What the CFP's new seeding means and how it would have affected the 2024 bracket

play

After months of meeting to discuss things to discuss at future meetings, the people in charge of the College Football Playoff actually made a decision on Thursday, and it was one we've assumed they'd make for a while. After last year's 12-team CFP gave byes to the four most highly ranked conference champions, this year's will not.

Conference commissioners voted to go to a straight seeding format (with five spots still reserved for conference champions) in 2025.

There are still plenty of things to discuss regarding what the CFP will look like in 2026 and beyond -- and good lord, don't even get me started on how much I don't like where we're probably headed in that regard -- but with the 2025 season starting in less than 100 days, we at least know how things will take shape this fall. Here are a few thoughts regarding these changes.

A 2024 simulation

To see what something might look like in the future, my first step is always to revisit the past. Last year's 12-teamer, the first-ever genuine tournament at the highest level of college football, indeed handed out byes to conference champions and gave us the weird visual of having two different numbers listed next to the teams in the bracket.

Boise State, for instance, was ranked ninth in the overall CFP rankings, but the Broncos got the No. 3 seed as the third-ranked conference champ. Arizona State was simultaneously 12th and fourth. Granted, the NFL does something similar, giving the top three seeds in each conference to the winners of each individual division (which occasionally gives us odd pairings such as 9-8 Tampa Bay hosting 11-6 Philadelphia in 2023 or the 10-7 Los Angeles Rams hosting 14-3 Minnesota in 2024). But from the start, it was clear there was some dissatisfaction with this approach. And when both BSU and ASU lost in the quarterfinals -- all four conference champions did, actually -- it became abundantly clear that this was going to change. It just took about five months to actually happen.

Regardless, let's look at how the 2024 playoff would have taken shape with straight seeding instead of conference-champ byes.

First round

12 Clemson at 5 Notre Dame (SP+ projection: Irish by 13.1, 79.4% win probability) 11 Arizona State at 6 Ohio State (OSU by 24.2*, 93.6% win probability) 10 SMU at 7 Tennessee (Tennessee by 7.0, 66.9% win probability)

9 Boise State at 8 Indiana (Indiana by 12.5, 78.3% win probability)

(* Here's your reminder that SP+ really didn't trust Arizona State much last season, primarily because the Sun Devils were a pretty average team early in the season. At 5-2 with a number of close wins and a sketchy-looking loss at Cincinnati without injured quarterback Sam Leavitt, they entered November ranked in the 50s. While they certainly rose during their late-season hot streak, they finished the season only 35th. They were genuinely excellent late in the season -- just ask Texas -- but they were 6-1 in one-score games heading into the CFP, and they were lucky to reach November with the Big 12 title still within reach.)

In last year's actual first round, the four home teams (Notre Dame, Ohio State, Penn State and Texas) were projected as favorites by an average of 7.2 points per SP+. The average spread was Home Team -8.9. The results were actually much more lopsided than that, and that probably wouldn't be any different with the matchups above -- here, home teams are projected favorites by an average of 14.2. Changing to straight seeding wouldn't have made the first round more competitive.

Assuming all four home teams win in this simulation, that gives us the following quarterfinals.

Quarterfinals

Rose Bowl: 1 Oregon vs. 8 Indiana (SP+ projection: Oregon by 5.9, 64.4% win probability) Fiesta Bowl: 4 Penn State vs. 5 Notre Dame (PSU by 0.7, 51.8% win probability) Sugar Bowl: 3 Texas vs. 6 Ohio State (OSU by 7.1, 67.1% win probability)

Peach Bowl: 2 Georgia vs. 7 Tennessee (UGA by 2.4, 55.9% win probability)

Interestingly enough, we got two of these four matchups in real life, but they were the two semifinals -- Ohio State's 28-14 win over Texas in the Cotton Bowl and Notre Dame's late 27-24 win over Penn State in the Orange Bowl. Now these games take place in New Orleans and Glendale, Arizona, respectively. We'll conveniently project those results to remain the same. Meanwhile, SP+ says there's only about a 36% chance that the other two projected favorites (Oregon and Georgia) both win, but we'll roll with that.

Semifinals

Cotton Bowl: 1 Oregon vs. 5 Notre Dame (SP+ projection: Oregon by 2.1, 55.3% win probability) Orange Bowl: 2 Georgia vs. 6 Ohio State (OSU by 6.8, 66.6% win probability)

With those win probabilities, there's only about a 37% chance that both projected favorites win, and this time we'll heed that and project an upset: Conveniently, we'll say Notre Dame upsets Oregon, giving us the exact same Fighting Irish-Buckeyes title game we got in real life.

