True Strength: The Belated Intro

This post is about the general functions and goals of the True Strength project. For more on the meat of how I got to the current formula, see the summary of the process.

As I said in an earlier post, I don't think a big introduction to this project was really necessary for much of the duration since these posts are mostly just standalone "How do ratings behave?" essays. This is my best effort at a somewhat understandable description of the project, though I might do some tweaking to this post as things develop and I think of other stuff.

What is True Strength (TRUE)?

You don't have to spend much time around TV ratings to realize that the raw numbers Nielsen puts out each morning just aren't good enough. There's a context beneath each rating. Maybe it had a crappy lead-in. Maybe it faced off against "Death Star" American Idol. Every day, PR people and real journalists and wannabe journalists like myself try to figure out and/or shape that context.

Much as I like doing that every day, I've long felt there was room for something even better: an objective way of quantifying all these external pieces of context. And thus was born the True Strength project: a quest for a formula that could input the rating and the surrounding circumstances and come up with a True Strength, or a number that objectively strips out those surrounding circumstances.

For example, a show that airs after American Idol on a Wednesday at 9:00 against four networks in repeats would post much higher raw adults 18-49 ratings than that same show airing on Friday at 8:00 against a bunch of originals, but it should hypothetically have about the same "true strength" in both situations.

What does True Strength adjust for?

As far as I can tell, there are four key things within the world of ratings that have a big impact on a show's ratings:

True strength: In other words, "how strong the show is in a vacuum." The true strength of a show should be relatively constant in a variety of situations, but then the other things below adjust the ratings further. To find "true strength," we have to be able to adjust out the big three factors below.

Viewing levels: This is the one I've been working on so far. A show's ratings are typically higher if there's a tendency for more people to be viewing TV at the time.

Competition: As I've explored a bit so far, competition is sort of an exception to the viewing levels rule. While heavy broadcast competition (and a very select few enormous cable events) tend to drive up the total number of people watching TV, it tends to drive down the ratings for other broadcast shows. We'll be starting into that one shortly.

Lead-in rating: Even in an age of DVR and on-demand and online viewing, lead-ins still have a meaningful impact on TV ratings. Given a much bigger lead-in, a show's ratings will assuredly go up as well.

What does True Strength NOT adjust for?

Network: It's possible that there might be an inherent advantage or disadvantage just from the network you're airing on, and perhaps there really are shows "on the wrong network." I just don't think there's a good way to quantify it. There are so few shows that actually switch networks, and even those are usually treated pretty differently.

Other demos: This system stays very strictly within the realm of live + same day adults 18-49 ratings. As I've shown in many a past post, it correlates pretty closely with ad rates. The only other stats that are available regularly (households and total viewers) have much more dubious correlations with ad rates. There are other ratings that contribute both to ad rates and renewal decisions, but those things are not just hard to consistently quantify but, more importantly, usually not even available to us randoms on the Internet.

The rest of the balance sheet: Basically the same explanation as the last one. This metric is purely "adults 18-49 ratings strength," not "economic strength." I am not connected enough to be able to incorporate DVD sales and iTunes sales and syndication revenue and production costs and licensing fees and all the other things that really determine the success of a show. I'm just trying to improve on the "adults 18-49 rating" and hope that still matters, at least for a while.

Multi-year: A possible future step for True Strength is to combine it with The War of 18-49, come up with a sort of expected trajectory for a show over the course of its entire run, and produce a sort of "era-adjusted rating" (sort of like the sabermetrics in baseball that amplify the stats of hitters playing in the "dead-ball era"). But we're almost certainly not going to get there this summer. So there should not be an expectation that a show should maintain the same "True Strength" each year across its entire run, just that it should more or less maintain it within one season. Year-to-year, True Strength will probably usually drop, much like raw adults 18-49 ratings.

What will True Strength DO?

Predict ratings: By the end of this project, we'll have a system that will convert a rating into a second metric. But it'll also be possible to reverse that, tweak the surrounding circumstances to simulate a big change a show will be undergoing and come up with an expected rating it should get in the new circumstances. How accurate will it be? Well, we shall see!

