Over the past several weeks, we have traveled together on a mathematical journey where we previewed the upcoming college football season. We started with a deep dive into Michigan State University’s schedule and season prospects. We then analyzed the Big Ten, the remaining Power Five, and the Group of Five. Today, it is time to put all of the pieces of the puzzle together and make some predictions about the 2022 college football postseason.
Here is a summary of how I see each conference championship playing out. In each case I show the projected record of each team following the conference championship game along with a note about each team’s key wins and losses.
In this scenario, the defending champions, Georgia, would likely be on the outside looking in with two losses. With so many other one-loss teams, the SEC Championship Game would become a play-in game for the playoffs.
Similarly, Notre Dame would own a head-to-head win over Clemson, which would give the Fighting Irish an edge and likely eliminate Clemson. As for Utah, the Utes would project to have the weakest schedule of all of the playoff contenders. As a result, I would project the playoffs result in the following match-ups, with my projected point spreads shown in paratheses for the two semifinals.
Clemson would automatically be placed into the Orange Bowl as the ACC Champion and Utah would be placed into the Rose Bowl as the Pac-12 Champion. The opponents for the Tigers and Utes would both most likely come out of the Big Ten, but the selection in this scenario would get interesting.
Based on overall record, Michigan State and Ohio State (both at 10-2) would be the most likely candidates. Penn State (9-3) could make a case, however, as in this scenario the Nittany Lions would have beaten both teams and finished in a three-way tie with the Spartans and Buckeyes for second place in the Big Ten East.
In the end, I believe that the overall record would prevail and Ohio State and Michigan State would wind up ranked higher than Penn State. But, which bowl would the Buckeyes and Spartans play in?
I would guess that Ohio State would end the season as the higher ranked team, especially considering that the Buckeyes would be coming off a win over No. 3 Michigan and Michigan State would be coming off a loss at Penn State. Strictly speaking, this would suggest that Ohio State would return to Pasadena in the Rose Bowl and Michigan State would go to Miami to play in the Orange Bowl as the Rose Bowl should get the “next best” Big Ten team as a replacement for playoff bound Big Ten Champions, Michigan.
But this would result in a Rose Bowl rematch between Ohio State and Utah. Furthermore, Clemson would likely be ranked higher than Utah, so an Ohio State / Clemson match-up would likely be more compelling. For this reason, I believe that the Selection Committee would send Michigan State to the Rose Bowl and Ohio State to the Orange Bowl.
The only remaining New Year’s Six Bowl is the Cotton Bowl. The top ranked Group of Five team in this scenario would almost certainly be Cincinnati and the Bearcats would automatically be sent to Dallas. As for their opponent, N.C. State at 11-1 would be the most likely as the best remaining candidate.
In summary, I project the rest of the New Year’s Six to be as follows, with the estimated final ranking of each team and the projected point spread for each game:
The first six columns of Table 1 below are a summary of much of the data that I have discussed over the past few weeks for a selected number of playoff and National Title contenders. The table contains the consensus preseason rank of each team, their strength of schedule, and my calculated odds to make the playoffs and win the championship.
The middle four columns contain the current betting lines for the playoffs and champion as provided by DraftKings SportsBook. For each money line, I have made the calculation to convert this line into probability so that their odds can be compared to my odds directly.
If we make the bold assumptions that my odds are correct, it is possible to calculate a theoretical return on investment (or ROI) for each potential bet. The final two columns show this ROI for each event, assuming that $100 was wagered.
In other words, if a team above has an ROI that is positive, that means that I project that team to have better odds to either make the playoffs or win the title than Vegas does. So, betting on that team is a “good investment” from a risk point of view. Conversely, teams with a negative ROI are a relatively bad investment.
Table 1: Summary of my calculated odds as compared to the odds provided by DraftKings SportsBook. For each bet, an ROI is estimated from a $100 bet, assuming that my odds are correct.
Table 1 suggests that there are several teams that might be worth a wager on either the National Title or a playoff berth for those who are inclined. Notre Dame has a very high ROI for both the playoffs and National Title. Oklahoma State is in second place on this table, followed by the ACC dark horse contenders Miami and N.C. State. Utah and Oregon also make the list, as do Michigan, Wisconsin, and Michigan State.
Interestingly, these teams all seem to have one thing in common and that is a relatively weak schedule. Of the top nine teams listed in Table 1, none of them have a schedule ranked in the top 40 of the FBS, based on my calculations. Many of these teams also tended to do well in my disruptive simulation, which all hints at a schedule advantage. While betting on any of those teams to win the National Title may be a bit too bold, a playoff wager or two seems reasonable.
It is also notable that the consensus top four teams in the preseason (Alabama, Ohio State, Georgia, and Clemson) all show fairly negative ROIs. This does not mean that one of those four teams will not win the National Title. It just means that the payback for a bet on one of those teams may be out of alignment with the real odds.
However, this also might simply be a failure of my system to properly handle the very good teams who are essentially outliers. Based on the odds from our friend in Las Vegas, it can be estimated that there is about a 90 percent chance that one of those four preseason teams will win the National Title. My calculations give those four teams combined about a 35 percent chance. The betting line certainly feel right, based on the past decade or so of college football.
