch places, as do Michigan State and Mississippi State, and Arizona's No. 10 CFP ranking looks rather odd considering both the polls and computers had Kansas State and Georgia Tech higher than the Wildcats -- but The Formula gave us the same names in the top four and the same names in the top 12.
This serves as a reminder of two things. First, it reminds us just how difficult separating Ohio State, TCU and Baylor was that year. Only seven thousandths of a point separated them in the averages above. Second, it reminds us that while there are pretty dramatic differences between a BCS-like process and the current committee approach, they take us to approximately the same place in the end. (I wrote that sentence for my own benefit as much as anyone's. Maybe that will tamp down a rant the next time the committee does something I don't like.)