Hopelessly Optimistic College Football Preview: Part Three, National Outlook and the Playoffs

In the first two parts of this series, I explained my general methodology to analyze a season of college football and I gave the results of my analysis of the Big Ten. Of course, this analysis was predicated on the idea that somehow the original college football schedule will get played, perhaps in the Spring. While I know that that is extremely unlikely, I hope that at least imagining some version of a normal football season, or normal life in general, brings with it some comfort. I know that it does for me.

That said, today I present the remainder of my hopelessly optimistic 2020 college football preview, starting with a whirlwind tour of the rest of the nation. Let's begin with the SEC.

SEC

Table 1 below gives an overview of the results of both my one-million cycle Monte Carlo simulation and the strength of schedule calculations for the SEC.

Table 1: Summary of the full season odds for the SEC based on a one million cycle Mont Carlo simulation

As the table shows, both the SEC East and West races are expected to be competitive, with a full seven teams total having at least a 10 percent chance to win their division. In the East, Florida appears to be the team to beat, based in large part on the fact that they have the easiest schedule in the conference. Out West, Alabama gets the nod over LSU, and overall, the Crimson Tide have slightly better odds than the Gators to win the SEC Title game.

As for my disruptive simulation, it concurs. It actually projects Florida to run the table at 12-0, while it projects Alabama to finish at 11-1 with their only loss coming at LSU. Unfortunately for both LSU and Auburn, their road games appear to be too tough to allow them to take advantage of Bama's potential stumble in Baton Rogue.  I would also project Alabama to edge Florida in the SEC Title game. That said, if Georgia were to beat Florida in their annual game in Jacksonville, the Dawgs would likely win the East instead.

Big 12

Table 2 below summarizes my calculated odds for the Big 12. Note that in this case, "Division Title" represents the odds of finishing as the No. 1 seed in the Big 12 Championship Game.

Table 2: Summary of the full season odds for the Big 12 based on a one million cycle Mont Carlo simulation

The straight odds for the Big 12 are so straight forward that it is almost boring. Oklahoma is likely to finish first, Texas is likely to finish second, and the Sooners are likely to win the Championship Game.  But, it should be noted that both Iowa State and Oklahoma State have odds of around 15 percent to take the conference crown instead.

As for the disruptive simulations, it basically concurs with the Monte Carlo simulation, but it does show the potential for some chaos. It this scenario, Texas is able to edge Oklahoma in the Red River Shootout, but the Longhorns stumble to Oklahoma State in the final game of the season to finish conference play at 8-1.  The Sooners are able to beat their in-state rivals from Stillwater, but they also drop at game in Ames to Iowa State to finish 7-2.  In this scenario, there is path for Oklahoma State to sneak into the Big 12 Championship Game over Oklahoma if they can eek out victories at TCU and at Baylor. My simulation does not predict that, but the numbers are close. Overall, my pick is still the Sooners to win the Big 12.

ACC

Table 3 below summarizes my calculated odds for the ACC. 

Table 3: Summary of the full season odds for the ACC based on a one million cycle Mont Carlo simulation

As expected, consensus preseason No. 1 Clemson has by far the best odds to win the Atlantic Division and has basically 50-50 odds to win the ACC. If anything, that is likely low. In the Coastal Division, things are obviously tighter, but North Carolina (38 percent) appears to have a clear edge over Virginia Tech (26 percent) and Miami (14 percent). 

The disruptive simulation concurs in this case as well, as Clemson and UNC also happen to have the two easiest schedules in the ACC (due in part to the fact that they don't have to play themselves...)  But, it does suggest that Clemson might stumble on the road at Florida State or (more likely) at Notre Dame. I project the Tigers to finish 11-1 with a loss in South Bend.  As for UNC, I project them to pick up their only ACC loss at Miami, but for them to also lose two in the non-conference to UCF and Auburn.  Either way, Clemson is still projected to truck the Tarheals and win yet another ACC Title.

Pac 12

To round out the Power Five, Table 4 provides a summary of the odds for the Pac 12.

Table 4: Summary of the full season odds for the Pac 12 based on a one million cycle Mont Carlo simulation

The Monte Carlo simulation shows a pretty clear favorite in each division, with Oregon having a significant edge in the North and with USC as the strong favorite in the South. Oregon would be favored to win the Pac 12 Championship Game.

