College football FPI release: Top tiers, conference races and playoff hopes

Ohio State's 2022 ended in dramatic heartbreak in the College Football Playoff, but immediate redemption could be on the horizon. The Buckeyes are the best team in the nation and the favorites to win the national championship, according to ESPN's Football Power Index and the Allstate Playoff Predictor.

The Buckeyes have a 37% chance to win the title -- well ahead of Alabama at 20% and defending champion Georgia at 19% -- according to the models.

Before we dive too far into the Buckeyes, other title contenders and the rest of the forecasted college football landscape, a quick refresher: What is FPI, and how do we project the season and College Football Playoff race?

FPI is our season-long ratings and projections system. In the preseason it relies on past performance on offense and defense, returning and transfer production and past recruiting data for players on the roster to form a rating. We then use those ratings to simulate the season 20,000 times, resulting in our projections.

Full FPI ratings and projections can be found here.

In addition, the Allstate Playoff Predictor uses those simulations to forecast the playoff committee's selection process, based on the committee's past behavior. All of which combined allows us to forecast a team's projected win total, chance to win its conference, reach the CFP and win the national championship.

Let's break down some of the top storylines emerging out of FPI's numbers ahead of the 2023 season.

A three-team top tier

So what makes the Buckeyes so dangerous in 2023, even relative to SEC powerhouses Alabama and Georgia? For starters, the best non-QB offensive returning production in the nation. That's most notable in the receiving game, where the Buckeyes return Marvin Harrison Jr. and Emeka Egbuka -- who recorded 1,263 and 1,151 receiving yards in 2022, respectively -- along with tight end Cade Stover.

The model ranks Ohio State No. 1 in offensive performance in recent seasons (it ranked No. 1 in offensive efficiency in 2022) and second in offensive talent (based on cumulative recruiting ranks on that side of the ball). Quarterback aside (we'll get to that), we can see why FPI makes the Buckeyes the best offense in the nation by a decent margin. Defensively, the Buckeyes rank third -- they don't quite boast the talent levels that FPI sees in Georgia or Alabama -- but the margin is smaller and thus the Buckeyes are the best team overall. Georgia's defense ranks No. 1 in FPI rating, driven more by the historical performance of the unit than anything else.

What's also interesting here is just how far behind everyone else is from the top three. More than five points per game separates Georgia (No. 3) from LSU (No. 4) in FPI's ratings. Each of the top three teams has at least a 19% chance to win the national championship, while no other team is over 6%. And there is a cumulative 76% chance -- in April! -- that one of the three wins the national title.

In addition, there is a 26% chance that all three of the Buckeyes, Crimson Tide and Bulldogs make the playoff, and the top six most likely combinations feature all three of those teams.

In other words: It's a crystal-clear top tier. And what's somewhat amazing is this is happening in a year when none of the three are returning their starting quarterbacks. C.J. Stroud, Bryce Young and Stetson Bennett are all NFL-bound, and yet the model still has confidence in these squads.

It means that the model has less confidence in each of their QBs than it does for, say, USC with Caleb Williams. Though in the case of Ohio State, for example, the fact that Kyle McCord (the most likely starter, in the model's mind) had an 87 grade as a recruit gives the model some confidence in him despite the lack of track record.

A QB with a track record of success is ideal, but as evidenced by FPI's ratings for Ohio State, Alabama and Georgia, it's not a prerequisite.

Our annual 'Texas is back (or is it?)' conversation

Sitting at No. 5 in FPI's rankings and No. 4 in terms of chance to win the national championship: Texas. It's a somewhat familiar place for the model, though. FPI was bullish on the Longhorns starting last year (rank: No. 7) and it finished bullish on them, too (rank: No. 7).

But now it's really in on them, with a 34% chance to reach the playoff and its first top-5 preseason FPI rank since 2012. They've got strong talent, based on recruiting grades, on both sides of the ball (fourth on offense, sixth on defense). And, despite losing running back Bijan Robinson to the NFL, the Longhorns return quite a bit of production on offense, led by wide receiver Xavier Worthy.

Quinn Ewers, the incumbent at quarterback facing competition from Arch Manning and Maalik Murphy, didn't have a great season last year (65.6 QBR, which ranked 50th) but does have a year of experience. And, as we saw with the teams at the top, a good situation around a quarterback can yield a high FPI rating.

