CharlieK

Nielsen ratings are issued every single day. They're tablualting viewing patterns among multiple people and multiple TV sets within a household at 1 minute increments for 1,440 minutes a day for 83 cable networks, 11 broadcast networks, PBS, and hundreds of local affiliates. Even in this day and age, that's a hell of a lot of data to go through every day for the sample they have, let alone the 284 million people you're suggesting.

Aside from that, there is no need to survey that many people. When you have a hearty enough sample size, to increase it would just be a waste of time and money. Statistically speaking, 10,000 is a large sample for the country. Political polls don't use samples that large.

And to compare dialing into 'AI' and all that Nielsen does is not a good comparasin at all. First, they release their results once a week, not daily. Also, the way 'AI' tabulates their votes is hardly more complex than the polling software on this forum.

I assure you, demographically speaking, you're represented in the sample and that's all that advertisers care about.

I'm not sure what you're getting at here. This sounds like you're saying that because you are not watching 'AI', that it shouldn't be the #1 show and that makes the methodology flawed. Millions and millions of people watch the show every week, you're just one guy.

Dr. DVD

This is why I like DVDs of TV series. Heroes is something I need to watch marathon style, like reading a graphic novel.

mrpayroll

That's because they've been around since the beginning (at least the 50's) and it's hard for the 'new kids' to break in.

I always thought that basing ratings on just a few thousand households was fine for the 50's thru the 70's, but with the advent of computers and the Internet, there has to be a better (more accurate) way to measure viewership.

I wouldn't be surprised if their figures every week are so far from the truth that if the real figures came out, there would be major upheaval in the television / advertising industry.

Look what happened in LA around 10 years ago. Arbitron (which measures radio audience) finally changed the way they measure Latino listeners and guess what, since then Latino radio stations have been at or near the top of the ratings every quarter.

Chris

mrpayroll

It's too bad you missed (and I forgot to post it here until it was too late) the mini marathon the Saturday before it came back. They showed the last 3 episodes and it was a great way for me to catch up. And it looked great in HD!

Chris

mrpayroll

Well then, as I said, they would have to break up the tv season into 3 1/2 seasons.

Fall (Sept thru late Nov sweeps) Comedies, non serialized dramas, Reality shows

Winter (late Nov thru early Jan) mostly repeats of comedies and non serialized dramas, some Reality shows and trying out new shows.

Spring (Jan thru the end of May) Serialized dramas and the rest (sweeps Feb & May)

Summer (June thru Aug) The usual stuff that's been on and it's been pretty good the last couple of years.

What does everything think of that idea?

Chris

mrpayroll

Please see my post above this regarding Arbitron. All those years Arbitron was reporting the amount of listeners in the LA area thinking they were pretty accurate.

When they finally updated their reporting method to include a larger, more accurate representation of the population, all of those Latino stations jumped from the bottom to the top of the rankings, all in one quarter.

I remember it created a pretty big stir here in LA radio.

Chris

CharlieK

Cable operators don't keep that kind of data. What good would it do them? They don't know who exactly is watching. All cable companies know is where you live and whose name the account is under. What are they supposed to do with such limited information. None of this is based on general viewership of a show, its very demooghraphically oriented. If they don't have specific info about you, its worthless.

Peep

I think it's funny when people comlain about DVRed shows not showing up in the ratings. The rating aren't really there to show how many people watch the shows, they're they to show how many people might be watching the ads. Since most people with DVRs skip the ads, it's OK for advertisers to get numbers that don't include DVRs.

Ratings are an advertising tool, primarily.

Rogue588

I'm betting there are some politicians that would like the way you think. I guess i'm silly for thinking that every person should be counted. I mean, nowadays we can fit 14,000 songs into a device roughly 3 1/2" x 2 1/2". We could also watch movies and TV shows on said devices AND it's also possible to watch videos on CELL PHONES. You're right. I'll stop with the crazytalk about having an accurate (read: TOTAL) accounting of viewers for any given television show. I mean, at the very least, households should be allowed to set up an account online (or something similar) to make sure their viewing habit is accurately represented. If the household doesn't have internet access, well, i'm sure something else could be invented. And if they don't want to participate, so be it.