Create a true "level playing field": I don't have any real delusions that this project will be picked up by the mainstream, so it's not like this number would "end the debate" or create "bragging rights" even if it were a flawless indicator of "true strength" (which it won't be). But it will be neat to have something I can point to without having to suffer through making all the adjustments/excuses every time. I'm really looking forward to creating weekly "true strength" top 25 rankings and stuff like that.

Enhance further research: Unless this thing really heads south from this point forward, I look forward to adopting "true strength" for future projects (see the Spotlights for some general examples). How big of a deal will True Strength become on this blog? That's not something I'm totally sure about. I can say with a lot of confidence that it will be a "better" measurement than raw adults 18-49 ratings for a lot of the stuff I want to do. The thing working against it is that it's something new and unfamiliar, so the numbers will seem obscure to an outside observer. There's a real risk of shutting out anyone who finds my posts in a search when I use some number that I plucked out of a (sort of) complex formula. If you've read all my True Strength posts and buy the number as gospel, great, but if not, there's a barrier to entry every time I use it. Definitely a problem to think about going forward.

What will True Strength NOT DO?

Predict renewals: Now, I say this because my vision for the initial "true strength" metric does not have anything to do with the network it's on. But I have considered doing a "network-adjusted true strength" that would take a stab at predicting renewals. I think it might end up being a little better than the Renew/Cancel Index at TVByTheNumbers simply because it'd make some objective adjustments that their system doesn't (but they still make them subjectively). I don't think it would make any real earth-shattering improvements in that area, simply because there's not a whole lot of room for improvement (they get most everything correct as is), so that's not a top priority.

Correlate with ad rates: I feel this is important to note because I actually think True Strength will be significantly worse at lining up with ad rates than raw adults 18-49 ratings. Why is that? Because advertisers want to buy eyeballs. They don't want to buy "ability to make the best of its situation." If it's in a crappy situation and is able to rise above that and get OK ratings, it's still only gonna get OK ad rates, because it only has an OK audience. And I'm not saying that's something about the industry that should change. It shouldn't change. I just think there's room for a number that puts all these shows with very different external factors onto a level playing field. I'll reiterate: this number is about the strength of a rating relative to its situation (or perhaps ratings potential is the better way of saying it), not about dollar value in reality.

How "will" this work?

I put "will" in quotes because I'm about halfway through this already, I think, and I've already sort of had to hint at the two big principles already. But here are the two "ground rules" I've used thus far.

Control group: The biggest assumption I need for any ratings to really be "usable" is that the show whose ratings I'm using is of a consistent "true strength." If the show isn't about the same strength throughout the season, then where would I have a foundation for looking at anything? There'd be nowhere to even start making a conversion to "true strength." Most of the consistent true strength shows are veterans (in their second year or later) and relatively standalone (procedurals, comedies, non-serialized reality like Extreme Makeover: Home Edition or Shark Tank) to prevent heavy influence from storylines.

Big changes: This rule isn't quite as important as the first one (and I've already sort of broken in it in the viewing levels portion, as I saved the "big events" for last), but I think a great way of getting a sense of how much a given thing impacts ratings is to look at really big changes in that thing. If you just look at small changes in the thing, then you're also looking at small changes in ratings, and at that point there's a good chance (unless the sample size is huge) that there's a lot of short-term luck/variance involved. When ratings change in a big way, you can be surer that it's because of the change in whatever you're looking at.

What's coming up?

As of this going live on July 18, 2011, the next post sometime this week should pretty much wrap up the "viewing levels" issue (with perhaps one more after that to formalize the formula), and then we'll embark on the relatively connected "competition" issue (which shouldn't take nearly as long, since we've done some of the stuff already). Finally, we untangle lead-ins, which is probably gonna take a little while. At the end, I'll have a formula that can convert adults 18-49 into "True Strength," and I'm planning on doing one relatively short essay-type thing that links to every single post in the project and tries to briefly put it all into layman's terms.