The final interesting bit of data in Table 1 are the three teams that are not ranked in the preseason top 10, but who have strongly negative ROIs: Cincinnati, Texas, and USC. The Bearcats seem to be getting a bump based on last year’s playoff appearance. My math suggests that they are a bad bet for 2022.
As for Texas and USC, there was some significant disagreement in the preseason rankings regarding these two teams. Texas was ranked in the mid to upper 20s by most sources, but ESPN’s FPI has the Longhorns at No. 7. The situation with USC is even more extreme. Athlon Sports has USC ranked No. 9 to start the season; Pro Football Focus, however, has the Trojans way down at No. 67. Most other sources are somewhere in-between.
In any event, the folks in Vegas seem to have more faith in Texas and USC than most other pundits. Perhaps they know something the rest of us do not. Or, perhaps it is just the knowledge that those fan bases are more likely than most to wager with their hearts than with their heads.
All of this talk of betting sure is interesting. Now that I have introduced the general method to compare my data to the Vegas spread, it would be interesting to take a closer look at some of the other season odds that are available, such as total wins and wagers on division and conference winners. That will be the subject of the final installment of this series. Stay tuned.
Review of Previous Predictions
In the previous three installments of this series, I made a prediction as to the final record of each team and the results of each conference championship game. My official process is to use the results of my “disruptive” simulation. This simulation uses the projected point spreads of all FBS games and adjusts the results to include a historically reasonable number of home upsets.Here is a summary of how I see each conference championship playing out. In each case I show the projected record of each team following the conference championship game along with a note about each team’s key wins and losses.
- Big Ten: Michigan (12-1, with a loss at Ohio State) defeats Wisconsin (9-4)
- SEC: Alabama (11-2, with losses at Tennessee and Ole Miss) defeats Georgia (11-2, with a loss at Kentucky)
- Big 12: Oklahoma (13-0) defeats Baylor (9-4)
- ACC: Clemson (12-1, with a loss at Notre Dame) defeats Miami (10-3)
- Pac-12: Utah (12-1, with a regular season loss at Oregon) defeats Oregon (11-2)
- AAC: Cincinnati (12-1, with a loss at Arkansas) defeats Houston (11-2)
- Mountain West: Boise State (11-2) defeats Fresno State (10-3)
- Sun Belt: Marshall (11-2) defeats Louisiana (10-3)
- Conference USA: UAB (10-3) defeats UTSA (8-5)
- MAC: Toledo (11-2) defeats Miami of Ohio (8-5)
- Notre Dame (11-1 with a loss at Ohio State, but wins over Clemson, BYU, and at USC)
- N.C. State (11-1 with a loss at Clemson and wins over Wake Forest and North Carolina)
- Ohio State (10-2 with a win over Michigan and Notre Dame but losses at Penn State and Michigan State)
- Michigan State (10-2 with a win over Ohio State, but losses at Michigan and Penn State)
- Oklahoma State (10-2 with a win over Texas, but losses at Oklahoma and Baylor)
- UCLA (10-2 with a win over USC, but losses to Utah and Oregon)
Analysis and Bowl Projections
If the scenario above were to come to pass, there would be seven viable candidates for the four playoff spots. Undefeated Oklahoma and SEC Champion Alabama (even with two losses) would be clear favorites, but the other two spots would go to some combination of Michigan, Georgia, Clemson, Utah, and Notre Dame.In this scenario, the defending champions, Georgia, would likely be on the outside looking in with two losses. With so many other one-loss teams, the SEC Championship Game would become a play-in game for the playoffs.
Similarly, Notre Dame would own a head-to-head win over Clemson, which would give the Fighting Irish an edge and likely eliminate Clemson. As for Utah, the Utes would project to have the weakest schedule of all of the playoff contenders. As a result, I would project the playoffs result in the following match-ups, with my projected point spreads shown in paratheses for the two semifinals.
- Fiesta Bowl No. 1 Oklahoma versus No. 4 Notre Dame (-2)
- Peach Bowl No. 2 Alabama (-2.5) versus No. 3 Michigan
Clemson would automatically be placed into the Orange Bowl as the ACC Champion and Utah would be placed into the Rose Bowl as the Pac-12 Champion. The opponents for the Tigers and Utes would both most likely come out of the Big Ten, but the selection in this scenario would get interesting.
Based on overall record, Michigan State and Ohio State (both at 10-2) would be the most likely candidates. Penn State (9-3) could make a case, however, as in this scenario the Nittany Lions would have beaten both teams and finished in a three-way tie with the Spartans and Buckeyes for second place in the Big Ten East.
In the end, I believe that the overall record would prevail and Ohio State and Michigan State would wind up ranked higher than Penn State. But, which bowl would the Buckeyes and Spartans play in?
I would guess that Ohio State would end the season as the higher ranked team, especially considering that the Buckeyes would be coming off a win over No. 3 Michigan and Michigan State would be coming off a loss at Penn State. Strictly speaking, this would suggest that Ohio State would return to Pasadena in the Rose Bowl and Michigan State would go to Miami to play in the Orange Bowl as the Rose Bowl should get the “next best” Big Ten team as a replacement for playoff bound Big Ten Champions, Michigan.
