Wednesday, May 12, 2021

Determining the League Average for the 2020-21 Season


As I wrote last summer, COVID-19 stay-at-home orders completely distorted the entire shape of the 2019-20 TV season, significantly increasing overall viewership and wiping out what would normally be a spring decline period. In response to the league average inflation, I created a second version of A18-49+ called "CVPlus," which used two separate league averages to  with typical spring declines.

In the 2020-21 season, COVID has remained with us, but the viewing inflations are pretty much gone. At least in terms of the external environment, the 2020-21 season is taking on a much more normal "shape," as shows have experienced their typical spring declines. Perhaps there could be some concern about even bigger spring declines, as the improving weather coincides with states lifting COVID restrictions, but this distortion is nothing like what we saw last year.

However, portraying the landscape in the 2020-21 season is still a unique challenge. This time, it's not about a change in the environment, but about a change in what the networks have actually aired. Due to production shutdowns, all the regular scripted options had their season premieres at least a month later than usual, and some delays were much longer. In place of these shows were a slew of filler programs, the likes of which we almost never see on the broadcast stage in the regular season. It's somewhat similar to the writers' strike year of 2007-08, which had several months completely dominated by unscripted programming.

So we don't need to use multiple league averages the way we did last year, because individual shows are behaving more normally. But we may need to make sure that the one 2020-21 league average is painting a fair picture. What do we do with the filler programming from the fall? Should it all be automatically included? In 2007-08, all of it was included, and (as I've acknowledged before) the LA that season was probably a bit deflated. Tinkering with what goes into the league average for 2020-21 might be setting a new precedent, so we should tread lightly.

What Goes Into the League Average?

The first thing I want to acknowledge is that there are a couple different kinds of filler that I was counting as "originals." Some of these shows are actual originals, airing for the very first time on broadcast, but they were low-priority leftovers or rush-jobs that would've never seen the light of day in a normal TV season. This category includes shows like the Fox dramas Filthy Rich and NEXT, maybe something like NBC's comedy Connecting..., and even unscripted filler like ABC's Emergency Call and Card Sharks that would usually be confined to the summer. I could drop them out of the league average, but it feels like a bit of a slippery slope to say, "I'm not including this in the LA because it flopped so hard." Where is the line? Do we start removing megabombs from seasons past? So even though they would not have aired in a "normal" season, I am hesitant to take them out.

The other type of filler was particularly unique to the 2020-21 season: the shows that had actually aired on some other channel or in some other country before premiering here. I gave a lot of these shows SpotVault pages, but I feel a little better about excluding these shows because it doesn't feel like it's fundamentally changing the age-old definition of "league average." They are not truly "originals," so they don't have to go in an average of "original, non-sports series." Please let me know if I'm missing any, but my current list of these exclusions is:

  • Manhunt: Deadly Games, Star Trek: Discovery, One Day At a Time (CBS)
  • Nurses, Transplant (NBC)
  • Cosmos: Possible Worlds, LA's Finest, Holmes Family Effect (Fox)

This type of filler is not exactly new to broadcast, but it is surprisingly rare that it shows up in the actual big four regular season, which is the only place it would affect the league average. Perhaps the most successful example was ABC's cop drama Rookie Blue, which aired 74 episodes on ABC (many after the Canada air date) but not a single one was part of a regular season. So I don't think this is gonna require me to pick back through history and make major wholesale changes. Maybe some of the CBS eps of Flashpoint should be removed? But that's about it.

Even in this season, removing them isn't a huge game-changer; they combined for 78 hours in a season where the league average currently has over 1400 with more than two weeks to go. The league average goes up by less than two hundredths without them. I think it makes for a more "fair" number, but there are still issues with the LA in this season...

What Does the 2020-21 League Average Look Like?