The Longhorns also benefit from their conference: There are fewer tough foes and no divisions in the Big 12, and that leads to a more straightforward path to a conference title than they will have in the SEC next season (Texas' 54% chance to win the Big 12 is higher than Georgia's 49% or Alabama's 41% chance to win the SEC). That doesn't mean Texas has an easy schedule, though. With a road game at Alabama on Texas' slate, the Longhorns have the toughest schedule of any of the FPI's top 10 teams.

USC leads a Pac-12 with a chance

A year ago at this time, FPI gave the Pac-12 a mere 8% chance to put a team into the playoff. We were artificially low on USC -- the Lincoln Riley effect wasn't fully accounted for -- but the point remained: The Pac-12 was a long shot. That ended up panning out, as the conference did not put a team in the playoff.

But the conference boasts the seventh-best team in the nation in USC, according to FPI. The Trojans alone have a 25% shot at the playoff, and the conference has a 34% chance overall at putting at least one team in the playoff thanks mostly to Oregon (5%) and Utah (4%). (This might be bittersweet for the conference, knowing it will soon lose USC to the Big Ten, but that's a next-year problem for the Pac-12).

With Williams -- who finished fifth in QBR last season and is a potential No. 1 overall pick in 2024 -- leading the way, USC enters the season with clearly the best quarterback room in the nation, according to FPI. That they have Williams back after he led the Trojans to a No. 3 rank in offensive efficiency only bolsters the model's confidence, and is part of why it considers USC to be the second-best offense entering the year only behind Ohio State.

The difference between USC and the top contenders mentioned earlier: The Trojans are elite only on one side of the ball. FPI sees them as the 36th-best defense, which is actually a huge forecasted improvement. The defensive unit ranked 82nd in defensive efficiency last season.

All of this gives USC a strong chance -- 50%, a true coin flip -- to win the Pac-12. And with an easier but not too easy schedule, a 1-in-4 shot at the playoff.

Slight pump of the brakes on Florida State

After finishing 10-3 with strong quarterback play from Jordan Travis (QBR: 85.3, which ranked seventh), the Seminoles are getting some heavy buzz. FPI likes Florida State ... but maybe not as much as everyone else. The Seminoles rank 14th in FPI, though they are just two-tenths of a point away from 12th.

The model is mostly buying the offense (rank: No. 10) but is less confident in the defense, which it ranks 26th. In general, Florida State has lower talent rankings (based on the recruiting ranks of players on the roster) than the teams above it in FPI. The Seminoles rank 17th in offensive talent and 24th in defensive talent. There's a lot to like here -- just not quite as a top-10 team yet.

It's in part due to the talent gap that FPI still likes Clemson (FPI rank: eighth) more in the ACC. The Tigers, who have higher talent ratings on both sides of the ball, have a 45% chance to win the ACC, while the model gives Florida State just a 17% chance.

FPI not buying Colorado just yet

Considering the hype around Colorado with Deion Sanders now the Buffaloes' head coach, as well as a slew of transfers in to play for him, it's a bit of a shock to see Colorado all the way down at ... 95th in FPI's rankings -- the lowest-ranked Power 5 team.

The model has somewhat of a handle on the transfers. They're recognized in Colorado's "talent" and returning production portions of the calculation, and the Buffaloes do have higher talent scores on both offense and defense than virtually all of the teams around them in the overall rankings. From the model's perspective, though, that isn't enough to overcome the recent poor play. The Buffaloes ranked fourth worst and 11th worst in offensive and defensive efficiency last season. So FPI actually is predicting a decent step up this year.

That being said, 95th is still rough. It's probably reasonable to think this is an unusual circumstance of roster turnover -- one-win teams don't usually transfer in high-end players -- that FPI might not be able to perfectly capture.

Conference playoff race

The SEC and Big Ten are each extremely likely (97% and 94%, respectively) to put a team in the playoff. But what about two teams, like the Big Ten had last year?

The SEC is the favorite to pull off that maneuver in 2023 with a 51% chance to put multiple teams in the playoff, while the Big Ten is at 25% (every other conference is under 1%).

The SEC has two major things going for it:

• While Ohio State is the best team in the nation, FPI believes the SEC boasts teams Nos. 2, 3 and 4 in Alabama, Georgia and LSU. By contrast, Michigan and Penn State -- the Big Ten's next-best teams -- come in at No. 6 and No. 10.

• The conference's top two teams are in separate divisions and aren't scheduled to play each other. That means Alabama and Georgia won't beat up on each other until the conference championship game at the latest, while only one of Ohio State, Michigan or Penn State can reach the Big Ten championship game.

Lauren Poe and Mitchell Wesson contributed to this article.