If you’ve known me for any period of time, you probably know I like to see what the data says about a situation before forming an opinion. It can be my downfall, but it can also put a lot of oddities into perspective.
In college football, there are several advanced analytics that can help understand what we are seeing on the field, helping to see the forest through the trees and help account for pace of play, garbage time, opponent strength etc. We will touch on many in future articles, but wanted to take a brief moment to discuss ESPN’s SP+ rating that was just finalized for 2022.
A brief description for anyone who hasn’t seen this metric before: “In a single sentence, it’s a tempo- and opponent-adjusted measure of college football efficiency.” Essentially, your team is given a rating that is a projected score difference if you were facing a perfectly average team. If your rating is 21.0, you would beat the nationally average team by 21 points on a neutral field. Similarly, your offensive and defensive ratings are the number of points you would score (Offense), as well as give up (defense). When you add these with your Special Teams score, it becomes your team rating.
Let’s look at a few of the numbers for examples:
As you can see, it lined up well with the teams that made the playoffs, sans Alabama with a nice bump after their blowout win over Kansas State (who the ratings have as 9th). But what do we care about on this site? Our beloved Sooners.
If you’ve listened to any of our round table pods, this is what has been eating at me all season, and I’ve tried to explain audibly, but the visuals are so much more striking. This was a top-20 team, a top-10 offense, and a hair below average defense. Surrounded by double digit win teams, and a meager 0.2 points worse than a particular team we’ve followed all season out west.
And that is the question… what kept OU from having a record similar to the teams around it, when the data says they were a much better team?
- 2022 Big 12 schedule – more on this in a moment
- QB play – good enough to be in every game, not great enough to win several
- In-game coaching decisions
Taking a look at each, let’s start with Big 12 competition. I think this is often overlooked when seeking out reasons for OU’s poorer than expected start in the Venables Era. Here is a look at where all 10 teams finished in SP+
Seven teams in the top 33, 100% of the teams ‘above-average’ (positive rating), and graded out 2nd in conference rankings only to SEC (15.5 SEC average to 13.4 Big 12). And while they certainly have earned their place at the top, the SEC is riding coattails as their conference rating without UGA/Bama is 12.4. In a season of razor thin margins, we had the unlucky break of playing one of the toughest Big 12 seasons in recent memory.
Next, QB play. I won’t do a deep dive here as much has been dissected even already, but I’ll bring up a quick look I threw on the message boards on Dillon Gabriel performance by down as well as the passing chart from @mattburns:
Simply put, Gabriel has to be better, especially in big game moments, i.e. third downs. His QB rating dropped to a meager 110.84 on third downs, compared to 166-174 on first-second. The ability to extend drives helps both the offense and the defense, and two of the most winnable games, came down to not being able to extend drives and expecting our defense to succeed with 100+ snap counts. It’s not a formula for success. The Baylor game had a 46% win expectancy despite Gabriel throwing three interceptions.
As for in game decisions… One of the coolest features that Bill Connolly provides alongside his ratings, is a game by game look at projected wins expected, and how your team fared. Here is the OU schedule:
I’ve highlighted the two games that the data projected out as wins, at West Virginia and at Texas Tech. The post-game win expectancy for these were a whopping 70% at West Virginia, and 61% at Tech. As we know, OU was up 22-13 heading into the 4th quarter in Morgantown, and amassed a 24-6 lead in the late 2Q in Lubbock. Whether it’s the offense rushing to the snap with a lead, finishing 3-and-outs in under a minute, or odd decisions to kick long FGs in the rain, the green experience of this coaching staff showed it’s ugly teeth in season one.
So what do we do with all this data? First, I think you are seeing a solid approach to the portal, filling in the gaps that should move us forward, and not a total breakdown and rebuild. We sent players to the portal that were not contributing, and are working to add either starters or quality depth at various positions. On top of that, OU’s historical SP+ chart isn’t worrisome considering the recruiting classes we are bringing in.
You can see the ebbs and flows of some of our best teams and worst, but look at the last four years, since Kyler left. Offensive rating was falling off every season until this year, where had a slight uptick even while trading out the eventual Heisman winner for Dillon Gabriel (stock up for Lebby). Defense hasn’t been the same since 2011, reaching historically bad levels in 2018. The defense was 1.5 points worse than last year, but consider the fact we lost these players to the NFL: Nik Bonnito, Isaiah Thomas, Brian Asamoah, Perrion Winfrey, Dellarin Turner-Yell. If any team other than Alabama/Georgia lost that much defensive talent in one season, everyone would have expected a huge regression. Another way to look at it? How the heck did they only have a 1.5 point improvement if you had all that talent last year?! USC fans have already started to figure it out.
Looking ahead, I will showcase this in season and we can see how OU is operating game to game, and we can try to have an unbiased data look at how these coaches and players are performing. More questions? More thoughts? Come to the message board and let’s have a conversation!
The Data Behind OU’s 2023 Season