Over the past few weeks, this series has covered a lot of ground. We started by looking back at 2025, then took a close look at Michigan State’s schedule before projecting the winners of all ten Football Bowl Subdivision conferences.

Now it is time to put all the pieces together and build a 12-team College Football Playoff bracket.

Last Year's Picks in Review

Now that we have mapped out the outcomes for all 10 FBS conferences in 2026, let's take a quick look back at how accurate last year's predictions were using the same methodology.

Table 1 shows my picks for the 2025 conference title game participants using both the Most Likely and Disruptive scenarios. The correct picks are shaded in yellow and listed on the right side of the table. The preseason odds for both the projected and actual winners are also included for reference.

Table 1: Summary of last year’s picks for the conference title game participants using both the Most Likely and Disruptive scenarios. The preseason odds for both the projected and actual winners are provided, with correct picks shaded in yellow.

The same data for the winners of each conference are given in Table 2.

Table 2: Summary of last year’s conference champion picks using both the Most Likely and Disruptive scenarios. The preseason odds for both the projected and actual winners are provided, with correct picks shaded in yellow.

For conference title game participants, the odds-based Most Likely Scenario correctly identified six of the possible 18 teams. The Disruptive Scenario produced four correct picks.

The Disruptive Scenario correctly picked Tulane, BYU, Boise State, and James Madison to make their respective conference championship games. BYU was the only unique pick made by the Disruptive Scenario that ended up being correct. The Most Likely Scenario correctly picked Ohio State, UNLV, and Alabama to make their respective conference championship games.

On the conference level, both scenarios got three picks correct out of nine champions and the two scenarios agreed on all three of those picks. Both scenarios correctly picked Boise State to win the Mountain West, Tulane to win the American, and James Madison to win the Sun Belt.

While that success rate may seem low, the sum of the odds in each table gives the preseason expected value for the predictions, and those numbers landed well within expectations. My analysis expected to get 7.3 title game participants correct and identified six. At the conference champion level, 2.6 correct picks were expected, and each model produced three.

As a final note, Table 1 also shows that while the odds-based lens suggests teams with at least a 30% chance to make their respective conference championship games are the safest bets, the reality is often much messier.

Of the 18 conference title game participants, five teams had preseason odds below 10%, and five more had odds between 10% and 20.1%. This year will likely play out similarly.

Back in August of 2025, no one predicted that Kennesaw State (3.1%), Virginia (4.4%), Western Michigan (6.7%), Duke (7.6%), or Indiana (9.9%) would make their respective conference title games. But they all did. College football, as always, is full of surprises.

Review of the 2026 Conference Champion Predictions

Table 3 summarizes the results shared in the previous parts of this series for the two teams from each conference that are expected to meet in their respective conference championship games from three different points of view.

Table 3: Review of the projected conference championship game participants based on the preseason simulation results, the Most Likely Scenario, and the Disruptive Scenario.

The left side of the table presents the likely conference championship participants based on the raw odds generated from my preseason Monte Carlo simulation of the full season. The middle of the table gives the results from my Most Likely Scenario, where all the favored teams win every game.

The results on the right show the Disruptive Scenario. This scenario starts by assuming the favored teams win every game on the full-season schedule, then adjusts the results to add a historically reasonable number of road upsets.

The right side of the table gives each team's record in the Disruptive Scenario, as well as some of the notable wins and losses for each team in that scenario. I use the results of the Disruptive Scenario as my official picks.

Table 4 gives the picks for each conference championship game winner using the three sets of assumptions discussed above.

Table 4: Review of the conference champion picks based on the preseason simulation results, the Most Likely Scenario, and the Disruptive Scenario.

There is a significant amount of agreement among the three sets of predictions, but there are also some notable differences. The second teams to qualify for the championship games in the ACC, Big Ten, and SEC are the biggest points of disagreement.

Also note that each table includes the sum of the odds for each set of picks at the bottom. Similar to last year, and every year that I have tracked these stats, only seven or eight championship game participants and two or three conference winner selections are expected to be correct.

That is not evidence of a systematic flaw in my preseason analysis, or anyone else’s. As discussed in Part 1, it is simply a reflection of the uncertainty surrounding team quality, the schedules those teams will face, and the role of plain old luck.

The uncertainty is the point. After reviewing how last year’s preseason picks performed, it is time to apply the same methodology to the 2026 season and build the full 12-team College Football Playoff bracket.

Who gets the automatic bids? Which contenders claim the final at-large spots? How does the committee sort out the seeding? And once the bracket is set, which team does the model project to win it all?

Subscribe to Spartans Illustrated to read the full playoff projection, including Dr. Green and White’s projection of the the complete 12-team field, bowl assignments, projected point spreads the 2026-27 national championship.