Improving directly after a humiliating loss is not an idea that’s exclusive to Oklahoma football, but it seems the Sooners are starting to make a habit out of dropping an easy one and then winning out.
Since Lincoln Riley arrived in Norman, Oklahoma is 37-0 in conference games when the Sooners enter with at least one loss on their record.
So we know that the Sooners’ winning percentage gets better after a loss. But does their actual performance get better? Or is this simply the result of flukey luck or even scheduling coincidences?
In an attempt to answer that question, I used CFBscrapR to compile data on every Sooner game against a power-five opponent since 2015. From there I classified each game under one of two categories: before or after Oklahoma’s first loss.
Once done, I averaged a handful of key stats (such as margin of victory) across each group. Here are the results:
As you can see, winning percentage is by far the biggest beneficiary. There’s no two ways about it, Oklahoma wins more, a lot more, once they’ve already dropped a game (and the singular loss was to Ohio State).
Next we’ve got average scores for each game. The Sooners’ offense averaged two more points a game while the defense held steady, but this doesn’t necessarily mean that the offense is improving solely (more on that later).
The next two stats, yardage differential and EPA margin per play, are there to ensure that OU is actually playing better on a granular level, not just winning by more points.
The answer, once again: yes, Oklahoma’s yardage advantage nearly doubled while its EPA margin increased by 38 percent. But we’re not done yet.
We still need to look at one more thing: who was Oklahoma playing against in these games? It’s possible that the Sooners’ simply plays easier teams down the stretch, creating bigger discrepancies in our metrics.
To get an idea, I averaged the SP+ Rating of every opponent in each set and found that Oklahoma’s opponents after the loss were actually slightly better (by about two points) than their opponents before the loss.
We can expect this to counteract our metrics a little bit, meaning the effect is actually slightly more pronounced than initially thought.
So do the Sooners play better after a loss? Since 2015, they certainly have. Can we say for certain this will continue indefinitely? No. This is a small sample size (p-rating of .3), but the early returns are promising.
Lastly, does this mean anything for this week’s game? Possibly. From what we’ve seen, the effect is at its highest directly after the loss and then diminishes slowly over the next several games.
Oklahoma’s average point differential in the first game after a loss is 21.5, compared to 14.8 when the Sooners are undefeated. Does that mean Oklahoma will come out seven points better this week than last? Not for sure, but it’s at least a little more likely now.
[vertical-gallery id=18376]