Against All Odds: 2025 Championship Weekend Recap

I have spent a lot of time over the last 15 weeks of college football trying to predict how each conference will shake out and how that will ultimately result in the field for the second ever 12-team college football playoff.

After 879 regular season games, nine conference championship games, and one controversial bracket reveal, we now have an answer to all of these questions. Champions were crowned, bowl spots have been filled, and the bracket is now set.

Did the committee get it right? Who will win it all? Did my computer provide any good advice last week?

Let's explore all of those questions, starting with the last one.

Championship Weekend Bad Betting Results

Let's start by taking a look at the overview summary shown below in Figure 1 which compares the actual result each all nine games in Championship Week with the opening Vegas spread.

Figure 1: Results of Championship Week showing. the actual point differentials relative to the opening spread.

The figure shows that three of the conference champions exceeded expectation by covering the spread by more than 14 points: Texas Tech, Georgia, and Boise State. MAC Champion Western Michigan covered the spread against Miami of Ohio while James Madison did not cover the spread in the Sun Belt Championship Game against Troy.

The underdogs won in the other four conference title games, including Indiana (+5.5) over Ohio State in the Big Ten Championship Game, Tulane over North Texas (-3) in the American Athletic Conference Championship Game, and Duke over Virginia (-2.5) in the ACC Championship Game. Table 1 below summarizes the upsets for the weekend and compares the results to the computers' predictions.

Table 1: Upsets in Championship Week based on the opening Vegas line compared to the upset projections from last week.

My computer got both upset picks correct, which brings the year-to-date record to 42-48 (47%). ESPN's Football Power Index went 2-1 to bring its record on upsets to 22-31 (42%) for the year.

There were no recommended bets against the spread this week, which leaves the record for my highlighted picks at 32-29 (52.5%). My recommendation based on an analysis of the FPI data remains at 20-16 (56%).

Overall my computer went 6-3 (67%) while the FPI went 5-4 (56%). My machine now has a record of 351-343 (50.6%) while the FPI is sitting at 331-363 (48%).

There was one recommended point-total bet and North Texas and Tulane obliged by scoring fewer than 67.5 points. This correct pick bumped up the record for selected picks to 83-89 (48%) for the season while the lock picks remain at a disappointing 26-31 (46%).

Did the Committee Get it Right?

No. No, they did not.

After a good effort in assembling last year's inaugural 12-team bracket, the committee made multiple, egregious errors this year.

The aspect of the field that is drawing the most attention is the inclusion of Miami at the expense of Notre Dame, and to a lessor extent, the inclusion of both Tulane and James Madison. There are at least two ways to look at this decision.

The first way is to consider the broad question of whether the committee picked the correct set of teams overall.

Regarding the inclusion of James Madison and Tulane, I am completely fine with those two teams making the bracket. The inclusion of the top five highest ranked conference champions as automatic qualifiers is one of the only rules in the current playoffs that actually makes sense.

The fact that the ACC is a weak conference that wound up with a champion in Duke with an 8-5 record is certainly not Tulane or James Madison's problem.

Much like the NCAA Basketball Tournament, the inclusion of "mid-majors" makes for a better, more fair event where all FBS teams have a chance to win the National Title. I would much rather have James Madison in the bracket than the sixth-best SEC team.

As for the top 10 power conference teams in the bracket, the question of inclusion is basically a judgment call to decide on the 10 "best" college football teams. But just defining "best" is not straightforward.

One definition of "best" is based on which 10 teams are the most likely to win the National Title. Practically, this boils down to identifying the top 10 teams that would be favored in Vegas over all remaining teams. This can be estimated based on power rankings such as the one that I have developed, ESPN's Football Power Index, and Bill Connelly's SP+.

Another definition of "best" is which are the 10 most deserving teams based on actual wins and loses over the schedule that they have played. Practically, this is what "strength of record" or "strength of resume" metrics attempt to measure.

The committee members are given various versions on these metrics for each team and must find a way to blend them together into a single ranking. The "blending" activity is a combination of math, science, and art.

For what it is worth, I have developed my own power and resume rankings and have determined the optimal way to "blend" them together in order to get a ranking that approximates the rankings made by past college football committees.

When I applied my methodology to this year's list of candidate teams, there are seven teams that finish in the top ten of both my power rankings and my strength of resume rankings. Those teams are Indiana, Ohio State, Oregon, Georgia, Texas Tech, Ole Miss, and Oklahoma.

If I focus on strength of resume, the remaining three most deserving teams are BYU, Texas A&M, and Alabama. But if I consider power rankings, the remaining three strongest teams are Notre Dame, Miami, and Utah.

When I apply my blended metric to generate my simulated college football rankings, Alabama and Texas A&M both make the field. The final team that I would have included is actually neither Notre Dame nor Miami. I would have put BYU into the field over both the Hurricanes and the Fighting Irish.

