I did this with Clemson, a couple of you may recall. I analyzed how well the offenses and defenses of Notre Dame and USC fared against their opponents' season averages in each game, totaled the results and divided them by the number of games played.
Notre Dame's offense has outperformed their opponents' defenses by 4.71 points per game; USC's by 12.93 ppg.
Notre Dame's defense has outperformed their opponents' offenses by 6.64 ppg; USC's by 0.49 ppg.
Total outperforming: Notre Dame 11.35 ppg; USC 13.42 ppg. So the slight edge should go to USC. But as Lee Corso says, "Not so fast, my friend!" Let's take a little closer look.
USC's defense has a disturbing (for them) trend: They have allowed each of their last 5 opponents to score more than their season average, by an average of 7.1 ppg. Meanwhile, Notre Dame's offense has started to click, outperforming each of their last 5 opponents' defenses, and by an average of 13.2 ppg. A fair inference can be made that Notre Dame has a great chance score quite a bit over their 30.1 season average. Another factor to consider is that in 7 out of their 11 games, USC's defense has allowed their opponents to score more than their season averages.
What about USC's offense? If they perform averagely (yes, that's a word!), they would score about 33 points (12.93 + 20.3). But that doesn't take into account that Notre Dame has, in 9 out of 11 games, allowed fewer points than their opponents' season averages. A fair inference can be made that USC will actually have a hard time scoring 33 points.
Do not be surprised if Notre Dame scores 38 points or so, and do not be surprised if USC scores 30 or fewer. In fact, the trends suggest this has a decent chance of happening. But as we all know, the game is not played on paper.