Every Saturday night, Andy Staples and Ari Wasserman react to the weekend’s slate of games on The Andy Staples Show & Friends. On Mondays, Andy revisits his and Ari’s biggest takeaway from Saturday night’s instant reaction. This week: Ari gave everyone homework — rank the 13 teams eligible for the College Football Playoff.
The only rankings that actually matter debut on Tuesday. And even these don’t truly matter. Remember, the first time the College Football Playoff selection committee released a ranking in 2014, these were the top four:
- Mississippi State
- Florida State
- Ole Miss
How many of those teams actually made the inaugural CFP? One. The Seminoles went 13-0, entered the bracket as a No. 3 seed and got crushed by Oregon in the Rose Bowl. So don’t despair if your team isn’t in the top four on Tuesday when the committee reveals its first ranking of the 2022 season.
As long as your team is one of the Lucky 13, of course.
On the postgame edition of The Andy Staples Show, Ari and I determined which teams remain eligible for the CFP. We might be wrong, but eight seasons worth of selections have established a fairly reliable pattern. The committee has yet to place a two-loss team* into the top four. You don’t have to be a conference champion to make the top four, but you’d better not have a blowout loss. (Unless you avenged said loss in the conference title game or beat the team that blew you out earlier in the season.) At the end of the show, Ari gave all of us a homework assignment: Rank these 13 teams.
*You’ll notice two-loss LSU is omitted from the Lucky 13. This is based on committee precedent. Should LSU beat Alabama and then beat Georgia or Tennessee in the SEC title game, perhaps that changes this year. A two-loss Auburn probably would have made the bracket in 2017, but the Tigers lost their rematch against Georgia in the SEC title game.
Entering this week’s games, these are the 13 teams that can still make the CFP, listed by conference.
- Ohio State
- Ole Miss
That this many teams remain in the hunt means we’ve had a pretty fun season so far. Also, it doesn’t feel as if there are one or two teams that would absolutely smash everyone else still in the hunt. When the CFP expands to 12 in a few years, we’ll be able to measure the teams still in the hunt at this point by the dozen. But for now, let’s be happy the number is this high.
To complete Ari’s assignment, I tried to imagine how I’d vote as a committee member. I collected some stats I know are important to the committee. I also used some that I find important. I used the SP+ predictive ranking created by ESPN’s Bill Connelly. This is my favorite of the predictive ranking formulas, but I won’t quibble if you want to use ESPN’s Football Power Index or Jeff Sagarin’s rankings. (Bill’s formula can’t seem to accept that Texas isn’t back this year, but I’m willing to forgive that.)
I do like the FPI’s strength of schedule measurement, though. So I also used that. The FPI also has a handy measurement of remaining schedule strength, but that isn’t necessary for this exercise since we can only go by the games that have already been played. I also used the FPI’s strength of record, which measures how difficult a team’s record is to achieve based on the strength of its opponents, travel time, rest time and other factors.
One stat I love is net points per drive. This is the number of points a team’s offense averages per drive minus the average number of points that team’s defense allows on each opponent drive. Brian Fremeau keeps this stat on his excellent site. He also keeps available yards, which is another fun one. If a team gets the ball at its own 20, it has 80 yards available. If it scores a touchdown, then it gained 100 percent of available yards. I didn’t want to get too in the weeds, though. So I left that out.
Instead of using wins against Top 25 teams, which seems fairly arbitrary and also would require me to rank 25 out of 131 teams, I stole a concept from the NCAA Basketball Selection Committee. In basketball, the committee weigh Quadrant 1 (games against teams in the top 25 percent of the NET ranking) wins heavily. Football doesn’t have as many data points, so I decided to count Quadrant 1 and Quadrant 2 wins using SP+ as the ranking. Quad 1 is the teams ranked No. 1 through No. 30. Quad 2 is the teams ranked No. 31 through No. 60.
I also wanted to use some raw numbers that aren’t adjusted by any proprietary formula. So I went with tried-and-true yards per play gained and yards per play allowed. This adjusts for tempo better than total offense and total defense, and it also helps identify outliers.
Even though I know enough about these teams to make educated guesses as to their identities based on their numbers, I stripped the team names off my spreadsheet before I started sorting stats. My hope was that I would forget which team corresponded to which letter. That way, I could rank based solely on what the team had done this season and not on brand name, past success or failure or conference affiliation.
Does that make this ranking objective? Of course not. Rankings are by their nature subjective. At a certain point, I have to look at two (or three or four) data sets that seem quite similar and decide which one to place above the other(s).
