As the 2014-15 television season hits the halfway mark this month, here are a few little-known facts about the major networks:

» CBS’ primetime lineup is drawing more viewers than it did in 2003.

» The premiere of ABC’s “How to Get Away With Murder” opened to a bigger audience than the debut of “Grey’s Anatomy” in 2005.

» Fox’s “Empire” after four weeks is on track to draw as large an audience as “House” at its peak in 2006-07.

» NBC’s fledgling “Chicago P.D.” is averaging 10 million viewers a week.

Wait a minute, you might say. How can these statements be true? They don’t square with the evidence that broadcast and cable network viewership has plummeted in the past few years.

But the Big Four networks have the numbers to back up each of these assertions. They’re just not Nielsen ratings — at least, not entirely.

Broadcast and cable networks alike are becoming more aggressive in disseminating viewership data that is cobbled together from many sources — from traditional Nielsen numbers to cable VOD-playback stats to a range of methods for tracking streaming activity.

The goal is to offer a more complete picture of the audience that is increasingly watching TV programming across multiple platforms. The great media migration of American consumers to time-shifted viewing options has accelerated to a gallop this season, and it shows no sign of slowing down, researchers say.

“This is real viewing. The growth and change of (consumer) behavior is just in the early days, and all of it is only going to increase,” said Alan Wurtzel, NBCUniversal’s president of research. “People are beginning to become adjusted to including digital (in TV ratings) because it is becoming more and more important.”

In an effort to combat the sky-is-falling perception that surrounds them, broadcast networks are increasingly going public with internally crafted numbers. These new metrics look at program performance over time frames of seven to as many as 35 days. The 30- to 35-day window is a key focus because it covers the typical time frame in which a new episode of a primetime series is available on ad-supported streaming platforms and on cable VOD.

The nets desperately want to expand the parameters of advertising sales, which are largely concentrated on the so-called C3 window, or commercial ratings that include up to three days of DVR and VOD playback. But the wider measurement parameters contained in this new, omnibus viewing approach are a long way away from being pitched as currency for advertising deals at the Big Four, despite inroads being made by the CW in selling combo linear/digital audience stats. Right now, researchers and the programming honchos they serve are still trying to get comfortable with television’s new math.

Illustration by Shout

“As these (metrics) become broader and more generally accepted, it is going to be central for us to have them vetted through organizations like the Media Rating Council,” said Will Somers, Fox’s senior VP of audience research. “It’s really empirical validity that we’re all going for at this point. We have to be able to report (viewer) information in a way that is representative of how people are actually doing the viewing.”

Nielsen at present doesn’t measure all of the viewing sources that can feed into a 35-day number, so research execs are devising intricate methodologies and patchwork quilts of data sources to arrive at what they believe are credible numbers. It’s all highly unorthodox by industry tradition, which is rooted in the standardization that came with having all ratings produced by the same Nielsen yardstick.

Each of the Big Four nets is trying to combat the focus on overnight ratings by issuing projections on a show’s performance in Nielsen’s live-plus-3 and live-plus-7 ratings. Key cablers, including FX, HBO and NBCUniversal’s entertainment channels, have stopped sending out overnight ratings altogether in a bid to shift the focus to L3 ratings as the bare minimum measure of a show’s commercial potential. Fox put the focus on the new methods of measuring multiplatform-viewing stats in a meeting held on the 20th Century Fox lot in September 2013 with reps from ABC, CBS, NBC as well as reporters from the trade press, including Variety. This season, the growing volume of publicly released data has execs hopeful that the Big Four, CW and possibly other nets will take the next step of sharing and comparing data on a consistent basis.

“The challenge for all of us in the marketplace for several years was that there wasn’t necessarily alignment on how each of the networks was talking about this in public,” said Brian West, ABC’s director of multiplatform research. “Now we’ve gotten to a place where there’s an accessibility to everyone’s understanding of multiplatform viewing. What we’re trying to push for is creating some consensus around this methodology so we can paint a picture across the industry of this viewing.”

