Back before cable spread into the wilds of Vermont, I was soooo ready to start purchasing whatever Star Trek series happened to be airing in syndication straight-from-video. I was tired of not having the choice to see shows I would otherwise have eaten right up.

I'd much prefer purchasing my shows directly, rather than relying on networks or cable stations to feed them to me when they decide the correct amount of viewers are interested in the show.

Where in the pipleline is that proposed 24: DTV series? Did that ever get off the ground or did it get axed? If any series could go the DTV route and sell a ton I think it would be 24. The format just lends itself to home video.

I believe he is working on a movie direct to dvd.

Actually it does, as of now Fox has only commited to airing 13 of the 35 episodes that are being produced. They have stated that if the ratings aren't that good the series would go straight to DVD and not even air all the episodes.

Fox really blows. I have said it before in other threads, but I will say it again here: Why are they (and other networks) still clinging to an old outdated ratings system like Neilsen? Maybe Neilsen was relevant back in the 60's and the 70's before there were more than three channels for people to watch, but you cannot ever convince me that having ratings boxes in a few thousand homes accurately determines how many people watch these shows. I know more people than not who watches both Family Guy and American Dad every Sunday night, and I am betting others are in similar situations. Yet none of those people have Neilsen ratings boxes, so their watching the show doesn't even count.

I guess Fox doesn't blow. It blows that somebody hasnt' come up with a better way to rate TV shows.

Exactly. I would love to see how many people are actually watching Arrested Development. It has a cult following. Not to mention everytime you see some interview with any celeb that asks what their favorite show is, what they are watching on dvd, or what show would they love to be on, more often than not(that I've seen), it is Arrested Development. Nielsen has definitley blown it on this show.

This is great idea waiting to happen when it's been implemented with some well thought out ideas . I know for a fact there is a good million roswell fans that would eat up new eps on dvd, even if it had a forced doritos menu or ads before it to get the cost down.

I don't know that working on is really the right status for the project. He never mentions it in his blog that I have seen. It certainly is one of the projects he has his name attached to, but I have learned that you can't count too much on those from Smith or any other artist. For instance, it sounds increasingly like Fletch Won is on the verge of either getting new talent attached to it or having nobody set to make it ever after years of talk about Kevin making it.

Actually, statistical theory has consistently shown that a sample size can accurately reflect the rest of the population if it is appropriately selected. Typically, the more random the sample, the better. Also, a sample size can be quite small and still be accurate, as long as it's not too small.

From
http://coolschool.k12.or.us/courses/...amplepoll.html
Neilsen also augments its rating box polling with other forms of polling as a check against any errors.

More info on Neilsen here:
http://www.nielsenmedia.com/whatratingsmean/

That's an anecdotal sample. Your personal group of friends is in no way random, they are more likely than not to share your tastes than others. Just because eveyone you know watches the show doesn't mean you can accurately extrapolate that out to reflect the population as a whole.

Well, I guess if it's such a statistical certainty, we can just take samples of the electorate in the next national election and everyone else can just stay home.

Music fans can make their voice heard when they buy music. Movie fans can make their voice heard when they buy a movie ticket. Book fans can make their voice heard when they purhcase a book. How do TV fans get heard? A few are asked their opinion and everyone else has to live with the consequences. No one will truly know how popular or unpopular a series is until everyone who watches it can be counted.

A cult following, by definition, is a small but dedicated following, emphasis on both small and dedicated. Now, fans and critics may appreciate the dedicated part, but not surprisingly networks are more concerned about the part about the following being small.

From http://en.wikipedia.org/wiki/Cult_following
Again, not music to a network's ears.

Unlike you, Nielsen relies on more accurate and scientific methods of polling. While I doubt every celeb mentions Arrested Development, even if they did, celebrities are an incredibly small portion of the population, and can't be accurately described as representative of the population at large. An overwhelming majority of them are passionately liberal, yet in the last presidential election John Kerry didn't win by a landslide.