But, my disruptive simulation has a different idea. In the North, Oregon is still picked to win the Division. In fact, the Ducks are projected to run the table at 12-0, including an early season upset win over Ohio State in the friendly confine of Autzen Stadium.  In the South, however, things project to end a bit differently. In the disruptive simulation, both USC and Utah finish with a 7-2 conference record, but the Utes edge the Trojans in Salt Lake City to earn the tie-breaker and a date with the Ducks (which they are projected to lose).

Group of Five and Independents

As this is an equal opportunity post, I will also quickly touch on the simulation results for the Group of Five and the Independents.  Table 5 summarizes the results of my simulation for those teams that refuse to join a conference.

Table 5: Summary of the full season odds for the FBS Independents based on a one million cycle Mont Carlo simulation

I will only note here that Notre Dame projects in my disruptive simulation of going 10-2 overall with losses to Wisconsin in Green Bay and at USC, but with an upset win over Clemson. Next up is the AAC, summarized in Table 6

Table 6: Summary of the full season odds for the AAC based on a one million cycle Mont Carlo simulation

The odds give the edge to UCF to win the conference, but my disruptive simulations suggests a three-way tie between UCF, Memphis, and Cincinnati may be likely. In this case, the final playoffs rankings might be the deciding factor, and I project Memphis to claim the Title in this scenario, due to their head-to-head win over UCF and the fact that Cincinnati projects to lose a non-conference game at Nebraska.

Table 7 below give the season odds for the Mountain West.

Table 7: Summary of the full season odds for the Mountain West based on a one million cycle Mont Carlo simulation

Boise State has the schedule to go 12-0, including a potential road win over Air Force and a notable win over Florida State on the blue turf. In the West Division, San Diego State has better odds, but the disruptive simulation likes Nevada to win the Division, based on their easier schedule.

Next up is the MAC, which is summarized below in Table 8.

Table 8: Summary of the full season odds for the MAC based on a one million cycle Mont Carlo simulation

The Monte Carlo simulation here likes Buffalo in the East and overall and favors Toledo slightly in the West due to their easier schedule.  The disruptive simulation comes to the same conclusion.

Table 9 provides a summary of the preseason odds for Conference USA

Table 9: Summary of the full season odds for C-USA based on a one million cycle Mont Carlo simulation

The race in the East appears to be the most competitive race in all of the FBS with the top three teams all ranked essentially equally and with very similar odds. The Monte Carlo simulation likes Marshall by a nose, while the disruptive simulation projects a three-way tie between Marshall, FAU, and Western Kentucky. I am not sure how that tiebreaker would shake out, but Florida Atlantic projects to have the best overall record, so I will go with them.  In the West, UAB has the clear edge in odds, both for the division and conference crown, and the disruptive simulation agrees. It only projects a single loss to Miami in the non-conference.

Finally, Table 10 gives the summary for the Sunbelt.

Table 10: Summary of the full season odds for Sunbelt based on a one million cycle Mont Carlo simulation

In the Funbelt, both divisions appear to have a clear favorite: Appalachian State in the East and Louisiana-Lafayette in the West, as both team figure to have a 60 to 70 percent chance to win their Division.  The disruptive simulations give the same result. Overall, Appalachian State is a heavy favorite to win the conference.

Playoffs and New Year's Six Bowl Predictions

Based on the simulations, we now have a good feel for how each conference race might shake out. As for the Playoffs, the Table 11 below give the odds that I calculate for every team that has at least a one percent chance to win the National Title, which works out to 21 teams total.

Table 11: Playoffs odds for the Top 21 teams in the FBS, based on one million Monte Carlo simulations

Based simply on this table, the four most likely teams to make the playoffs are Clemson, Ohio State, Penn State, and Alabama.  However, these numbers can be a bit misleading. For example, Table 11 says that in 33.5 percent of the one million simulations, Ohio State made the playoffs and in 28.7 percent of the simulations, Penn State made the playoffs. But, at the end of the day, there is only one scenario that will actually take place, and certain outcomes are mutually exclusive. 

While the odds for both Penn State and Ohio State are good individually, a closer look at the simulation results show that the Big Ten only gets multiple playoff teams in about 19 percent of all cases and only 13 percent of all cases are those teams from the same division. So, it is unlikely that both OSU and PSU will reach the playoffs together. A similar analysis on the Power Five conferences, Notre Dame, and the Group of Five is shown below in Table 12.

