ND-USC: A Rivalry Recalibrating

The calendar has moved again, and the thing I keep doing — pulling the daily pageview counts off the Wikimedia Foundation's public tallies and laying them next to each other — has produced, this week, a shape I want to walk through carefully, because it is the first time since the Week 0 hinge that the two lines in this thread have moved together rather than apart.

Chapter 11 left us on the September 20 Notre Dame spike, the 5,638-view Saturday that confirmed the summer curiosity gap had not closed so much as it had been buried under the noise of actual football. What I want to do here is pick up on the far side of the first full October week and note, honestly, what the daily counts have been doing since the end of September — because they have started doing something new.

“Great honest evaluation Big big Notre Dame band, but made the best team win And honestly, you got a great program over there, high academics, and a great football program I'm also a big Danny ANIGE He was truly one of…”

— a viewer comment on Notre Dame Fighting Irish, YouTube, Oct 13

Here are the numbers. The Notre Dame Fighting Irish football page, from October 6 through October 12: 1,348 on October 6, 1,137 on October 7, 1,156 on October 8, 1,389 on October 9, 4,573 on October 10, 3,467 on October 11, 2,630 on October 12. The USC Trojans football page, over the same stretch: 849 on October 6, 987 on October 7, 711 on October 8, 784 on October 9, 1,538 on October 10, 2,541 on October 11, 1,119 on October 12. Read those two columns side by side and the thing that jumps out is not the Notre Dame weekend spike, which by now is a shape we have seen three or four times — a weekday floor somewhere in the low-to-mid four digits, a Friday step-up, a Saturday peak, a Sunday that stays elevated and then bleeds back toward the floor by Monday.

What jumps out is that the USC line, this week, finally did the same thing.

Go back and look at the pageview walkthroughs in Chapters 9 and 11. The USC curve, through most of September, did not have this shape. It had the one enormous Week 0 Friday that Chapter 9 built an entire argument around, and then it had a long stretch of weekday numbers that sat lower than Notre Dame's weekday numbers and that did not reliably spike into the weekend. The two lines were not running on the same schedule. One of them was keeping football hours and the other one was not.

This week they are. The USC Trojans football page went from 711 on Thursday October 8 to 1,538 on Saturday October 10 to 2,541 on Sunday October 11, which is a bigger Sunday number than the preceding Saturday — a shape I have not seen on the USC line in any week I have walked through in this thread. It is the shape of a game that happened and that people kept reading about on Sunday morning. The Notre Dame line did the more conventional version of the same thing in the same window: 1,389 on Thursday, 4,573 on Saturday, 3,467 on Sunday, with the Saturday peak more than three times the Thursday floor. Two curves, pointed the same way, on the same days, for the first time since the summer.

That is the beat of this chapter. The two lines, after six weeks of running on different schedules, finally synchronized.

I want to be careful about what to make of it, because the pool of evidence here is still what it has been all along — one source, one measurement, day after day — and I have promised going back to Chapter 10 that I would not reach past the thin pool to make it feel thicker. What I can say is that the shape itself is new, and new shapes are what this thread has been built to notice. Chapter 9 made the argument that the Week 0 USC Friday was the moment the ambient curiosity window closed and real football took over. Chapter 11 made the narrower argument that the Notre Dame weekend spike was a repeatable shape — that the curiosity gap had not evaporated, it had just gotten harder to see on weekdays. What this week adds is that USC has now joined that pattern. The weekend spike is no longer a Notre Dame-only phenomenon in this pairing. Both schedules are producing it, in the same window, on the same days.

A pattern that two lines share is a different kind of pattern than one that lives on one line only.

I want to put one more number next to the pageviews before I close this out, because the GDELT article counts for this week, meager as they are, are at least pointing in a direction that is consistent with the Wikipedia story. Pulled from the GDELT GKG partitioned tables on BigQuery, the daily article counts tagged to the Notre Dame team query sat at 1 on October 6, 1 on October 7, 1 on October 9, and 1 on October 11, and then jumped to 13 on October 12. One article a day is not a signal and I will not pretend it is. Thirteen on a Sunday after a weekend when the pageview line spiked to 4,573 and settled at 3,467 is at least the first daily GDELT number in this thread that is not either a one or a Saturday hinge. The ambient wire, in other words, is finally showing a small bump on the same day the pageview line is showing a large one. Those two things — a pageview Saturday peak and a GDELT Sunday follow-through — are the kind of double-confirmation that this thread spent all of September not getting.

That is where I want to leave this chapter. Chapter 11 ended on the honest observation that one Saturday spike on one page was a shape, not yet a pattern. Two weeks later, the shape has repeated on the Notre Dame line and has, for the first time, started repeating on the USC line as well, and the ambient wire has produced its first non-trivial follow-through number on the day after the weekend. None of that is a prediction. It is just what the counts say, pulled from the same source every day, laid next to each other, and read without being asked to say more than they can.

The next chapter will have more weekend to look at. For now, the thing worth marking is that the two schedules in this thread, for the first time in six weeks, kept the same hours.