And I assure you, that until i'm allowed to log the shows that I watch, then I am NOT represented.

Not discounting the "millions and millions" of people who watch AI, but i'm sure there are more "just one guys" out there..

CharlieK

That's the key! If its large enough and representative enough, the methodolgy is sound. It sounds like Arbitron screwed up.

I know that Nielsen recruits for its sample based on the the latest information from the Census department so they can make it as representative as possible.

Also, although I work in the TV industry, I have no affiliation with Nielsen.

pinata242

Maybe they don't know specifically what *I* watch, but they know how many viewers any particular program has.

It would help them gauge interest in their channels and tiers/packages. How to best adjust pricing and offerings based on viewer demographics.

They may not have a breakout of what each individual customer/box watched, but they can certainly benefit from aggregate information.

wildcatlh

Neilsen's ratings are not gospel. Like any other poll that uses a representative sample, they're subject to a significant margin of error.

Considering that most people nowadays use some kind of box for their cable or satellite TV, and that box regularly communicates back to the cable/satellite company, whatever data they're keeping now, those could certainly be used to calculate a much larger sample that lowers the margin of error. Studying the viewing habits of all 300+ million Americans is impractical, yes, but you could very easily up that 10k to 1-2 million.

CharlieK

You know who else likes the way I think? Every single person who has ever done any kind of research! Survey research is all about sample sizes. You never have the opportunity, time or money to talk to everyone you want to. Please give me an example of some research on a large scale that has interviewed every single person they're doing research on. I would love to see how they did it.

As with your 'AI' / voting example, this just isn't comparable to how they're able to retreive the information. I mean, since they can do all of the things you just mentioned, how come I can't teleport to Europe?!?!?

This idea is even worse than the one you hate so much. A self selecting sample? That won't give you any kind of accurate read. The logistical problems with your idea are also insurmountable at this time.

You are absolutely right. There are more people not watching 'AI', than are. However, there is no show that is seen by more people than 'AI', and since we are talking about TV viewership, that's all that matters.

Groucho

I remember a couple years ago, USA Today started listing viewing figures provided by Tivo. What ever happened with that?

mrpayroll

You can't?

Chris

CharlieK

I'm not familiar with what you're referring to exactly. I do know that Tivo and Nielsen did a study together a few years ago that compiled data from roughly 10,000 Tivo owners. They were looking at time-shifted viewing of shows and how people watched commercials. If you own a Tivo brand DVR, then they have the ability to see exactly how you're watching your TV.

There was talk that Tivo may start a service similar to Nielsen's, but with only 15% of the country owning a DVR, and not necessarily a Tivo, that's a long way from being legitimate. (Its all about sample size. )

Right now Nielsen is able to calculate how many people watch TV 'live', and 'live + 1 hour' through 'live + 7 days' and practically every increment in between. 'Lost' is one of the most time-shifted shows right now. Since the ratings we see on this board are 'live' ratings, that can partially explain why we're seeing a decline in that shows ratings. People are just watching it at a later time.

CharlieK

The networks only put on shows they think will attract the most desirable viewers to advertisers. They rarely have the loyalty to a show that viewers do.

Breakfast with Girls

If I knew that watching the advertisements for shows that I like would help them stay on the air, I wouldn't skip the ads. I would only make that effort for a handful of shows (Heroes, The Office, Lost, Battlestar Galactica) and shows that I watched that were canceled (Veronica Mars, Day Break, Raines, Stargate: SG-1, Andy Barker: PI).

Since the television industry is based on ad revenue, it seems that measuring the ads that are watched would work best. And if people knew this, and could opt-in on a mass scale to watch ads for shows they liked, you might see shows with dedicated fan bases not get canceled so quickly.

Rogue588

Of course they think the way you do - why fix it if you think it ain't broke, right? This way we'll keep getting shows that pander to the lowest common denominator (or someone who's not as smart as a fifth grader). This ain't rocket science. It's done every week by the MPAA and RIAA. I'm surprised technology can't handle this "massive" task for network television.

