Hello!
It's late July and as you may have noticed, my usual summer marathon review of the previous broadcast TV season has not yet commenced. I really only have myself to blame for this, but I do have a couple excuses as well. I had to port everything over to a new computer during the spring, and so I had to make a bunch of changes to adapt to new versions of the program and the packages I use for coding. I was back up and running pretty quickly for the daily tables, but Schedules Plus and The Breakdown were more complicated beasts.
And before I could start on the usual summer routine, I had to make one of these posts describing the landscape in general. As predicted last summer, the writers' strike made 2023-24 the most heavily disrupted broadcast TV season of the 25 in my archive. And these posts always take awhile. But here we finally are! I'll get into the logistics for this summer's The Breakdown at the end, but first we'll dive right into what made 2023-24 so crazy, and what it means for evaluating the shows that aired this season.
We'll start by looking at the number that drives all of our A18-49+ metrics, the "league average" (the Live+SD adults 18-49 average for all original, non-sports series on the big four networks during the broadcast regular season). Here are the trends over the last decade.
| Year | LeAv | y2y |
|---|---|---|
| 2014-15 | 1.68 | -11% |
| 2015-16 | 1.46 | -13% |
| 2016-17 | 1.22 | -16% |
| 2017-18 | 1.10 | -11% |
| 2018-19 | 0.95 | -13% |
| 2019-20 | 0.82 | -14% |
| 2020-21 | 0.64 | -21% |
| 2021-22 | 0.51 | -21% |
| 2022-23 | 0.43 | -16% |
| 2023-24 | 0.38 | -11% |
By this metric, -11% was the healthiest season for broadcast TV since 2017-18, and quite a return to form after what seemed like a sharply worsening trend from 2019-22. (As we said in 2020, the 2019-20 season was headed for a 20%ish decline and was only rescued by COVID-boosted viewing levels in the spring.)
Scripted vs. Unscripted
But what's going on under the hood? The simplest way to approach this is to look at the scripted vs. unscripted breakdown, a sensible approach since we know that the writers' strike wiped out most original scripted programming for about two thirds of the season. Here's how that shapes up in terms of all the hours counted in the league average each season:
| Year | Unscripted | Scripted | Scripted% |
|---|---|---|---|
| 2014-15 | 635.0 | 1188.5 | 65% |
| 2015-16 | 642.0 | 1249.0 | 66% |
| 2016-17 | 580.5 | 1279.0 | 69% |
| 2017-18 | 595.0 | 1207.0 | 67% |
| 2018-19 | 622.5 | 1159.5 | 65% |
| 2019-20 | 651.5 | 1063.5 | 62% |
| 2020-21 | 835.0 | 754.5 | 47% |
| 2021-22 | 809.5 | 923.0 | 53% |
| 2022-23 | 773.5 | 970.0 | 56% |
| 2023-24 | 1001.0 | 470.5 | 32% |
For a decade-plus before this table starts, you could pretty much count on the league average being right around 60% scripted/40% unscripted, and that number actually grew into the mid- to upper-60's for much of the period in this table. That was derailed by the COVID production-delayed 2020-21 season, and while there was some recovery in the two years following, it didn't quite get back to normal. Then came a shocking drop in the strike-shortened 2023-24 season, with less than one third of the league average consisting of scripted programming. So what does that mean in terms of ratings? Here are the Plus averages for these two categories:
| Year | Unscripted | Scripted |
|---|---|---|
| 2014-15 | 104 | 98 |
| 2015-16 | 106 | 97 |
| 2016-17 | 109 | 96 |
| 2017-18 | 107 | 97 |
| 2018-19 | 110 | 95 |
| 2019-20 | 116 | 91 |
| 2020-21 | 108 | 91 |
| 2021-22 | 106 | 95 |
| 2022-23 | 110 | 92 |
| 2023-24 | 104 | 92 |
In the last decade, unscripted programming on average is consistently stronger in Live+SD than scripted programming. That very much held true even in 2023-24 when unscripted was shouldering much more of the real estate. The Plus average for the entire category was lower, since there was a lot more "spackle" type stuff going into the unscripted average, but in general more unscripted programming means a "better" league average.