We could do this, but it'd take a massive rewriting of our voting laws. Heck, we can't even get rid of the Electoral College.

Besides, I said that polls can be accurate, not certain. Polls can give a very impressive amount of accuracy in their results, but there will always be a certain amount of inaccuracy. I personally wouldn't want a president voted in with a +/- 2% margin of error.

Anyway, it's good to occasionally take a complete survey of everyone (or in the case of voting, everyone elegible that bothered to participate) to make sure the polling is accurate.

The difference is that all those other methods use a finite method of delivery that's COD. They don't have to count how many people read/saw/heard their product, just how many units they've sent out the doors. Even this method has errors in it, since it doesn't take into account sharing between friends or from libraries. However, it doesn't matter to the companies how many people actually read/see/hear the product, just how many they sell.

TV is different in that their delivery system for each show is to the maximum number of viewers. They can't count the number of people they delevered it to as a measure of success, rather they have to count the number of people who actually watch it, something the other methods don't have to do. Since every particular show reaches everyone that can receive that network, the polling number for each of their shows is huge. Potentially the entire nation. They can't poll the entire nation every night for every show. Thus representative samples come into play.

Also, while music, movies, and books can judge their success after they release their products, they do use representative polling beforehand a lot. Or haven't you heard of product testing or test screenings?

While we may never "truly" know, we will have a pretty accurate guess, with only a minimal range of error. If you could find the budget and time to completely monitor everyone in the nation (250 million+) and what they're watching (not just every TV mind you, but every person that's watching that TV) all the time every day, you'll still only end up being about 2% (maybe less) more accurate than a poll of only 5,000 or so people.

Just for some perspective:

Over the years I've loved plenty of shows that have been prematurely cancelled. Some last 2-3 seasons, some lasted less than 1. Everytime I've been upset and dissapointed, but not at Neilson.

The problem isn't the ratings, it's the network's overreliance on them to the exclusion of any context.

When Family Guy was cancelled, it was because of low ratings. When it was released on DVD, the sales went trhough the roof. Does this mean that the ratings were wrong? No, the ratings were probably pretty accurate. What it showed was that the potential audience was there and much bigger than FOX realized, because their constant moving and pre-empting of the show made it difficult to watch. Ironically, the DVD sales are themselves a form of polling, since if only everyone who bought the DVDs watched the show, the ratings would be horrible. Instead, the DVD sales were taken as being representative of the potential TV audience for new episodes. Sounds a bit like a form of polling, doesn't it?

The same has happened to other shows. Firefly was cancelled because of low ratings. Do I think the ratings were wrong? No, I know for a fact that most people I know didn't see the show while it was on. Yet DVD sales have been good and steady for this show, through word of mouth. This isn't an example of flawed ratings, but flawed promotion on the part of FOX. Most fans, when complaining about the show's early demise, don't pin the blame on Neilsen's polling for the show, but on FOX's mishandling of it.

Over-reliance on just the ratings can cause bad decisions by networks. Shows like X-Files and Cheers had low ratings their first (few) seasons. If the network hadn't stood behind these shows, they wouldn't have later gown into the ratings dominators that they did. Again, the ratings weren't wrong, people really weren't watching these shows in their first seasons, it's just that the risk the networks took keeping them on paid off in eventual seasons.

It looks like FOX may be standing behind Arrested Development, since going for a 3rd season of a show that's never had great ratings shows a good amount of commitment. Of course, even if a show is a critic's darling and has a cult following, and even if the network pushes it, it may never become a mainstream success. The mainstream is fickle, and impossible to predict.

But how do you truly know? How do you account for people who tape shows and watch them later? Or people who watch a program in groups instead of on individual sets? To my mind, there are too many variables to account for to comfortably proclaim a survey method as reliable. It's all we have, so that's what the networks use, but I'm sure they'd rather have hard numbers than a statistical approximation of their audience.

Did you bother reading the links I provided?

From http://www.nielsenmedia.com/whatratingsmean/
So they're able to detect when people are watching a tape, and they're also able to tell what's on the tape.