Final

5 Notre Dame vs. 6 Ohio State

Again, we saw this one.

Who would have benefited from this change?

In all, using my pre-CFP SP+ projections from December, here's a comparison of what each team's national title odds were heading into the tournament versus what they'd have looked like with straight seeding.

Not surprisingly, Arizona State's and Boise State's odds would have sunk without receiving a bye, but their title odds were minimal regardless. The teams that actually ended up hurt the most by the change would have been 2-seed Georgia, original 5-seed Texas and original 11-seed SMU. The main reason for the downshift in odds? They'd have all been placed on Ohio State's side of the bracket. Meanwhile, Ohio State's and Tennessee's odds would have benefited from the simple fact that they would no longer be paired with unbeaten No. 1 Oregon in the Rose Bowl quarterfinal. Obviously Ohio State handled that challenge with aplomb, but the Buckeyes still had to ace that test, then win two more games to take the title.

Beyond Ohio State and Tennessee, both Indiana and Oregon would have seen their title odds improve a bit with straight seeding, though for different reasons. Indiana would have gotten a first-round home game instead of having to travel to South Bend, while Oregon would have avoided Ohio State until a potential finals matchup.

Takeaways

Good: The No. 5 seed isn't quite as uniquely valuable now

We never got to see the 12-team playoff as originally envisioned, with six conference champions earning bids from a universe that featured five power conferences. Instead, between the announced adoption of the 12-team playoff and its actual arrival, the SEC officially added Oklahoma and Texas to its roster while the Big Ten, with help from the Big 12, cannibalized the Pac-12. With only four power conferences remaining, we ended up with only five conference champions guaranteed entry, and with the distribution of power getting further consolidated (we still have four power conferences, but it's clearly a Power Two and Other Two), that left us with an awkward bracket.

For starters, the new power distribution meant that the No. 5 seed -- almost certainly the higher-ranked team between the losers of the Big Ten and SEC championship games -- would get an almost unfair advantage. As I wrote back in December, "the odds are pretty good that the teams earning the No. 4 and 12 seeds (aka the two lowest-ranked conference champs) will be the weakest teams in the field .... Texas, the top-ranked non-champion and 5-seed, is indeed pitted against what SP+ thinks are the No. 17 and No. 30 teams in the country and therefore has excellent odds of reaching the semifinals."

As you see above, Texas actually entered the CFP with better title odds (17.2%) than Georgia (16.6%), a higher-ranked team in SP+ and the team that had just defeated the Longhorns in the SEC title game. In theory, giving a team a bye and asking them to win three games instead of four would be a massive advantage. But in practice Texas' odds of winning two games (against Clemson and ASU) were better than Georgia's odds of winning one (Notre Dame). That's not particularly fair, is it?

Bad: Conference title games mean even less now

Making this change would have indeed given the SEC champion better title odds than the SEC runner-up. That's good, but it comes with a cost. In the re-simulation above, you'll notice that both the winners and losers of the SEC and Big Ten title games ended up with byes and top-four seeds. That means there were almost literally no stakes -- besides a quest to avoid major injuries like what afflicted Georgia -- in either game.

Meanwhile, in the ACC championship, SMU lost to Clemson but barely fell in the CFP rankings (and, more specifically, still got in) because the playoff committee didn't want to punish the Mustangs for playing a 13th game while others around them in the rankings were already done at 12. Add to that the fact that the straight-seeding approach diminished the above title odds for four of the five conference champions in the field, and it leads you toward a pretty easy question: Why are we even playing these games?

Commissioners of the power conferences have pretty clearly had that in their minds as they've discussed a convoluted (and, in my own opinion, patently ridiculous) new playoff structure that hands multiple automatic bids to each of the top four conferences: up to four each for the SEC and Big Ten and likely two each for the ACC and Big 12. With this structure in place, they can drift from title games and toward multiple play-in games within each conference. I absolutely hate this idea -- if you want to wreck the integrity of the regular season, nothing would do that faster than a 7-5 or 8-4 Big Ten team potentially stealing a bid from a 10-2 or 11-1 comrade that was vastly superior in the regular season -- but you can at least understand why the commissioners themselves, facing a world with diminished conference title games (and always looking for more TV spectacles), would try to get creative in this regard.

Straight seeding doesn't change all that much. Ohio State was given a harder title path last year than would have existed with straight seeding, but the Buckeyes cruised regardless, winning four games by a combined 70 points. Meanwhile, even with a bye, Boise State and Arizona State weren't likely to win three games and go all the way. The team that best peaks in December and January will win 2025's title just like it did in 2024, we'll enjoy ourselves all the same, and we'll be facing another change in 2026 no matter what.

The countdown toward 2025 continues.