Tuesday, June 28, 2011

True Strength: Viewing Levels 102 - How TV Events Affect Viewing


Seemingly about 20 years ago, I looked at some of the basics of how many adults 18-49 watch TV during primetime. Ultimately, we want to look at how a program's rating is affected by the level of viewing. But before we look at how viewing affects ratings, we're gonna look at the reverse: how do ratings, specifically big events with really big ratings, affect viewing levels?*

*- Why big events with big ratings? I'd have noted this in the intro if I'd managed to get a good version done, but most of these posts over the course of the summer will revolve around big changes. I do that because big changes make things more obvious. Trying to figure out why the viewing went from 35% to 36% in a given week seems almost impossible to nail down, if it really even happened at all (margin of error!). The big events tell much clearer stories. (And the big ratings/shares mean much smaller margins of error.)

Untangling the causations between "viewing levels" and "ratings" and vice versa aren't trivial things to do. It's easy to look at viewing levels as just some vague measurement that comes out of nowhere, but we've got to keep in mind all of the 30-however-many percent of the adults 18-49 viewing at a given moment are watching something. Each "viewing level" is actually a combination of umpteen show ratings, and when something really big happens in the world of TV, all those people have to come from somewhere. In a world without that big event, most viewers would've been watching something else on TV, and some wouldn't have been watching any TV.

VIEWING LEVELS 102 - CHANGES IN VIEWING

For purposes of this exercise, we'll break a big TV event's audience into two types of viewers. 1) The "Thin Air Viewer" is a viewer who would not have normally been watching any TV in a timeslot but turns on the TV to see the event. So they arrive and watch the program, effectively, out of "thin air."* 2) The "Taken Viewer" is a viewer who would have been watching other TV in a timeslot but instead flips to the event. So the event "takes" that viewer away from some other program.

*- Worth noting, though not terribly meaningful: the Thin Air Viewership is always a little higher than the below figures, because there's always a small subset of viewers who would watch whatever the network would "regularly" program but instead shun both the event and everything else on TV. But for the sake of simplicity, we'll just count those who turn off the TV and those extra who turn on the TV as cancelling each other out.

So what's big enough to actually cause a noticeable increase in overall viewing? I basically count the NFL, big awards shows and American Idol. So how much does viewing go up when those air? It's kind of tough to really know, but here's my best effort:

Event Rtg Reg PUT ThinAir %Thin
Super Bowl 39.9 35.2 51.6 16.4 41%
AFC Champ 19.7 35.2 43.2 8.0 41%
Oscar 11.1 36.5 39.2 2.7 24%
BCS Title 10.1 36.5 40.9 4.4 44%
Grammys 10.0 36.5 37.3 0.8 8%
Idol Wednesday 7.6 34.2 36.3 2.0 27%
Sun Night FB 7.4 36.5 39.5 3.0 40%
Idol Thursday 7.2 32.4 34.8 2.4 33%
Mon Night FB 5.8 36.5 38.0 1.6 27%
Globes 5.2 36.5 38.3 1.8 35%

"Rtg" is the 18-49 rating of the event in question. "Reg" is the 18-49 viewing level in the event's timeslot under "regular" circumstances (in other words, when the event is not on the air). "PUT" is the 18-49 viewing level in the timeslot when the event is on the air, "ThinAir" is "PUT" minus "Reg" (or the number of "Thin Air Viewers," AKA how much the event increased the viewing level) and "%Thin" is "ThinAir" divided by "Rtg," or how many viewers of a big event are Thin Air Viewers. For example, the Super Bowl got a 39.9 adults 18-49 rating. The 18-49 PUT level in that timeslot (approximately... I counted it as 7-10pm) is 35.2% under "regular" circumstances. On Super Bowl night, it was 51.6%. That means there was a 16.4 percentage point increase in 18-49 viewing in the approximate Super Bowl timeslot; in other words, the Super Bowl created a 16.4 rating out of Thin Air. And it took the remaining 23.5 away from other programs.

*- The particularly observant might notice that some of these numbers don't look exactly right (for example, the Idol Wednesday rating includes some smaller stuff like Human Target and Breaking In just so I can do entirely 8-10pm comparisons), but it would probably take thousands more words to explain everything about that table, so let's just say all the fudgery is in an effort to reduce the error in the comparisons as much as possible. Any specific questions, feel free to ask.

One key observation here: not all big events change viewing levels equally. What sticks out most obviously is that sporting events drive viewing levels up the most (as a percentage of the event rating). The above sports events, on average, are around 40% Thin Air Viewers, a higher percentage than anything else.

Why is that? If you know anything about the demographics of sports, it's not that hard to figure out. The sports audience is much younger and much more male than the general primetime audience. Compare that with awards shows and American Idol, where you have an audience that looks much more like the general primetime audience and much more likely to be interested in entertainment programming (award shows actually sort of require an interest in that kind of stuff). So the big ratings those shows get are noticeably more likely pound-for-pound to "take" viewers from other programs rather than from Thin Air. And if we could break it down further we would probably see that even a lot of the Taken Viewers would not necessarily be watching originals on broadcast. That is a big part of why a show like Desperate Housewives can thrive all fall against football but then gets clobbered in the spring when it faces relatively similarly-rated awards shows.

Anyway, the only regular entertainment program that I really feel is big enough to use is American Idol, and it seems in both cases like that is about 30% Thin Air Viewers and 70% Taken Viewers. That's more Taken Viewers than sporting events, and it shows in the dips many programs take as soon as Idol arrives. I think, factoring in how the awards shows fared, it's probably a pretty good baseline number to use for original entertainment programs. Presumably it would be much lower for repeats, as I don't think nearly as many people show up out of thin air to watch those.

A little down the line, we'll look at how competition affects ratings, and it'll help having some sense that more or less about 70% of the competition's ratings actually dip into the potential audience pool. We'll see if I actually need to use that number.

I'm gonna end this post, as it is becoming clear I will never finish this if I try to make each post a Homeric epic. So I'm basically just gonna try to make one point per post rather than store up some fully formed lecture. We may have to get a bit more in-depth on the subject discussed herein, but next time we'll get into the flip side of this post, which is how "non-TV events" (like holidays and weekends) affect viewing. Exciting...!

No comments:

Post a Comment

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