2021 College Football Championship Odds
Do the Bearcats have reasonable college football championship odds or are they more likely to go down as a footnote in history? From Andy Dalton’s 0-4 playoff record to the Reds’ catastrophic meltdown against the 2012 Giants; From both Xavier and Cincinnati being upset in the second round of the 2018 March Madness Tournament (in the same arena on consecutive days!) to the torture of Brian Kelly leaving the Bearcats before the 2010 Sugar Bowl matchup against Tim Tebow; From Carson Palmer’s career altering injury against the Steelers in the playoffs, to the #1 Bearcats losing the player of the year and eventual #1 overall pick Kenyon Martin in the CUSA tournament effectively ending their championship hopes; Cincinnati sports has had a lot of negative outcomes over the last 30 years. Unfortunately for Cincinnati fans, though, there have been almost no bright spots.
In this article we’ll preview each team’s college football championship odds using our Bayes Ensemble model. This model uses a mixture of information sources to generate aggregate rankings. On the prediction tracker, our model is the second most accurate being tracked. Our model has correctly predicted 74.2% of college football games this year.
The first ingredient in the model is a team’s performance measured relative to the quality of their opponent. This measure of quality of win is a way of getting past the ‘strength of schedule’ problem that arises in college fotoball. If we want to compare Cincinnati to an SEC or B10 team, this becomes crucially important.
The second ingredient is a prior guess about how good the teams are. The college football season is short enough that statistical measures about team quality can have a lot of variance; there are too few data points to form good models. One game can unduly affect an estimate of how good a team is. Our model uses a prior based on how good a team was the previous year. This prior is worth about an additional three games worth of data.
The remainder of this article will contain a few parts. The first two contain the college football semifinal odds and the college football championship odds. After that, we’ll also dig into the New Year’s Day games to give you a little extra to chew on. Finally, we’ll give you our top 25 heading into the CFB semifinals.
College Football Championship Odds
The way to use our ratings is that the difference between two team’s ratings is the predicted margin of victory on a neutral field. The ratings for the four college football championship semifinalists are contained in the table below.
Georgia | Alabama | Michigan | Cincinnati |
|---|---|---|---|
33.1 | 29.7 | 24.9 | 21.7 |
These rankings give you a good idea of what the college football championship odds and probabilities will be. The most likely championship matchup is Georgia-Alabama and Georgia should be the favorite to win the whole thing because they’re the best team. But how likely or unlikely are the other outcomes?
Before presenting the data, here is a quick example of how the calculations are made. Georgia’s rating is 8.2 points higher than Michigan’s which implies a predicted line of Georgia -8.2. Using standard spread to probability calculators for college football, this results in us making the prediction that Georgia wins their matchup against Michigan about 74% of the time. The table below shows the college football championship odds for each team as well as their odds to advance to the championship game. The Champ ML column is what money-line odds returns ‘fair value’ for that team. To bet, you want to compare the Vegas college football championship odds to the values in that column. If Vegas’ values return more, then there is probably good value available.
Team | Advance % | Championship % | Champ ML |
|---|---|---|---|
Georgia | 74 | 48.2 | +110 |
Alabama | 74 | 34.2 | +192 |
Michigan | 26 | 10.8 | +825 |
Cincinnati | 26 | 6.7 | +1390 |
Note: to compute the college football championship odds, one needs to multiply the probability of winning in the semi-final by the probability of facing each different opponent in the championship game and beating them. This is just a simple application of the law of total probability. An important step in the college football championship odds is the matrix of probabilities for each potential matchup shown below computed with the same spread-to-probability conversions:
Team | Beat UG | Beat Bama | Beat UM | Beat UC |
|---|---|---|---|---|
Georgia | x | 60 | 74 | 80 |
Alabama | 40 | x | 64 | 74 |
Michigan | 26 | 36 | x | 58 |
Cincinnati | 20 | 26 | 42 | x |
Just quickly comparing my model’s college football championship odds to the Vegas lines, it seems as if Vegas is higher on Alabama than I am. However, my model is pretty close to what Vegas thinks about Michigan and Cincinnati. So, even though this is the first season a G5 team makes the college football playoff, it seems as if it is quite unlikely that the championship belt leaves the SEC this year.
New Year’s Day Bowl Odds
There are three main New Year’s day bowls, but let’s give the odds for each of the five matchups on that day. The five games are:
- Outback Bowl: Penn St. v Arkansas
- Citrus Bowl: Iowa v Kentucky
- Fiesta Bowl: Notre Dame v Oklahoma St.
- Rose Bowl: Ohio St. v Utah
- Sugar Bowl: Ole Miss v Baylor
Note that these games have no championship implications so it is often the case that high profile players will sit the game out so as not to risk injury and, eventually, their NFL draft stock. If that is the case, then my model will be quite a bit less accurate as it is not able to determine the impact of individual players. The odds for each matchup are given in the table below:
Bowl | Line |
|---|---|
Outback | Penn St -4.5 |
Citrus | Iowa -4 |
Fiesta | Notre Dame -3.5 |
Rose | Ohio St. -16.5 |
Sugar | Push |
Most notably, the Ohio State number should be taken with a grain of salt. A significant number of their players have opted out in favor of the upcoming NFL draft. Similarly, Notre Dame is missing their best player on defense which may shift things a bit in Oklahoma State’s favor. As always, use these numbers as a baseline and update your beliefs based on how important you think these departures actually are.
College Football Pre-Semifinal Rankings
The top 25 teams in college football according to our rankings are given in the table below.
| Team | Rating | |
|---|---|---|
| Georgia | 33.1 | |
| Ohio State | 31.3 | |
| Alabama | 29.7 | |
| Michigan | 24.9 | |
| Notre Dame | 22.6 | |
| Cincinnati | 21.7 | |
| Wisconsin | 19.6 | |
| Oklahoma State | 19.2 | |
| Clemson | 18.9 | |
| Pittsburgh | 18.0 | |
| Oklahoma | 18.0 | |
| Iowa State | 17.8 | |
| Texas A&M | 16.9 | |
| Penn State | 16.8 | |
| Iowa | 15.4 | |
| Baylor | 14.9 | |
| Mississippi | 14.8 | |
| Utah | 14.7 | |
| Auburn | 14.4 | |
| Wake Forest | 14.4 | |
| North Carolina State | 14.0 | |
| Purdue | 13.7 | |
| Nebraska | 13.5 | |
| Minnesota | 12.8 | |
| Texas | 12.7 |
There are two takeaways one can make from these rankings. First, Ohio State fans should be pretty upset about not having a chance to play for the championship. The committee made the right choice giving Michigan the nod over Ohio State because, well, Michigan won the conference. However, our numbers say that Ohio State is the better team. If Ohio State was in the playoffs, my guess is that they would have about a 35% chance at winning the championship.
Even more interesting is that our model thinks Notre Dame is better than Cincinnati. Almost identically to the Ohio State-Michigan conundrum, the committee would take an unbelievable amount of heat if they put the Irish in the playoffs instead of the undefeated Bearcats. However, if you rule Ohio State out for not winning their conference, the 4th best team according to our model is Notre Dame.
