College Football 2023 Simple Talent Tiers
A simple model for estimating college football success and Jim Harbaugh is the best coach in college football, but how long can he keep making up for Michigan's talent deficit?
There’s an old saying in football that it’s not the Xs and Os that matter, but the Jimmys and the Joes. Meaning that players matter a lot. And having the right athletes on your team can overcome differences in game plan.
I believe that less in the NFL, but it is certainly true at least a little bit in all sports. And it is certainly true a great deal in college football. If you watched last week’s opener—Notre Dame vs. Navy—you can’t think any different. If you watch any of the Big State University versus Regional College games this weekend, you’ll come away with the same impression.
So when I think about college football, that’s where I start: talent and recruiting. And more importantly, how we can use knowledge of recruiting success to predict on-the-field success.
Of course, this won’t be a perfect predictor. Other things matter too. Coaching matters. Player development matters. Team composition matters. Injuries, academic eligibility, interpersonal relationships between players, and the student section all matter. But talent matters a lot.
And so we’re going to measure it and use it to predict how good college football teams will be this year.
College Football Simple Talent Tiers
Which brings us to the Simple Talent Tiers: a model for predicting how college football teams will finish, given the talent they have brought into their programs in the last four years.
The goals of this model are two fold:
Be informative
Keep things simple
The first objective is self-evident: we want the rankings to provide some real value. And this model is reasonably effective at that. More than 70% of the teams in a given tier will finish above the ranking threshold that defines that tier.
That may seem like a low number, but consider that 50% of teams ranked at the start of the season drop out of the rankings by the end of the season. College football is a sport where a single loss—caused by a single errant throw, bounce, or bobble of the football—can eliminate a team from serious national title contention. The margins are narrow.
And the second objective—keep things simple—stems from that last point. Because college football is chaotic and small hiccups can cause big messes, we want a model that is simple. A simple model will be less prone to overreacting… less prone to reacting altogether.
How does this all work?
Which raises the question, how does the model work?
The Simple Talent Tiers model considers the last four seasons of recruiting rankings for each team and awards credit for the following things:
Number of Top 20 recruiting classes
Average recruiting class rank
Best recruiting class rank
For example, Notre Dame’s last four recruiting classes were ranked 11th, 6th, 9th, and 22nd. They would get credit for three top-20 classes (‘23, ‘22, ‘21), a 12th place average recruiting ranking, and a top recruiting ranking of 6. Those three numbers—3, 12, and 6—go into the model and produce an estimated end-of-season ranking for Notre Dame.
Those ratings are then used to produce tiers, based on whether or not they exceed specific end-of-season ranking-expectation thresholds.
Tier 1 is expected to be a Top 5 team.
Tier 2 is expected to be a Top 10 team.
Tier 3 is expected to be a Top 20 team.
Tier 4 is expected to be a Top 25 team.
We chose to focus on end of season ranking because end of season ranking matters more than wins and losses. A 10-2 that makes the College Football Playoff out of the SEC will be happier than a 11-1 team that misses from the Mountain West.
Only 16 top 25 teams?
But wait, this season we’re only ranking 16 teams—what gives?
The model is designed to predict the finish of teams based on their talent. And not all the teams that finish in the Top 25 do not have “Top 25 talent”. Consider the Boise States, UCFs, and Cincinnatis of the world. These teams have good seasons—but even in their best seasons, do they have the talent compete with the elite teams?
The final ratings last year include Tulane, Troy, and the University of Texas San Antonio. Quality institutions, but not ones whose football programs bring in top 25 talent. Combined, those three schools have fewer players in the NFL than Stanford.
And while The Simple Talent Tiers model doesn’t directly account for this, it certainly recognizes the fact.
So without further ado… let’s review the College Football Simple Talent Tiers.
Tier 1: Expected top 5 finish
We begin in tier 1, with three teams: Alabama, Georgia, and Ohio State. These teams are all expected to finish in the top 5. I dare say, we’re not adding too much value here. You have to go back to 2019 to find a season in which at least 2 of these teams didn’t make the college football playoff.
That said, both Georgia and Alabama have had down years in the recent past. In 2020, a two-loss Georgia was ranked 9th. And in 2019 a two-loss Alabama was ranked 13th.
You have to go back to before the College Football Playoff to find the last time Ohio State finished outside the top 10. In 2013, an undefeated Ohio State lost their last two games of the season—the Big 10 championship and the Orange Bowl—to Michigan State and Clemson.
Notable: Georgia is the unanimous top team and pre-emptive favorite. Coming off back-to-back National Championships, the press, Vegas, and our model all have them as top of the pack. But the road to the title is tricky and only one team (2020 Alabama) in the College Football Playoff era has started and finished the year ranked #1.
Tier 2: Expected top 10 finish
Next up is Tier 2. We expect teams in Tier 2 to finish within the top 10, based on the talent on their roster. Clemson, LSU, and Texas A&M all find their home in this tier.
LSU and Clemson are both establishment picks. Both are ranked in the top 10 by the AP and Vegas, and have are a bit better than +350 to make the College Football Playoff.