Again, from http://www.nielsenmedia.com/whatratingsmean/
Neilsen also uses other survey methods like diaries and random phone surveys to check their data for irregularities.

Which is why Neilsen uses multiple survery methods and checks them against each other, as well as against Census information. Again, the ways Neilsen cross-checks its samples and survey results is available at http://www.nielsenmedia.com/whatratingsmean/

Of course they'd rather have the "hard numbers," if such numbers were easy to come by. They're not though. It'd be prohibitively expensive and time consuming to try and sample every TV and every person watching TV. The Neilsen ratings, while a statistical approximation, are a fairly accurate statistical approximation, and falls within the realm of actual possibility.

Your points are valid, and things I have considered before. However, what about the statistics that these shows sell so well on DVD? I know Fox thinks that these DVD viewers came out of the wood works and scratches their heads wondering where these viewers were when they were rating the shows, but I am willing to bet most of those people watched that show when it was on TV. Like the first season of Chappelle Show. Chappelle wasn't a ratings winner for Comedy Central, if you recall, which is why they grossly under produced season one on DVD, and which is why most stores were sold out before the end of business day the day season one came out.

I fully understand my cross section I am exposed to isn't all that vast, but I would meet a lot of different customers when I worked at a video store and I would see quite a different cross section into stuff like Family Guy. And who knows? Maybe if Neilsen put more ratings boxes in more homes, they would get similar results, but which homes would they put more boxes into? I knew a family who became a ratings family, and I was also considered by Neilsen myself. The fact of the matter is they often go for the typical family of mom, dad, and 2.4 children and skip over the homes of single folks or people who won't watch hours upon hours of TV a week. Which, I suppose, is understandable, but I just have to wonder in all those homes they are skipping over based on a few demographic technicalities are they skipping over a massive audience that would boost the ratings for some of these shows?

I would like to add that I think TVtome.com (RIP) had a great ratings system, and was actually probably more accurate. I have no scientific data to back it up, but I feel this way because it allowed anybody of all walks of life to post their ratings for the show on the site, not just a certain demographic targeted by Neilsen. Again, Neilsen may have been relevant back 40-30 years ago, but I think it has become a thing where it panders more to the advertisers than it does anybody else. Advertisers would never go for a ratings system like TVtome.com had because they would have no idea which demographic was watching which particular show.

If it's straight to DVD, they how do you call it a TV series?

Wouldn't it just be a 13-hour made-for-DVD movie? There's no "TV" in the equation.

That makes about as much sense as saying a movie is direct to DVD, how is it a movie? The episodic nature of the final product would make it a series by nature. Maybe not a "TV" series in the original sense, but it would be a series nonetheless. As another poster pointed out, in Japan, original TV animated series are created for DVD all the time.

According to the last real numbers we had for Firefly DVD, it had sold over 500,000 copies. Now that's a lot, for DVD sales. For network TV, 500,000 viewers represents a 0.5 rating, or less than 0.5% of viewers. That's a horrible rating. Firefly regularly score ratings more than 5 times that size and still got the axe. All of the people who bought the DVD could've been watching the show, their numbers weren't enough to save the show.

Remember that FOX, the network, doesn't care how well a show sells on DVD. They don't ever see that money. The only way FOX makes money off of shows is through the advertising they sell for it. If a show doesn't get good ratings, they can't sell the advertising for a high enough price, and they lose money. The ratings for Firefly were low enough that not airing all of the episodes, even though they paid for all of them, made more economic sense.


Actually, I do recall, and it's certainly not as you recall.

http://press.comedycentral.com/press...ne_Ratings.xml
What show wasn't a ratings winner?

As noted above, Chappelle's Show episodes averaged 5 million viewers. The DVD sold 2 million, after 8 months on the shelf. So the potential was there for those sales, especially after 8 additional months of reruns and cross-promotion with the 2nd season. The potential for the DVD sales was there, although so far the show's DVD sales have been unusually high.