But this would result in a Rose Bowl rematch between Ohio State and Utah. Furthermore, Clemson would likely be ranked higher than Utah, so an Ohio State / Clemson match-up would likely be more compelling. For this reason, I believe that the Selection Committee would send Michigan State to the Rose Bowl and Ohio State to the Orange Bowl.
The only remaining New Year’s Six Bowl is the Cotton Bowl. The top ranked Group of Five team in this scenario would almost certainly be Cincinnati and the Bearcats would automatically be sent to Dallas. As for their opponent, N.C. State at 11-1 would be the most likely as the best remaining candidate.
In summary, I project the rest of the New Year’s Six to be as follows, with the estimated final ranking of each team and the projected point spread for each game:
- Rose Bowl: No. 7 Utah (-4.5) versus No. 9 Michigan State
- Sugar Bowl: No. 5 Georgia (-5) versus No. 11 Oklahoma State
- Orange Bowl: No. 6 Clemson versus No. 8 Ohio State (-1.5)
- Cotton Bowl: No. 10 N.C. State (-0.5) versus No. 12 Cincinnati
Bad Betting Advice
I have just spent the last few weeks presenting a tremendous amount of data on the upcoming college football season. It is reasonable to wonder how some of this data might compare to the betting lines for the 2022 season. I would like to officially kick-off the 2022 season of Bad Betting Advice with the following analysis of the playoff and National Title landscape.The first six columns of Table 1 below are a summary of much of the data that I have discussed over the past few weeks for a selected number of playoff and National Title contenders. The table contains the consensus preseason rank of each team, their strength of schedule, and my calculated odds to make the playoffs and win the championship.
The middle four columns contain the current betting lines for the playoffs and champion as provided by DraftKings SportsBook. For each money line, I have made the calculation to convert this line into probability so that their odds can be compared to my odds directly.
If we make the bold assumptions that my odds are correct, it is possible to calculate a theoretical return on investment (or ROI) for each potential bet. The final two columns show this ROI for each event, assuming that $100 was wagered.
In other words, if a team above has an ROI that is positive, that means that I project that team to have better odds to either make the playoffs or win the title than Vegas does. So, betting on that team is a “good investment” from a risk point of view. Conversely, teams with a negative ROI are a relatively bad investment.
Table 1: Summary of my calculated odds as compared to the odds provided by DraftKings SportsBook. For each bet, an ROI is estimated from a $100 bet, assuming that my odds are correct.
Table 1 suggests that there are several teams that might be worth a wager on either the National Title or a playoff berth for those who are inclined. Notre Dame has a very high ROI for both the playoffs and National Title. Oklahoma State is in second place on this table, followed by the ACC dark horse contenders Miami and N.C. State. Utah and Oregon also make the list, as do Michigan, Wisconsin, and Michigan State.
Interestingly, these teams all seem to have one thing in common and that is a relatively weak schedule. Of the top nine teams listed in Table 1, none of them have a schedule ranked in the top 40 of the FBS, based on my calculations. Many of these teams also tended to do well in my disruptive simulation, which all hints at a schedule advantage. While betting on any of those teams to win the National Title may be a bit too bold, a playoff wager or two seems reasonable.
It is also notable that the consensus top four teams in the preseason (Alabama, Ohio State, Georgia, and Clemson) all show fairly negative ROIs. This does not mean that one of those four teams will not win the National Title. It just means that the payback for a bet on one of those teams may be out of alignment with the real odds.
However, this also might simply be a failure of my system to properly handle the very good teams who are essentially outliers. Based on the odds from our friend in Las Vegas, it can be estimated that there is about a 90 percent chance that one of those four preseason teams will win the National Title. My calculations give those four teams combined about a 35 percent chance. The betting line certainly feel right, based on the past decade or so of college football.
The final interesting bit of data in Table 1 are the three teams that are not ranked in the preseason top 10, but who have strongly negative ROIs: Cincinnati, Texas, and USC. The Bearcats seem to be getting a bump based on last year’s playoff appearance. My math suggests that they are a bad bet for 2022.
As for Texas and USC, there was some significant disagreement in the preseason rankings regarding these two teams. Texas was ranked in the mid to upper 20s by most sources, but ESPN’s FPI has the Longhorns at No. 7. The situation with USC is even more extreme. Athlon Sports has USC ranked No. 9 to start the season; Pro Football Focus, however, has the Trojans way down at No. 67. Most other sources are somewhere in-between.
In any event, the folks in Vegas seem to have more faith in Texas and USC than most other pundits. Perhaps they know something the rest of us do not. Or, perhaps it is just the knowledge that those fan bases are more likely than most to wager with their hearts than with their heads.
All of this talk of betting sure is interesting. Now that I have introduced the general method to compare my data to the Vegas spread, it would be interesting to take a closer look at some of the other season odds that are available, such as total wins and wagers on division and conference winners. That will be the subject of the final installment of this series. Stay tuned.