ABC led the charge among broadcasters in making shows available for streaming, starting in 2006. But for a long time, the networks still weren’t able to interpret digital ratings the way that they pore over every tenth of a Nielsen rating point. The focus was on unique users and streaming starts rather than on tracking the average audience for any episode over a defined period of time.

“Until we were able to add up some of the online viewing and VOD viewing and translate it into a TV context, it was hard to put our hands around how the two compared to each other,” said Lisa Heimann, VP of multiplatform research for ABC and ABC Studios.

Certainly there are both PR and advertising-driven motivations for getting multiplatform viewing stats out on a regular basis. A look at live ratings for the Big Four networks over a 10- or five- or even year-over-year span is not a pretty picture. Collectively, live viewership for the nets has dropped nearly 30% since the 2008-09 season.

But the overnight ratings that were once the be-all and end-all for performance now capture only a portion of the audience for any given program. The amount of incremental activity that falls even outside the seven-day window is enough to make big hits bigger, and make mediocre L3-L7 ratings look respectable.

“It does call for adjustment on everyone’s part, but I don’t see any reason why we can’t get into a mode where we have a consistent ability to look at a show’s 30- to 35-day performance,” said Fox’s Somers. “We’d be remiss if we didn’t do that. If you look at those live-same-day and 30-day comparisons, it’s commonplace for shows to more than double. Live ratings don’t represent how the audience is consuming our shows.”

The nets are scrambling to come up with ways to milk more ad dollars out of the post-C3 part of the 30- to 35-day frame. CBS and Comcast have agreed to an experiment with “Madam Secretary” this spring that will see a new batch of ads inserted uniformly across VOD episodes of the show once the C3 window ends. (Part of the problem for Nielsen in tackling multiplatform ratings is that its intricate system of coding advertisements to generate commercial ratings is geared toward linear TV standards, not the new world order of delayed viewing and ad-switching.)

Research execs acknowledge that the best way to keep one another honest is to hold their work up to scrutiny by their peers. In the battle for bragging rights, nets won’t hesitate to cry B.S. if numbers appear inflated just enough to edge out a competitor for “No. 1” status in anything.

The Coalition for Innovative Media Measurement, an org founded in 2009 by major TV nets, media-buying agencies and advertisers, is working on protocols for multiplatform ratings. The end game is to develop a system that can be anointed as the industry standard by the Media Rating Council, the New York-based watchdog org that has governed all manner of media ratings since it was founded at the behest of Congress in 1964, to bring accountability to advertising sales.

Moreover, the need for a number that fully includes cumulative viewing goes beyond advertising concerns, according to David Poltrack, chief research officer for CBS Corp. With SVOD outlets growing in importance as off-network buyers of programming, and cable operators paying big retrans dollars in exchange for VOD rights, stats that show a program’s level of circulation, or its binge-worthiness, on emerging platforms are important metrics for those deals.

“The business is changing, in that the subscription side is growing relative to the advertising side,” Poltrack explained. “The total audience is becoming more and more important from a
revenue standpoint.”

The biggest challenge, research execs say, is managing the massive volume of data that pours in.

Rentrak provides numbers for cable VOD viewing that falls outside of the live-plus-3 window measured by Nielsen. Streaming numbers are particularly hard to corral. Nielsen offers online ratings that include such viewing on many sites, as do Internet measurement firms ComScore and Omniture. Congloms can also get hold of detailed data from Internet ad-serving firms such as Comcast-owned FreeWheel, which inserts and tracks online blurbs for clients. And each net has the proprietary data harvested from streaming initiated via its own websites and mobile apps.

Today, the trenches in the war to save the broadcast TV business model are packed with number crunchers armed with spreadsheets and finely tuned stats that measure digital demographics.

“We’re fighting the battle of patience,” West said. “Go back 10 years — you got all your ratings the next day. Today, if you want live-plus-7 ratings, you’re waiting almost three weeks. VOD numbers come out monthly. (But) we have our 35-day timeline pretty compressed. We’re building it as each stream of data comes in.”