Over the past few weeks, I have been breaking down the upcoming college football season through the lens of math and simulation. I started by performing a breakdown of Michigan State's schedule and season outlook. I then performed a deep dive on the Big Ten race. Most recently, I looked at the conference races in the remaining Power Five conferences.
Today, it is time to look at my computer's projections for the remaining FBS-level teams who as a whole make up the Group of Five conferences. Let's jump right in, starting with the American Athletic Conference.
American Athletic Conference (AAC)
In the following sections, I will present two data tables to summarize the results of my calculations and full season simulation of the 2022 college football season. The first installment of this series explains my methodology in more detail. Briefly, most of the data is derived some a series of 240,000 Monte Carlo simulations where I use a consensus of the preseason rankings of all 131 FBS teams as the primary input.
Most importantly, I also use the historical uncertainty of these preseason rankings in my calculations. This twist results in much higher accuracy numbers than similar calculations from other sources. I use this simulation to calculate the odds for each team to win a certain number of games and handicap certain season odds.
Finally, I also make a calculation of the strength of each team's schedule and make two separate projections for the final record of each team. The baseline "most likely" projection is based on the scenario where the favored team wins every game. The modified "disruptive" simulation adjusts the baseline scenario by introducing a historically accurate number of home upsets.
Starting below with the ACC, Table 1 shows the win distribution matrix. Table 2 summarizes the consensus preseason ranking of each team, expected win total, projected records, strengths of schedule, and the overall odds to win the division, conference, make the college football playoffs and win the national title.
 |
| Table 1: AAC win probability matrix derived from the results of the 240,000 cycle Monte Carlo simulation, including the known uncertainty in the preseason rankings. |
 |
| Table 2: Summary of the preseason projections for the AAC, based on the consensus preseason rankings and a 240,000 cycle Monte Carlo simulation of the full college football season. |
Last year, the Cincinnati Bearcats made history by becoming the first Group of Five team to be selected for the College Football Playoffs. While the math gives the Bearcats a less than three percent chance to once again finish in the top four, Cincinnati does enter the 2022 season as both the top ranked team in their conference as well as the top ranked team in the entire Group of Five.
As such my calculations give the Bearcats a very good chance (74 percent) to return to the AAC Championship game. Their most likely opponent is the Houston Cougars (46 percent), but Central Florida (26 percent), Southern Methodist (19 percent), and Memphis (16 percent) all have solid odds as well.
Cincinnati does play Houston in the regular season this year, but the Bearcats do need to travel to both UCF and SMU. In addition, Cincinnati opens the season with a tough road test at Arkansas in Week One. These three road games will ultimately determine if the Bearcats are true contenders this year, or simply pretenders. My disruptive simulation suggests that Cincinnati will fall at Arkansas to open the season but will run the table after that to claim the AAC title over Houston and very likely another New Year’s Six bid.
As for other games that will likely impact the AAC race, watch for Houston at Memphis on Oct. 8. In addition, Houston at SMU and UCF at Memphis, both of which will be played on Nov. 5 will both likely define which team ultimately makes it to the AAC Championship Game.
Mountain West Conference
Similar to the analysis above for the AAC, Tables 3 and 4 below summarize the win distribution matrix and the season odds and other data from the Mountain West Conference.
 |
| Table 3: Mountain West win probability matrix derived from the results of the 240,000 cycle Monte Carlo simulation, including the known uncertainty in the preseason rankings. |
 |
| Table 4: Summary of the preseason projections for the Mountain West, based on the consensus preseason rankings and a 240,000 cycle Monte Carlo simulation of the full college football season. |
In the Mountain Division of the Mountain West Conference, the usual suspect (Boise State) once again enters the season with the highest preseason ranking and the best odds (42 percent) to advance to the conference championship game.
However, the odds and season simulations suggest that the Air Force Falcons pose a real threat to the Broncos. First of all, Boise State needs to travel to Air Force this year. Based on the initial rankings of each team, this game appears to be a virtual toss-up. If Air Force can win that game, the Falcons will have the inside track in the division race.
The challenge for the Air Force Academy, however, is that they have to travel to both Utah State and San Diego State this year. As a result, both my baseline and disruptive simulations have Boise State winning the division.
As for a possible spoiler, Utah State odds (15 percent) are notable, but the Aggies are still very much a long shot.
In the West Division, Fresno State is the clear favorite (61 percent odds to win the division) thanks in some part to the fact that they hold the easiest conference schedule in the Mountain West. While the Bulldogs do travel to Boise State in early October, they will face the projected two next best teams in their division (San Diego State and San Jose State) at home.
As for those two potential spoilers, San Diego State has better odds (25 percent versus nine percent) due to the fact that the Aztecs are projected to be a little better and they did the benefit of playing San Jose State at home.
My official prediction is that Boise State will edge out Fresno State in the Mountain West Championship. As for a potential berth in the New Year's Six this year, several of the conference contenders will have high profile non-conference games that could serve as resume builders.
Boise State will host BYU and travel to Oregon State. Fresno State travels to USC, San Diego State visits Utah, San Jose State will travel to Auburn, and Utah State will square off with Alabama in Tuscaloosa in Week Two.
However, Air Force's toughest non-conference game in 2022 current appears to be a neutral site contest with Army. Due to this fact, the Falcons may be the Mountain West's best chance to reclaim the Group of Five bid into the New Year's Six.