As for Miami and Notre Dame, the two teams are basically in a dead heat in my rankings. I would likely have put Notre Dame into the playoffs over Miami, but it is mathematically extremely close. A coin flip would have been an equally valid way to make the decision.

The second way to think about how the bracket was assembled is to consider the validity of some of the procedures used by the committee. This is where the committee dropped the ball the most. More precisely, this is where the individuals who wrote the rules governing the selection and seeding process failed the worst.

Notre Dame had been ranked ahead of Miami in every single ranking issued previously by the committee. But a loss by BYU in the Big 12 Championship Game created a situation where Miami and Notre Dame were ranked in sequence at No. 10 and No. 11.

With the two teams now ranked next to other, the guidelines dictate that head-to-head results come into play. Miami defeated Notre Dame in Week One by three points at home, and for this reason the committee bumped Miami ahead of the Irish in the final rankings despite neither team playing a game in the final weekend.

As discussed above, I do not have a strong opinion about whether Notre Dame or Miami should have made the field. But the overall situation that resulted in Notre Dame slipping below Miami when neither team played is ludicrous. If the committee truly believed that Miami is the better and more deserving team overall, they should have ranked them appropriately in the first place.

The college football playoff committee needs to take notes from the NCAA basketball tournament, which just so happens to be the most popular and successful bracketed tournament in the history of American sports.

The fact that the football committee is obligated to release a ranking in the weeks leading up to the final bracket is the root cause behind the majority of the angst surrounding the Notre Dame versus Miami debate. This is a situation more transparency does more harm than good.

More importantly, the seeding principles of the basketball tournament are designed to minimize the number of regular season rematches as one of the most important bracketing principles.

By contrast, the guidelines of the college football playoff specifically prohibit the shuffling of teams in the rankings for the purpose of avoiding rematches. The fact that the football playoff specifically chooses to ignore this best practice is both bizarre and, frankly, stupid.

The "logic" behind this rule is that moving teams around on the bracket would potential create a competitive imbalance. As we will see later, this concern is completely unfounded. As discussed above, the process used to rank the teams is not purely based on evaluating the strength of each candidate team. Regardless, the committee has shown no ability to accurately estimate the true strength of each team.

As a result of this flawed policy, this year's college football playoff bracket is filled with numerous, easily avoidable rematches that result in a lower quality tournament.

First, the committee matched up No. 8 Oklahoma with No. 9 Alabama in the first round. Oklahoma already played Alabama back on November 15 in Tuscaloosa and escaped with a narrow two-point victory. Both the Sooners and the Crimson Tide deserve to play a new opponent to open the playoffs. This placement is an egregious error in bracket construction.

The easy fix to this problem would be to swap Alabama for Miami. This would create a matchup of Alabama and Texas A&M in the first round. While the two teams are currently both in the SEC, they have not faced each other since 2023. This obviously superior solution would have also worked if Notre Dame would have made the playoffs instead of Miami.

But that is not the only first-round rematch. No. 6 Ole Miss will face No. 11 Tulane in the first round as well. Those two teams met on Sept 20 in Oxford and the Rebels won 45-10. In what universe does it make sense to force them to play each other again in the playoffs?

To make matters worse, if Ole Miss wins in the first round, the Rebels would next face No. 3 Georgia in the next round. Georgia already beat Ole Miss 43-35 at home on Oct. 18 of this year. Once again, why should these players and fans have to experience and watch this movie again?

There are a few potential ways to solve this rematch problem, including swapping Tulane and James Madison or swapping Georgia and Ohio State. But the best solution appears to be a swap between No. 5 Oregon and No. 6 Ole Miss.

Making the two proposed swaps would eliminate all potential rematches until the third round of playoffs.

There is also one, more minor quibble that I have with the final playoff bracket, and that is in the location of the four quarterfinal games.

As the Big Ten Champion, Indiana earned the right to go to the Rose Bowl while Georgia similarly earned the right to play in the Sugar Bowl. But the location of the other two games has No. 2 Ohio State assigned to the Cotton Bowl while No. 4 Texas Tech was assigned to the Orange Bowl.

For the Buckeyes, the travel distance to Miami or Dallas is roughly the same. Texas Tech would rather play in the Cotton Bowl as would the most likely opponent Oregon due to proximity. But the committee decided to reverse that assignment to avoid Ohio State potentially having to play Miami in their home stadium.

This logic makes sense, but it only reinforces the benefit of swapping Alabama for Miami in the No. 9/No. 10 seed slots. This swap would solve both the rematch problem and the venue problem in one fell swoop. The result would be a better overall tournament.

If only the committee were wise enough to have made these simple adjustments.

College Football Playoff Analysis

Despite the fact that multiple, correctable errors were made in the construction of the 2025-26 college football playoff bracket, it is the bracket that we have.

So the analysis shifts to how the actual tournament will shake out. Figure 2 below shows the odds for each team to win the National Title based on my post-championship weekend power rankings.

