Tommy Rees Departs for Alabama
A tough blow for Notre Dame, but also a potential blessing in disguise?
News broke on Friday that offensive coordinator Tommy Rees would be leaving Notre Dame for the same role under Nick Saban at Alabama. This shouldn’t come as a major surprise considering his brief flirtation with Miami last offseason after Marcus Freeman was hired, as well as the speculation that Sean McVay was considering him for the vacant offensive coordinator role with the Rams.
But nonetheless, Marcus Freeman now has to make the most important hire of his young head coaching tenure. There’s no doubt Rees is a great coordinator and has earned the opportunity to choose where he wants to be. However, from watching the games to analyzing the data, I’ve always found something missing in his performance that’s held him back from being a truly elite play caller.
This is not to say he isn’t a good play designer. You could point to many plays each season that were so creative and left your proverbial jaw on the floor wondering how he schemed them up. Yet, play design is separate from the responsibility of knowing in which situation to call what play, an area I’ll show he seems to have struggled in.
These struggles are not necessarily indicative of future performance, either; he may very well develop into a great play caller in talent-rich Tuscaloosa. Regardless, I’m going to dive deep into his tenure here and hopefully illustrate why his departure from the university might be best for both sides.
Chip Long vs. Tommy Rees
Maybe the first thing we should do to understand Rees’ impact on the Irish offense is analyze his units’ performance compared to those of his predecessor. When doing so, it’s not clear that, from an efficiency perspective, Rees took the Irish to another level.
If we discount the 2017 season, when Brandon Wimbush was the starting quarterback and Notre Dame leaned into “33 Trucking” Josh Adams and the monster offensive line they had, the opponent-adjusted offensive results were almost identical every year.
And before we get into the “he didn’t have a good enough QB to succeed” argument, here are the PFF grades for each starting quarterback since 2017:
Brandon Wimbush (2017): 74.9
Ian Book (2018): 70.2
Ian Book (2019): 77.9
Ian Book (2020): 78.5
Jack Coan (2021): 83.5
Drew Pyne (2022): 69.1
You could potentially argue that Rees had slightly better quarterback talent at his disposal than Long considering he was the coordinator for the two highest PFF quarterback grades in the last 6 years. The difference isn’t massive, however, and they produced very similar offenses that excelled through the air while remaining fairly pedestrian on the ground.
Since Rees didn’t really improve the offense’s efficiency relative to his predecessor, let’s take a look at his play calling and see if he pushed Notre Dame further with more optimal performance in that area.
Optimal Play Call Rate and Pass Rate Over Expected
Since the offensive coordinator is not fully in control of the offensive results, another way we can analyze their performance is investigating the area they do have full control over: play calling.
I’ve built two different models that try to accomplish this, the first being an optimal play call rate model. This works by utilizing different game-state features (down, distance, score, win probability, etc.) and team strength features (our position unit elo ratings) to predict the probability of a successful play if the offense were to run or pass. From there, we can calculate the rate at which the offense ran the preferable play, e.g. their optimal play call rate. It’s important to note that this model doesn’t take into account any schematic data nor can it parse out scrambles and designed QB runs, categorizing them all as rushes. It’s thus not an exhaustive answer to evaluate play calling, but an important tool to help illustrate the greater picture of Rees’ performance.
As seen above, Tommy Rees does not shine the brightest through this lens. While Notre Dame ranked 15th in optimal play call rate in 2021, 2020 and 2022 saw them ranked 77th and 100th, respectively. Additionally, the Irish rushing attack struggled heavily to begin the 2021 season, the one year he rated very well in the metric.
For comparison, Chip Long ranked 93rd in 2017, 44th in 2018, and 50th in 2019. Therefore, optimal play call rate presents another on-field area where it’s difficult to parse out where Rees improved on what Long left him.
The second model is an expected pass rate model. This predicts the probability of a pass on every play given different game-state characteristics. From there, we can measure a team’s pass rate over expected (actual pass rate - expected pass rate) and in different scenarios (1st & 10, 2nd & long, etc.). This gives us a lens to evaluate how conservative or aggressive a team was in throwing the football compared to what an average team would have done.
Yet again, it does not show Rees in the most flattering light.
The Irish were one of the least aggressive teams in the country throwing the football despite its relatively efficient passing offense, ranking 101st in pass rate over expected.
To me, the most notable finding from this visual is his 2nd & 7+ performance which appears to show a lack of situational awareness. 2nd & long should never be a rushing scenario unless you’re trying to bleed the clock. Frustratingly, Rees consistently treated it like a 3rd & 1 and threw the ball 14% (!!!) fewer times than expected, which rated 121st in the country. While Notre Dame ranked 39th in 2021, in 2020 they came in at a paltry 96th. Moreover, the ‘21 number could also be explained away by the offensive line’s struggles in run-blocking to begin the year.
So where does that leave us?
I truly believe both Notre Dame and Tommy Rees are going to be better off going forward after the separation.
Rees needs to evolve as a play caller to be more pass-first. Alabama is the perfect place for him to do so as they have shed their run-first identity and embraced a forward-thinking offense. Here are the Tide’s optimal play call rate and pass rate over expected rankings since 2020:
Optimal play call rate
2020: 26th
2021: 13th
2022: 19th
Expected pass rate
2020: 9th
2021: 9th
2022: 14th
Now, this potential evolution is dependent on Rees embracing Alabama’s status quo, not the other way around. Given the receiving firepower Rees will inherit in Tuscaloosa, he would be foolish not to.
Notably, I don’t think this development would have been possible at Notre Dame. Marcus Freeman still believes in a run-first offense (which we will address in our article breaking down the offensive coordinator candidates) and likely wouldn’t have challenged Rees to be more aligned with the analytical idea of aggressiveness in the passing game.
Sometimes a relationship is great and a lot of fun, but reaches a point where both parties have different needs in their life than when it started. For both Notre Dame and Tommy Rees to improve and reach their goals, a break-up was unfortunate but likely necessary. Only time will tell how it plays out.