2021 Season Win Projections

1,000 Simulations. Who has a good season ahead of them?

We are still in the spring season swing at the moment, but now seems like a good time to take a peak forward at the 2021 fall season. Why? Because why not! It is interesting to see what our projections tell us now, and we might be able to seize on some early value when win totals are released in the coming months. First things first, we need a way to gauge a team’s overall strength.

To do this we will use the ELO rating system which will give us a rating for each team coming into the 2021 season. For those that are unfamiliar, an ELO rating system is a system originally used to rank the relative skill of chess players. The idea being players with equal ranks would be expected to win an equal amount of times against each other. After some slight modifications to the model and additions such as a home field advantage modifier, the model was ran on every game in the CFB Playoff Era to date (2014-Present).

Now that we have our ratings we can compute the probability a team will win against another team, simulate the games 1,000 times and see how many wins each team comes out with at the end of the season. It is still early in the offseason and a lot of things can still happen over the course of the summer (transfers, injuries etc.). So this is more of a baseline expectation. Without further ado, here is the 2021 season!

ACC

I occasionally get yelled at on twitter about leaving Notre Dame out of power 5 content, so I will throw them into the ACC for now. It also helps add some power into the conference, because without them Clemson is once again the runaway favorite to win the ACC. Clemson was a virtual lock to win 9+ games in all 1,000 simulations, and ended up averaging a near perfect season. Filling the shoes of number 1 overall pick Trevor Lawrence will be difficult, but Clemson is in good hands with former 5 star recruit DJ Uiagalelei. Excluding Notre Dame, 8 teams reached 11 wins at least once in the simulations, showing us that there should be at least once valiant challenger to Clemson’s reign at the top. Could it Sam Howell and the Tar heels? Or perhaps one of the more dark horse teams like a Pittsburgh? One interesting game to watch: Clemson vs NC State September 25th in Raleigh. NC State returns a lot of production including promising QB Devin Leary. Even if they don’t come away victorious, NC State could be in for a promising season. If FSU has any chance to shatter expectations and the ceiling for this season, UCF transfer McKenzie Milton has to show the world he’s back and better than ever after a 2 year hiatus.

SEC

Alabama built itself a sizable lead in the ELO ratings after a historically dominant 2020 season. Similar to Clemson they will be replacing a 1st round QB with a former 5 star recruit, Bryce Young. In what seems like a yearly tradition at this point, Alabama will also be replacing most of their assistant coaching staff after the previous staff went off to bigger and better things. Assuming Alabama hits their projected 11+ wins, the battle for 2nd in the SEC will be the biggest question this season. Texas A&M appears to be the media favorite for that role, but my model actually prefers Georgia, who will get to enjoy a full season from former USC QB J.T Daniels. The most interesting thing about these simulations is nearly every team in the league got to 8+ wins at least one time (Sorry Arkansas, you have the hardest schedule in America per ESPN FPI). Will we see new Tennessee coach Josh Heupel defy the odds and create a high flying offense in year 1? Probably not, but wouldn’t that be something! I will leave you with this thought….. the Lane Train hitting the recruiting trail with a 10 win season under his belt.

Big Ten

Hey look, another dominant program replacing a future first round QB! This time its the Buckeyes going from Justin Fields to…. to be determined? They will still be the runaway favorite to win the league, but don’t count out any of the challengers just yet. Simulation results had 10 teams (excluding Ohio State) reaching double digit wins, with Maryland and Illinois topping out at 9 wins. Many of those teams have less than 0.00% chance to achieve those maximum range wins, but still have respectable chances to reach bowl eligibility at 6+ wins. Indiana gets Michael Penix Jr back from injury, which will be crucial if the Hoosier want to parlay their 2020 season into something greater. As always Iowa seems to have a solid floor, my model giving them a 56% chance to hit at least 9 wins. It seems to be another status quo year for Michigan under Harbaugh which begs the question… is it time to move on?

Big 12

Aha we have a conference favorite that isn’t losing a QB! Which is ironic given that Oklahoma went multiple years churning through one and done QBs. Spencer Rattler and the Oklahoma Sooners looked poised to return to the CFB Playoffs. To do that they will have to go through Matt Campbell and the Iowa State Cyclones, who return a lot of talent including star QB Brock Purdy. Its amazing to see what Campbell has done for that program, turning a bottom dweller into a program who looks to be on track for 8+ wins in 2021. Despite the coaching change, expectations will be sky high for Steve Sarkisian and the Texas Longhorns. Given the model doesn’t take this into account, we may need to temper expectations a bit in year one. Kansas finally has an AD, so a coaching search will begin to look for a replacement for disgraced former coach Les Miles. 1.36 wins on average may be generous here.

Pac 12

Last, but certainly not least, we head out west to check out the Pac 12. This is easily the tightest conference with the most uncertainty. USC looks to take back control of the conference from Oregon, and they’ll do it through the arm of QB Kedon Slovis. UCLA is perhaps underrated by the model given their unlucky finish to 2020 (Losing to USC and Stanford despite 95.5% and 93.3% post game win probabilities, respectively). The team I am most interested to watch is Arizona State, especially the development of former blue chip QB Jayden Daniels. The model gives them a 73% chance to make it 6+ wins, but doesn’t like their overall ceiling. If Jayden Daniels can take that next step forward, you may see a broken ceiling.

Conclusion

College Football is still many moons away and there is still plenty of time for roster shakeup (also Kansas needs a coach). Still, this should serve as a baseline to the exceptions we have for teams going into the season. This is just one tool we can use to gauge how the season is going to play out. If you plan on betting on over/under win totals when they are released (bet responsibly folks), you can use this as a starting point in your quest to find edges in your betting strategy. In any event, I hope you enjoyed a little look ahead into the 2021 season, and hope you were able to find someway to lie about your team to your friends or coworkers!

If you want to dive in to the data like I do, check out @CFB_Data and @cfbscrapR on Twitter, where you can learn how to get started in the world of College Football data analysis!

If you want to see more charts and one off analysis, follow my twitter page, @CFBNumbers