One thing that Nielsen doesn't capture is the rabitity of a fanbase for a show, or even really the fanbase at all. All it does is capture individual watchers. There's no real way to tell whether or not those viewers are regular watchers or enjoy the show enough to buy it on DVD. And ultimately, those factors don't matter much to Networks, who just want "asses in seats." If not enough asses are in enough seats watching their show, it gets the act, no matter how rabid the tiny fanbase is.

Here's a nice article on why DVD sales and ratings may not match up:
http://www.buffalonews.com/editorial...06/1032029.asp
Of particular note:

The fact of the matter is that your "fact" is plain wrong.

Again, from http://www.nielsenmedia.com/whatratingsmean/
The TVtome poll would get laughed out of the room by any statistician. It is in no way near the standards for an accurate representative sample. For one, the people who took the poll had to A) know it existed, B) be able to access it, and C) be inclined to actually use it. TVtome didn't poll every show a user watched, the user decided which shows they would rate, which already creates a bias. And so far, I think the average demo for regular internet users is still a bit skewed from the norm.

That's what it's always done. In fact, that's what it was designed to do. Networks aren't in the business of selling you shows. They're in the business of selling advertisers viewers. Aside from PBS and commercial free pay channels, all networks use a business model where advertising is King. Nielsen is built aroung this business model, and ranting against it is like blaming seismic meters for only being able to detect possible earthquakes and not hurricanes. It does what it was designed to do, and it does it well.

I'd sure buy a 6th season of Angel or a 4th season of Roswell! Heck, I'd even go for an 8th Season of Buffy!

Chapelle show 1st season has gone past the 3 million DVD's sold mark. Season 2 has sold around 2 million already and hasn't even been out that long.

I am beginning to wonder if you work for them!

And considering how you showed Chappelle did so well in the ratings, it makes the studio and the stores look more idiotic for not having enough DVDs of season one available on the date of release to meet what clearly would be a large demand.

ps...I still think Neilsen sucks. I was just voicing opinion, and even made it clear I wasn't posting what I thought was fact (which you so politely bolded in your quote of me). But way to go on your research. I don't think I've had a more through reply to one of my threads before!

Just a note in advance, this gets a bit long... I just really, really dislike the Nielson rating system. Don’t bother reading if you don’t care about any of that stuff. Another vote for "works at Nielson" though.

I’m going to ignore the numbers issues even though I simply cannot believe a few thousand can truly give an accurate representation of 100,000,000 within a couple percentage points for every single television show over hundreds of networks at a time...

With DVRs and digital receivers, the technology is now in place for huge amounts of people to participate in a truly accurate and representative system. There’s no reason such an archaic system sampling miniscule amounts of people should still be in place. But forget all of that while I point out some unacceptable flaws in the Nielson system:

Does Nielson include Tivo/VCR/HD-OTA viewers? Nope, not the last time I checked. I've read they are "in the process" of doing tests to include such things in their system, but God only knows if/when they will ever get around to monitoring those viewing methods. But they definitely did not sample any viewers using those methods to watch TV when Firefly aired.

Firefly aired on Friday nights... A night a lot of people go out, along with Saturday. Many people record shows that air on these nights on their VCRs and watch them later when they come home. But according to Nielson, unless you watch the television show as it airs, you don’t count. And surprise, these two nights are ratings wastelands, as reported by Nielson, even though a lot of people I know go out and still watch the shows when they come home. But they don’t count.

Let's go more into detail on Firefly specifically though. When Firefly aired, almost anyone worthy of being called a geek owned a DVR at that time. And the geek/tech-enthusiast was the core of Firefly’s fan-base. Don’t take my word for it: http://www.dvdtalk.com/forum/archive.../t-241500.html Did Nielson include any of these people in the ratings? Nope.

Bottom line: Nielson has/had massive holes in its sampling audiences, holes that would alone have a serious outcome on a show’s ratings, but would combine to have a huge effect on a show like Firefly. Then there's the whole issue of Nielson not counting digital HDTV broadcasts, but that wasn’t much of an issue when Firefly aired.