Table 12: Breakdown of the different playoff scenarios

This table gives a better view of which specific scenarios are most likely.  Overall, there is about a 60 percent chance that one conference will put multiple teams into the playoffs, with the SEC (32 percent) and the Big Ten (19 percent) being the most likely sources. The SEC has 84 percent chance to get at least one team in the playoffs, and the odds for at least one Big Ten team are 76 percent.  

As for the ACC and the Big 12, both leagues have about a 50-50 shot at getting one team into the playoffs, while the odds for the Pac 12 getting one team are right at one-third.  As for Notre Dame and any the Group of Five, I estimate both of their odds to be about one-in-six.

If we take this information into consideration, it seems the most probable playoff make-up is to have one Big Ten team, one or two SEC teams, and an ACC and/or Big 12 team.  Based on the Table above, the most likely two combinations appear to be:

• Clemson
• Ohio State
• Alabama
• LSU

or

• Clemson
• Ohio State
• Alabama
• Oklahoma

I am not sure about you, but these are two pretty super-boring outcomes. Fortunately, if I look to my disruptive simulation, I can identify some potentially more interesting scenarios that are plausible. If I extrapolate the results of that simulation, this is what the landscape of college football might look like after the conference championships:

• Penn State (11-2) beats Wisconsin (10-3) in the Big Ten Championship game. Penn State has losses at Michigan and at Virginia Tech.
• Ohio State (10-2) has loses at Oregon and at Penn State and did not make the BTCG
• Alabama (12-1) beats previously undefeated Florida (12-1) in the SEC Championship game.  Bama's only loss came at LSU
• Georgia, Auburn, and Texas A&M finish at 10-2, while LSU is only 9-3. Note also that Georgia beats Auburn and Auburn beats A&M in head-to-head conference action.
• Oklahoma (11-2) beats Texas (10-3) in the Big 12 Championship Game. Oklahoma's losses came at Texas and at Iowa State, while Texas lost at LSU and at Oklahoma State.
• Clemson (12-1) beats North Carolina (9-4) in the ACC Championship Game. The Tigers only loss came at Notre Dame.
• Louisville and Virginia Tech also finish the regular season at 10-2, with Louisville owning a win over the Hokies in the regular season.
• Oregon (13-0) beats Utah (10-3) to win the Pac 12 Championship. Arizona State also finishes at 10-2.
• Notre Dame finishes the regular season at 10-2 with loses at Wisconsin and USC, but with a win over Clemson
• In the Group of Five, Boise State runs the table and beats Nevada in the Mountain West Championship game to finish a perfect 13-0, including a win over Florida State.

In this scenario, the top three teams in the final College Football Playoff poll would almost certainly be undefeated Oregon, Clemson, and Alabama. The final team could possibly be Penn State or Oklahoma, but I bet Florida with 12 wins would get the nod.  That would result in the following playoff match-ups:

• Rose Bowl: No. 1 Oregon (13-0) vs. No. 4 Florida (12-1)
• Sugar Bowl: No. 2 Alabama (12-1) vs. No. 3 Clemson (12-1)

Based on the preseason rankings, I would project Clemson to beat Alabama and then Florida in the National Title game.

As for the remaining New Year's Six Bowls, there are fewer contractual constraints in the 2020-21 cycle, but the Orange Bowl must still select an ACC team to play against the highest ranked non-Champion from the SEC or Big Ten.  With that constraint, I think the following NY6 pairings would be the most likely:

• Orange Bowl: Louisville (10-2) vs. Georgia (10-2)
• Cotton Bowl: Oklahoma (11-2) vs. Penn State (11-2)
• Peach Bowl: Ohio State (10-2) vs. Auburn (10-2)
• Fiesta Bowl: Boise State (13-0) vs. Notre Dame

If the full season were somehow to be played, this is how I would project it to go down. In the next few days or weeks, the actual trajectory of the college football season should become clear. Once it does, I will once again dust off my simulation and report back on what the math tells us to expect.  If needed, maybe I will just simulate the entire college football season. Literally no one wants that, but these are strange times in which we are living. It pays to be prepared.

Until next time, wear a mask and social distance if you want college sports.  Enjoy, and Go Green.