Sun Belt Conference
Tables 5 and 6 below summarize the win distribution matrix and the season odds and other data from the Sun Belt Conference.
 |
| Table 5: Sun Belt win probability matrix derived from the results of the 240,000 cycle Monte Carlo simulation, including the known uncertainty in the preseason rankings |
 |
| Table 6: Summary of the preseason projections for the Sun Belt, based on the consensus preseason rankings and a 240,000 cycle Monte Carlo simulation of the full college football season. |
With all of the conference realignment happening in the Power Five over the past few years, it is easy to forget that the teams in the Group of Five have also been making a few moves. This year, Marshall, Old Dominion, and Southern Miss will now be playing in the Sun Belt Conference instead of their previous home in Conference USA.
As a result, the Sun Belt's ranks have swollen to 14 members. Based on Table 6, the "Fun Belt" looks to be one of the more competitive leagues in the FBS. In the East Division, Appalachian State is regarded as the preseason favorite and owns the best overall odds at 33 percent. However, the three teams behind the Mountaineers all have odds to win the division that are over 10 percent. Coastal Carolina (27 percent), Marshall (19 percent), and Georgia State (12 percent) all look like contenders.
While Appalachian State would seem to have the upper hand, the disruptive simulation highlights a different possible scenario. This year Appalachian State has to travel to play the two next best teams (on paper) in their division: Coastal Carolina and Marshall. Furthermore, Marshall has a very favorable road schedule. The Thundering Herd play the top four teams in the Sun Belt East all at home. As a result, I project Marshall to be the surprise champion of the East in their inaugural champaign.
The Sun Belt West Division generally projects to be weaker overall that the East and as a result the picture is a bit clearer. Louisiana Lafayette has the best odds to win the division (42 percent) and the disruptive simulation suggests that the Ragin' Cajuns will win the West by two games.
Based on this analysis, I project that Marshall will defeat Louisiana Lafayette to claim their first Sun Belt Championship.
As for the Sun Belt Champion claiming a spot in the New Year's Six, this would seem to be a longshot as no team in the conference is expected to win more than 10 games this year. That said, there will be opportunities for several Sun Belt teams to score some high-profile upset victories.
Appalachian State hosts North Carolina and travels to Texas A&M in the first two weeks of the season. Marshall has a road game at Notre Dame, Southern Miss will take on Miami in South Beach, Troy State will face Ole Miss, Coastal Carolina travels to Virginia, and Louisiana Lafayette has a late season date at Florida State. A win to two over a Power Five team could but the Sun Belt in the conversation for a New Year's Six Bid.
Conference USA
Table 7 and 8 below summarize the win distribution matrix and the season odds and other data from Conference USA.
 |
| Table 7: Conference USA win probability matrix derived from the results of the 240,000 cycle Monte Carlo simulation, including the known uncertainty in the preseason rankings. |
 |
| Table 8: Summary of the preseason projections for Conference USA, based on the consensus preseason rankings and a 240,000 cycle Monte Carlo simulation of the full college football season. |
Conference USA (C-USA) looks a little different in 2022 with the departure of Marshall, Old Dominion, and Southern Miss for the Sunbelt conference, but the trimmer 11-team league looks to be very competitive.
Based on Table 8, C-USA looks like a three-team race for the two spots in the championship game. UAB (49 percent odds to make the conference title game), Texas – San Antonio (UTSA, 45 percent) and Western Kentucky (31 percent) seem to all be on the inside track.
Those top three teams all have one road game and one home game against the other two contenders, so there is no clear schedule advantage at the top. That said, my calculations suggest that Western Kentucky has the toughest conference schedule overall of those three teams as is therefore the most likely team to finish on the outside looking in.
As for potential dark horse candidates, Florida Atlantic, North Texas, Middle Tennessee State, and UTEP all have between 10 percent and 20 percent odds to reach the C-USA Championship Game, but the disruptive simulation does not project any of these teams to finish with better than a .500 record in conference play.
Based on my disruptive simulation, I project UAB to defeat UTSA in the Conference USA Championship Game.
As for New Year's Six potential, the eventual champion would almost certainly need a high-profile non-conference win to have a chance. UAB will have a chance at LSU late in the season and UTSA travels to Austin to face Texas in Week Three. Western Kentucky will also get their shot to build their resume at Auburn on Nov. 19.
Mid-American Conference (MAC)
 |
| Table 9: MAC win probability matrix derived from the results of the 240,000 cycle Monte Carlo simulation, including the known uncertainty in the preseason rankings. |
 |
| Table 10: Summary of the preseason projections for MAC, based on the consensus preseason rankings and a 240,000 cycle Monte Carlo simulation of the full college football season. |
If the preseason predictions are accurate, the race in the MAC looks quite straightforward in 2022. Both divisions have a clear favorite: Miami of Ohio (45 percent chance to win their division) in the MAC East and Toledo (40 percent odds) in the West.
Miami has a definite advantage of playing in the easier division and having the overall easiest schedule in the conference. The disruptive simulation projects that the RedHawks will win the MAC East by three games. A road game at Northern Illinois looks to be the only potential stumbling block.
As for Toledo, the Rockets project to be the best team in the conference this year, but the MAC West is a bit tougher than the East. Toledo also has to travel to Northern Illinois in 2022 and that game projects as a toss-up. My disruptive simulation projects that Toledo will win that game. However, if Northern Illinois can protect their home field, the Huskies have a great shot to repeat as MAC Champions.
That all said, my official prediction for the MAC is for Toledo to defeat Miami of Ohio in the MAC Championship Game.
The MAC is an extreme longshot to earn the bid to a New Year's Six Bowl. That said, Toledo will get a crack at Ohio State this year, Miami will face Kentucky, Cincinnati, and Northwestern, while Northern Illinois will face Vanderbilt and Kentucky.
Over the past four parts to this series, we have now explored the entire FBS landscape. In the next installment, it will be time to put all of the pieces together to start making predictions for the Playoffs, New Year's Six and other Bowl Games. Stay tuned.
Comments
Post a Comment