Whether Nielson has truly entered testing devices to correct these massive oversights or not doesn't matter. The bottom line is they do not give accurate representations within a couple percentage points all the time, or even most of the time as far as I can tell.

Heaven only knows what an accurate rating system might have had Firefly getting, or other sci-fi shows or things airing on Friday/Saturdays. Count me as one of the people anxiously hoping for a day an accurate rating system that is flexible to new technology and variables is put in place, one not controlled or created by Nielson Media Research.

And to remain on topic... I can't see it ever happening for a new show, that's never aired a single episode on television. An existing show, yes. It will happen eventually, it's just a matter of time. But only for shows with that "cultish" factor, something that already has a built in fan base of dedicated fans.

It'll take a very unique kind of show, and I have to think it will always be a very rare thing. Obviously a lot of canceled shows are getting unaired episodes out on DVD... But for a show to have episodes produced without an order by a network, strictly for the DVD set and financed strictly by the anticipated sales of the DVDs? It'll be lucky to happen a couple of times over DVD's life-span. Enterprise almost became a first, but Paramount shut any notion of such a thing happening down. The fact that Enterprise got as close as it did means that such an event is probably going to be at most three years down the line.

I don't. I was, originally, one of those who was actually critical of Nielsen and their polling. After all, there seemed to be such a wide fanbase for these cacelled show I liked online, surely we weren't alone? Surely the fanbase was larger than the numbers said? However, I then actually read about Nielsen and its polling procedures, and about statistics in general, and Nieslen polling suddenly seemed a lot more reasonable. It's not surprising that the people who criticize Nielsen are mostly those who are fans of cancelled shows. But that's just shooting the messenger. Nielsen doesn't cancell shows, they just relay the numbers.

The Simpsons has regularly had a much larger viewing audience than Chappelle's Show yet hasn't sold as well on DVD. DVD sales are not directly proportional, or even directly related, to the number of TV viewers. Firefly was recently (still is?) the #2 best-selling DVD at Amazon. The show has been cancelled for 2 years, with no reruns on TV as of yet, so those sales weren't based on TV viewers at all.

That's funny, because I clearly remember you saying this:
I bolded out the relevant parts to show what you claimed as "fact" was wrong.

To paraphrase Harlan Ellison (quoting someone else) "Everyone is not entitled to their own opinion. Everyone is entitled to their own informed opinion." While I never may change your opinion, I can point out all the erroneous "facts" that inform your opinion.

Well, initially I just provided a few links. However, it looks like nobody has bothered to read them.

If you doubt this, than you doubt the entire field of statistics.

DVRs and digital receivers are a premium service and only available to those who can afford them. Their demographic is not "truly" accurate or representative, and any numbers off of them would be akin to "Dewey defeats Truman." They're also owned by less than 100% of the population, so the whole "how can you get accurate results by polling only a section of the population" comes into play (not a criticism I have, but one others on this thread, including yourself, seem to have).

Where did you check? I've already provided a quote that said they monitor VCRs.

That's because, to advertisers, you don't count. Again, TV is not in the business of selling shows. They're in the the business of selling advertising. Time-shifters overwhemlingly skip advertising. Why should the advertisers care how many people watched the show but not their advertising.

Aside from the time-shfiting issues, that article you cite says that Tivo had 500,000 subscribers. That means that if every single one of them had watched Firefly, they would've only accounted for less than 20% of the viewership for that show. Of course, not every Tivo viewer did, and we don't know how many actually did, so those numbers are useless. Regardless, I don't think they could be counted as the "core" of the fan-base.

Where are your statistics to back this up? "As far as I can tell" is a really poor basis for comparison, especially when compared to a long-established, rigorously stuctured and applied method of polling.

What's your beef against Nielsen? Even if they created a more accurate polling system, you sound like you would doubt it simply because it's from them. This sounds more like prejudice than any real criticism of its methods.