Here’s my spreadsheet. Feel free to rank the teams as you see fit…
The actual committee chooses a bucket of about six teams in order to select its top three. It then scrubs through the list three at a time until it reaches 25. The six that seemed to belong at the top here were teams E, F, K, J, M and L.
So I moved them into a different spreadsheet and tried to parse them. Team J leads everyone with four Quad 1 wins but has a loss. Team M has three Quad 1 and two Quad 2 wins and the No. 1 strength of record. But Team M is one of only two on this list with a yards per play number above No. 15 in the nation. Its defense is No. 39 in yards per play allowed. But its offense is No. 3 in yards per play gained, and it is No. 5 in net points per drive. In other words, its defense might be giving up yards, but Team M usually is winning its games by a healthy margin.
Team K and Team F look cleaner. Neither has a loss, and both have single-digit ranks in the yards per play stats. Team K is No. 2 in net points per drive and has one Quad 1 win and three Quad 2 wins. Team F is No. 2 in strength of record and No. 1 in net points per drive. The drawback to these two? Their schedules haven’t been as difficult as Team J or Team M’s schedules.
Still, these two have been so consistent that I feel like I need to place them in the top two. So I’ll make Team F No. 1 and Team K No. 2. I’m only choosing the top three now, so I have to decide between Team J and Team M and then send the remaining team back to the pool. Team M’s No. 1 strength of record suggests that’s who I should pick, but I suspect Team M handed Team J its loss. I like using head-to-head results as a tiebreaker. (Otherwise why bother playing?)
So I peek at my key, which confirms my suspicion. Team M will be No. 3. Team J goes back in the pool.
My top three look like this:
- Ohio State (Team F)
- Georgia (Team K)
- Tennessee (Team M)
Now let’s move on. You’ve probably guessed by now that Team J is Alabama, but let’s try to ignore that knowledge and compare it with the next group.
We take the three remaining teams from the first group (J, E, L) and add three more teams (H, C, G).
The two that jump off the page are Team J and Team E. We’re trying very hard not to make any assumptions because we know who J is. What happened from 2009-21 is not important here. E has a similar strength of record, two Quad 1 and two Quad 2 wins and a better net points per drive rank. It seems the defense has been stingier but the offense isn’t quite as explosive. The biggest difference is strength of schedule. Team J’s strength of schedule is 10th out of 131. Team E’s is 79th, the lowest in this grouping of six. So let’s give the nod to Team J. Then Team E.
4. Alabama (Team J)
5. Michigan (Team E)
Now let’s choose No. 6 from the remaining four on our list (H, C, G, L). All of these teams have more flaws than the others, and those flaws seem to show up on defense. Team G has a loss but only one Quad 1 or Quad 2 win. So that team goes back in the pool. Team C’s strength of record is No. 3, meaning it has achieved something difficult relative to its schedule. Team L has the best net points per drive rank and has two Quad 1 wins and one Quad 2 win.
I think I’m going with Team C. After peeking at my key, I see I’ve ranked:
6. TCU (Team C)
I’ll spare you most the gory details, but I ranked the next 13 the same way:
7. Ole Miss (Team L)
8. Clemson (Team A)
9. Oregon (Team G)
10. UCLA (Team H)
11. Illinois (Team D)
12. USC (Team I)
13. North Carolina (Team B)
The biggest surprise? Ole Miss at No. 7. If I had the team names next to the stats, I probably would have placed Ole Miss around No. 10. After watching the Rebels against Auburn, LSU and Ole Miss, I have no faith in their defense to hold up enough to allow them to beat Alabama, Arkansas, Mississippi State and the SEC East champion. But their defensive stats are not as bad as I thought, and Clemson’s were not as good as I thought. Plus, Ole Miss has an elite offense and Clemson has a pedestrian one.
That said, I think it’s much more likely that Clemson goes undefeated and makes the CFP than Ole Miss goes 12-1 and makes the bracket. But after looking at these numbers, I have less faith in the Tigers to beat Notre Dame, Louisville, Miami, South Carolina and the Coastal Division champion (probably North Carolina) in consecutive weeks than I did before. Taken individually, Clemson should beat each of those teams. But it feels as if the Tigers aren’t playing with the same margin for error they had when they were making the CFP every year. Another game as sloppy as their Syracuse matchup could result in a loss.
But that’s why they play the games. Clemson could prove me wrong and wind up in the field.
The bigger question: Will this be a Lucky 13 next week? The Tennessee-Georgia loser probably stays on the list. But can everyone else?
(Photo: Eakin Howard / Getty Images)