So here's what the "projected league average" table we do every year would look like, with those shows excluded. This is the same table that has been updated regularly on this site in previous years:

WeekEndingLAy2dy2dy2yProjLALAwkwky2yOrigHrsLYHrs
0


0.66*



19/27/20200.67-31%0.62*0.67-31%38.060.0
210/4/20200.70-24%0.61*0.74-15%32.064.0
310/11/20200.68-23%0.60*0.64-24%30.563.5
410/18/20200.69-21%0.590.71-12%34.064.0
510/25/20200.71-18%0.620.76-5%35.051.0
611/1/20200.71-16%0.640.74-6%48.052.5
711/8/20200.72-14%0.650.78-2%32.062.0
811/15/20200.73-13%0.670.78-3%46.556.5
911/22/20200.73-12%0.670.73-7%52.559.0
1011/29/20200.74-11%0.680.80+3%36.038.5
1112/6/20200.73-11%0.680.67-13%48.540.5
1212/13/20200.72-12%0.680.65-12%42.048.0
1312/20/20200.72-12%0.680.74-19%34.027.0
1412/27/20200.73-12%0.680.82-10%16.027.0
151/3/20210.72-12%0.680.61-12%11.06.0
161/10/20210.72-11%0.680.70+7%38.020.0
171/17/20210.72-13%0.680.69-26%44.057.0
181/24/20210.71-13%0.670.65-18%44.050.0
191/31/20210.71-13%0.670.65-16%47.548.0
202/7/20210.71-13%0.660.69-17%36.546.0
212/14/20210.71-14%0.660.66-24%57.543.0
222/21/20210.70-14%0.660.64-19%58.565.0
232/28/20210.70-15%0.650.61-22%50.564.0
243/7/20210.69-16%0.650.60-27%55.557.0
253/14/20210.69-16%0.640.64-20%52.551.5
263/21/20210.69-16%0.640.66-24%46.062.5
273/28/20210.68-17%0.640.55-41%52.059.5
284/4/20210.68-18%0.640.60-35%53.559.5
294/11/20210.67-19%0.640.58-36%57.546.0
304/18/20210.67-20%0.640.55-38%53.543.5
314/25/20210.66-20%0.640.59-26%49.053.0
325/2/20210.66-20%0.640.51-30%48.545.5
335/9/20210.65-21%0.640.54-23%59.044.5
345/10/20210.65-21%0.640.56-27%9.010.0

So I think this illustrates a few points better than I could by just explaining them. The first big one is that the projection (ProjLA) is based on how the league average developed two years ago and three years ago. Last year is thrown out completely because of the aforementioned lockdown craziness. You can really see this in the spring ratings, where the projection has been steady even as the y2y numbers have tanked. The projection is expecting the spring decline, rather than the weird steady period from last year, so all the y2y ugliness has not really budged the projections.

Second, I hope it's obvious why I was scared to have a continually updating projected league average all season... because of the production delays, the data from the fall part of the season was so off-kilter that the projection was quite volatile, even using my 2019-20 formula which continues to weigh the preseason projection heavily in the early weeks. The numbers were very ugly early on, especially in premiere week. Then, much of the fall, right after more normal programs came back, was actually quite healthy from a y2y standpoint. However, a big part of the key was in the "OrigHrs" column, where you can see that there is just so much less programming going into the league average than a year ago. So while the averages in that period may have looked good, that period ended up not counting much once we had a winter with typical original volume.

One last thing I wanted to do to see if the league average is "OK" was to check how the league average decline compares with the actual individual shows themselves. So the "Show Median" column is just looking at the median y2y number among the big four shows on the Daily Year-to-Date Rankings.

YearLAShow Median
2011-12-7%-14%
2012-13-11%-16%
2013-14-11%-14%
2014-15-11%-12%
2015-16-13%-17%
2016-17-16%-17%
2017-18-11%-20%
2018-19-13%-20%
2019-20-14%-20%
2020-21-21%-24%

Every year, the median returning show tends to do a few points worse than the actual league average decline. The LA decline might creep up to -22% by season's end, but it's possible the Show Median one will also go up as more shows get compared against the inflated spring ratings from last year. Either way, the gap is gonna be a bit smaller than it has been in recent years. That may have something to do with the new shows hurting the LA (we got stuff like Filthy Rich and Connecting... in the fall when we would usually have "normal" hyped shows), but it may also be an indication that the "true" league average decline should be less than what I'm calculating. But I don't think it's to an alarming degree, and if anything it makes me feel better about removing the "repeat" filler from the LA (because the gap would be even smaller if they were included).

Anyway, that's just a little preview of what it will look like. I may or may not put out the usual Climate Center posts between now and the end of the season, but I'm doing this just so you don't get shocked when a 0.64 shows up as The Breakdown begins this summer.


No comments:

Post a Comment

© SpottedRatings.com 2009-2018. All Rights Reserved.