I look at it now as an extremely vicious cycle.

Networks use them to measure how many people are watching...which determines ad rates and which shows advertisers advertise on...which determines the shows that nets keep producing and spending money on...which is determined by ratings...which are what networks used to measure how many people are watching...etc, etc...

Tracer Bullet

VM hasn't been canceled yet. Don't give me a heart attack.

CharlieK

Are you saying that the vast majority of survey research ever done is suspect because they didn't talk to every member of the group they were surveying? If I ask America what flavor of ice cream they like best - chocolate, vanilla or strawberry - you think I would get a significantly different result if I ask 10,000 people or 280 million? After a while you're going to see a pattern that will carry through the rest of the group.

And with regards to the MPAA & RIAA, they can quickly tell you how much a movie made or records sold. But that's about it, and that's not too difficult to do. Have you ever heard of CinemaScore (I think) the company that reflects the audience's reaction to a movie with a letter grade? Do you think they ask every single person that saw that movie? Of course not. They ask a certain number of people, then extrapolate. Nielsen does the same thing. To ask every single person is impractical and unneccesary.

As for pandering, here are the top shows last week. In my opinion, there are maybe 5 or 6 shows out of 30 that I would characterize as such. Most of these shows are smart, mature, well-written entertainment. Is the system really that broken? I think this is a great time for television. Can you not find enough good things on?

American Idol - Wednesday (Fox: 26.93 million), American Idol - Tuesday (Fox: 26.55), House (Fox: 20.81), CSI (CBS: 20.20), Grey's Anatomy (ABC: 20.08), Dancing With the Stars - Monday (ABC: 18.87), Desperate Housewives (ABC: 17.17), CSI: Miami (CBS: 17.03), Dancing With the Stars - Tuesday, 9 p.m. (ABC: 16.56), NCIS (CBS: 14.17), Shark (CBS: 14.10), Survivor: Fiji (CBS: 13.83), Without a Trace (CBS: 13.71), The King of Queens (CBS: 12.63), Two and a Half Men (CBS: 12.27), Cold Case (CBS: 12.14), Heroes (NBC: 11.96), Lost (ABC: 11.86), CSI: NY (CBS: 11.40), Brothers & Sisters (ABC: 11.15), 60 Minutes (CBS) and Deal or No Deal (NBC: 10.98 each), Extreme Makeover: Home Edition (ABC: 10.73), 24 (Fox: 10.41), Numb3rs (CBS: 10.15), Are You Smarter Than a 5th Grader (Fox: 10.13), The Unit (CBS: 9.83), Criminal Minds R (CBS: 9.75), Ugly Betty (ABC: 9.62), Boston Legal (ABC: 9.54)


What you describe as an extremely vicious cycle, I see as a business. If you're spending money on something, you need to earn money as well. Some type of ratings system is how all of that is decided, and it always will be. Besides, you're already getting it for free, what more do you want?

lukewarmwater

I totally hijacked this thread. Awesome.

Rogue588

I'm guessing all the responses where I say the ratings system in place now is antiquated and needs to be replaced to account for ALL television viewers (or at least the ones that WANT to be accounted for) didn't explain what more I want?

And in case you didn't catch it...yes, the system is broken.

Now if you'll excuse me, I have to teleport over Europe..be back in a minute or two.

DJariya

Is this thread trying to imply that every serial drama needs to start in January in order to not lose viewers?

CharlieK

Oh, I caught it. I'm just throwing it back. If this system is broken, then all of survey research is broken. Your ideas for change will never happen because it is impractical, unneccesary, and a waste of resources, so get used to disappointment. Good shows get cancelled all the time because not enough people watch them. That doesn't mean the system is flawed, it simply means from the network's view not enough people watch them.

From a research standpoint, your theory that you would get a more accurate read on all TV viewing by only talking to those who opt in to some kind of system is laughable. A far more diverse population watches TV than would ever volunteer to be in your sample. All you'd get with your methodology is what a bunch of white, middle class males watch.