Note that the data discussed in the opening section above confirms that my algorithm has outperformed ESPN's FPI this year and is over 50% overall against the spread. Therefore, the data shown below is likely a bit more accurate than either the overall odds or the Vegas money lines.

I should also note that last year my analysis correctly predicted the result of all 10 playoffs games prior to the start of the playoff. This included the prediction that No. 8 seed Ohio State would defeat No. 7 Notre Dame in the National Title Game.

Figure 2: Odds for each team to win the National Title based on Dr. Green and White's final power rankings

This year's analysis suggests that despite the results of the Big Ten Championship game, No. 2 Ohio State has the best odds to repeat as National Champions (38%). No. 1 Indiana (28%) has the second-best odds.

Interestingly, No. 4 Texas Tech (15%) and No. 5 Oregon (6.7%) have the next best odds. Despite having a bye, No. 3 Georgia (6.2%) has only the fifth best odds to win the tournament.

There is a large gap in the odds between Georgia and the next team, which is actually No. 10 Miami (2.4%). No. 8 Oklahoma (1.8%) and No. 6 Ole Miss (1.4%) are next in line.

Rounding out the field are the longshots No. 7 Texas A&M (0.28%) and No. 9 Alabama (0.21%) as well as the two extreme longshots from the Group of Five: No. 11 Tulane (1-in-6,700 odds) and No. 12 James Madison (1-in-16,000 odds).

As a general rule the odds generated in Figure 1 will mirror the power rankings used to make the calculation. My algorithm has both Texas Tech and Oregon ranked ahead of Georgia. It also has Miami ranked ahead of the four other SEC teams.

Another way to analysis the 2025-26 College Football Playoff bracket is to calculate the strength of the draw for each team. Mathematically, this is equal to the odds for an average playoff team to win the National Title if that reference team were placed in each of the 12 positions on the bracket.

The results of this calculation are shown below in Figure 3. For comparison, I included a reference calculation of the strength of draw for this year's tournament if the teams were seeded based on my power rankings.

Figure 3: Strength of draw for all 12 teams in the 2025-26 College Football Playoffs.

A few conclusions can be drawn from the results shown in Figures 2 and 3 above. The headline conclusion is that more likely than not (66%) a Big Ten team is going to win the National Title. In fact, there is over a 30% chance that the National Title Game will be a rematch of the Big Ten Championship Game between Indiana and Ohio State.

A secondary conclusion is that the bracket appears to be slightly unbalanced. My analysis shows that three of the strongest four teams in the tournament (Indiana, Texas Tech, and Oregon) are all located on the "top" half of the bracket. As a result, all three teams have a slightly harder path to the title than the teams on the "bottom" half of the bracket.

This is good news for Ohio State and to a lessor extent for Ole Miss, and Miami. As Figure 3 shows, these three teams on the bottom half of the bracket have a slightly easier path in the tournament than they would have if the teams were seeded based on actual strength. The primary driver of this imbalance is that No. 3 Georgia appears to be slightly overrated.

As a result, the odds for the Buckeyes are a few percentage points better in the current bracket than they would be if Ohio State were to be the No. 1 seed in a playoff where the teams are seeded based on my power rankings.

This lack of balance, despite the committee's strict reliance on their own rankings to fix the bracket, only highlights the ridiculousness of the guideline against shuffling teams to minimize rematches.

Game-By-Game Picks

Unlike last year, my computer is projecting a fairly "chalky" result in the 2025-26 playoff. In the first round, it only predicts one upset based on seeding: No. 10 Miami (-3, projected) over No. 7 Texas A&M. My algorithm also likes No. 8 Oklahoma (-7) to beat No. 9 Alabama despite the fact that the Crimson Tide are a 1.5-point favorite in Vegas.

My computer is forecasting a big win by No. 5 Oregon (-20) over No. 12 James Madison and by No. 6 Ole Miss (-14) over No. 11 Tulane. That said, my math suggests that both Group of Five teams will cover the current Vegas spread of 17.5 points and 21.5 points, respectively.

In second round, my computer has the top four seeds winning all four contests. My computer would have No. 1 Indiana favored by 10-points over No. 8 Oklahoma. No. 2 Ohio State would also be a 10-point favorite over No. 10 Miami.

No. 3 Georgia is projected to be a four-point favorite over No. 6 Ole Miss. No. 4 Texas Tech is projected as a three-point favorite over No. 5 Oregon.

In the national semifinal, my computer has No. 1 Indiana as a three-point favorite over No. 4 Texas Tech. No. 2 Ohio State would be a 9.5-point favorite over No. 3 Georgia.

If this were all to come to pass, my computer has No. 2 Ohio State as a 1.5-point favorite to get revenge over No. 1 Indiana for the National Title.

As a final note, from a betting point of view there are three teams in the playoffs that have a current money line that I project to have a positive return on investment:

Texas Tech (+850), ROI = 40.5%
Ohio State (+230), ROI = 27%
Indiana (+270), ROI = 3.2%