Clemson’s playoff odds seem a bit low, because other than November games against Notre Dame and UNC, they don’t face other top-recruiting powers. Vegas will have them as a favorite for every regular season game they play this year. But unless Notre Dame plays impressively, they may need to run the gauntlet undefeated to be considered.
Brian Kelly’s LSU is a consensus choice for outside-looking-in team. At 5, the AP and Vegas have them just missing the playoff. Of course, all eyes will be on their yearly contest with Alabama.
Rounding out tier 3 is Texas A&M, coached by Jimbo Fisher. Fisher is an excellent recruiter. The talent ratings bear out as much. But A&M is notorious as an also-ran school, always playing second fiddle to Texas, and Fisher has a reputation as a bad culture guy. His left Florida State in 2017 after a 3 win campaign through the weaker ACC, while on the brink of sanctions from the NCAA. The Aggies won only 5 games last year. Can Fisher right the ship? A&M has the talent, but not the confidence of the media or oddsmakers. Vegas has Texas A&M at +1300 to make the playoff.
Tier 3: Expected top 20 finish
Tier 3 is the deepest tier and it spans twice the distance as our other tiers. Tier 3 teams are expected to finish in the top 20 based on their talent. It includes six teams: Michigan, Texas, Oregon, Notre Dame, Oklahoma, Florida.
Of these, Michigan is the biggest story. A classic Michigan Man, Jim Harbaugh has coached Michigan to 6 top 20 finishes in his 8 seasons as head coach. And the last two seasons Michigan has been knocking on the national championship door: beating rival Ohio State, winning the Big 10 and making the College Football Playoff both years.
With their talent—or relative lack thereof—this is a remarkable accomplishment. And Michigan returns 81% of their productivity this year, according to ESPN. Both the AP and Vegas have them poised for a big season this year and ranked #2 going into the season. Vegas, however, has Michigan at the 3rd best odds to make the college football playoff—despite being narrow favorites over Ohio State for their end-of-season showdown.
Texas and Oklahoma bring the Big 12 into the picture in Tier 3. Many are optimistic about Texas’ recent recruiting and transfer-portal success—surely tied in part to the rise of NIL and pay-to-play, and in part to tortured-but-talented head coach and former Saban assistant Steve Sarkisian.
Oregon, the only Pac 12 team in our ratings, and the always noteworthy Notre Dame, also feature in Tier 3—in agreement with the press and Vegas. While Ole Miss has the toughest schedule in college football—playing Alabama, Georgia, and LSU—Notre Dame has playoff-relevant matchups against USC, Ohio State, and Clemson. Depending on how the season plays out, the Irish could play kingmaker or spoiler for any of those schools.
Tier 4: Expected top 25 finish
And finally, the last four in: Penn State, Tennessee, North Carolina, and Miami. We expect all of these schools to finish the year in the top 25, based on their talent. That means at most 4 losses; 3 or fewer if you’re North Carolina or Miami.
Penn State will have their hands full, as always, with Ohio State and Michigan—but will be favored in their other games.
Tennessee has matchups against Alabama and Georgia to look forward to, and needs to watch out against Texas A&M and Florida, who also make our list.
And the ACC teams—UNC and Miami—just need to live up to their talent in a conference that is so often full of disappointing or inconsistent teams.
2023 College Football Bets
If Clemson’s road to undefeated is blocked by conference rival Florida State, and Notre Dame and North Carolina—both of whom are rated in our Simple Talent Tiers. But they will be favorites in all of these matchups. Further, all of these matchups are home games for the Tigers. If you’re optimistic about Sophomore quarterback Cade Klubnik, there’s good reason to like Clemson’s +320 to make the playoff, or even +1600 to win the National Championship.
In part, betting Clemson is a short on Florida State—who is ranked highly by the AP and Vegas, but has relatively little talent on their roster. FSU is a media darling because of they had a competitive season last year and return more than 80% of their players. But they’re only two-years removed from a 5-7 season, and their 34-28 loss to Clemson last year was not as close as advertised.
Ohio State, +175 to win the Big 10, is another interesting target. They have a talent advantage against all their opponents, but opponents Notre Dame (Sep 23), Penn State (Oct 21), and Michigan (Nov 25) are no schlubs. Further, the Michigan game is at the Big House this year. If played today, Ohio State would be a touchdown underdog in that game.
This bet would be a bet on talent of OSU over the coaching and culture of Harbaugh and Michigan, who return 84% of their offense. But when else are you getting near 2:1 odds one of the top programs in the country?
Is Texas back? If you like the Longhorns, Texas +320 to make the playoff seems like a good bet. The Big 12 is not a massively talented conference and Texas returns 85% of its offensive starters, including elite recruit and former Ohio State quarterback Quinn Ewers. Rival Oklahoma, meanwhile, is starting a 22-year old Central Florida transfer who had an underwhelming showing last year, his true senior year.
Lastly, Fan Duel has North Carolina projected to win 8.5 games. There is give on the over (+122), and Clemson is the only ranked team the UNC plays all season. Even if the Tar Heels lose trips to South Carolina and NC State, in addition to the Clemson game, they can still come in above the 8.5 wins. Apart from Clemson, no teams on North Carolina’s schedule make our Talent Tiers.