Scripted Returnees vs. Newbies
Still, you might be thinking that scripted programming must have done OK if it still managed a 92 average, despite the big guns missing out on the first five months of the season. But I wanted to drill just a bit deeper, to try to get a sense of what's going on with returning shows. The following chart is breaking out just the scripted shows into returnees and newbies:
| Year | Returnees | Newbies | Newbie% | Returnee+ | Newbie+ |
|---|---|---|---|---|---|
| 2014-15 | 778.0 | 410.5 | 35% | 101 | 92 |
| 2015-16 | 838.0 | 411.0 | 33% | 104 | 81 |
| 2016-17 | 877.0 | 402.0 | 31% | 97 | 92 |
| 2017-18 | 904.0 | 303.0 | 25% | 99 | 91 |
| 2018-19 | 850.5 | 309.0 | 27% | 98 | 87 |
| 2019-20 | 794.5 | 269.0 | 25% | 97 | 72 |
| 2020-21 | 624.5 | 130.0 | 17% | 95 | 75 |
| 2021-22 | 712.0 | 211.0 | 23% | 101 | 75 |
| 2022-23 | 755.5 | 214.5 | 22% | 95 | 78 |
| 2023-24 | 394.5 | 76.0 | 16% | 93 | 83 |
This one isn't as dramatic, but there's a similar effect going on with how few new shows there were. While the volume of scripted content was down 51% from last year, that was -48% among returnees and a whooping -65% among newbies. Putting the two charts together, scripted newbies were a mere 5% of the total league average. This number was 8% in the COVID-disrupted 2020-21 season, and 12% in the other two years since, but historically it has usually hovered around 20%! It's been a fairly precipitous decline across the whole decade, and this year only accelerated it.
As a general rule, the returnees do better than the newbies, so having so few scripted newbies will tend to help the scripted average and the league average as a whole. It was also (at least by recent standards) a good newbie class ratings-wise, largely carried by CBS' Tracker, which helped mitigate its damage to the average.
Full-Season Averages vs. Episode-By-Episode
Finally, I wanted to look at the individual returnee trends and see if there was evidence that they were actually healthier this year, in line with the healthier league average. I started doing this by looking at the y2y numbers listed on the yearly indices, but I quickly remembered that these are episode-by-episode counts; in other words, if a show only aired eight episodes, the "y2y" numbers on those pages will only compare vs. the first eight episodes of the previous season. The idea here was that comparing episode-by-episode is more predictive of where the final y2y trend will end up, since the season average tends to decline.
However, when we do The Breakdown, we're comparing full season averages, regardless of how many episodes are in the order in both seasons. Usually this won't be a big issue, but what about in a strike-shortened season when the scripted returnees overwhelmingly have far fewer episodes than last year? Only three of the 42 returnees aired the same or more episodes this year, and the vast majority (31 of 42) aired 8 to 12 fewer.
So I decided to chart the median decline in full-season averages ("full") vs. the median decline just compared vs. just the same episodes last year ("rolling"), and threw in the league average declines for good measure ("LeAv").
| Year | LeAv | full | rolling |
|---|---|---|---|
| 2014-15 | -11% | -15% | -15% |
| 2015-16 | -13% | -21% | -22% |
| 2016-17 | -16% | -23% | -24% |
| 2017-18 | -11% | -22% | -22% |
| 2018-19 | -13% | -21% | -22% |
| 2019-20 | -14% | -22% | -23% |
| 2020-21 | -21% | -26% | -25% |
| 2021-22 | -21% | -21% | -21% |
| 2022-23 | -16% | -21% | -21% |
| 2023-24 | -11% | -15% | -20% |
The big takeaway here is that, indeed, this is the first season where this distinction has really mattered. The only time in the last decade it was more than a one-point difference was this season, when it was a whooping five points. In other words, shorter seasons tend to boost full-season averages.
What this means is that, while the league average may be inflated, quite a few of these full-season averages will also be inflated, since they didn't have the second halves of their seasons in which to keep dropping. For that majority of shows that had 8-12 fewer episodes, comparing inflated full season averages vs. the inflated -11% league average decline won't be too far off. And we should be a little more generous with shows that were much closer to even in terms of episode orders. This lesson is perhaps most insightful for the unscripted shows that were not strike-shortened.
The "rolling" is more in line with the historical trends, but it's also worth noting that these individual episodes almost all occurred much later in the season than their counterparts in 2022-23. It's possible they overachieved because absence made the heart grow fonder, but it's also possible they'd have done better airing in their usual part of the season. It actually seems likely that the scripted environment was a bit healthier than usual this year, but probably not quite as much as the improved league average trend would suggest.
The Plan Going Forward
So the goal here is to speedrun the entire season of The Breakdown up in the next six weeks, covering one day per week and one network each weekday starting on Monday. Due to the time crunch, it's probably a safe bet there will be less text on these posts than usual, and I will not be doing grades on everything I have in the past. I don't really want to waste brain cells trying to judge random spackle from the fall/winter periods, so the plan for now is to do all the scripted shows and only the staple unscripted shows that already have long ratings histories to draw back on. If you think those other shows deserve more commentary, feel free to add it in the comments! I might try to squeeze in a handful of new War of 18-49 posts, but more likely will just wait on those till later (fall/holidays/next summer).
I had initially thought about dropping the CW or consolidating it into one post, since they aren't really running like what I consider a normal broadcast network anymore. But it quickly became clear that setting up a whole new one-post format with only All American and Walker would take much longer than just spitting out the posts in the existing format, so I think I will just put all the CW posts out but likely have no text at all unless something strikes me as particularly interesting.
See you Monday!
No comments:
Post a Comment