That’s less than a percent of a percent of the American over-18 population. And yet people think that this is representative? You can’t compare it to elections. Election polls work because a majority vote determines a winner. A majority can exist even among 10 people. With Nielsen, however, 20,000 people is not a majority. Also, Nielsen is not a one-click or one-view system.

My idea would involve limiting results to one click=one view for each household. At most, 20,000 households equals ~80,000 viewers based upon the standard two-parent, two-child model. Expanding allows for way more than ~80,000 viewers to determine shows’ fates. If the nation can’t participate, then obviously something somewhere does not work as intended. Nielsen touts itself as national? Well, its time for Nielsen to either adapt or perish. If good shows aren’t given the time of day, Nielsen shouldn’t have a backdoor method for defending its flawed system.

I’ll sum it up at this: Why defend a system that doesn’t count you when determining what airs? Your views only fail to matter if you resign yourself to defending a half-baked idea.

The info would still help networks/ advertisers have a broader picture of viewing by household make up and we all know that it doesn’t matter that Mommy and Daddy don’t watch something as long as the kids are watching, because children often dictate purchasing decisions. This is especially true when considering media industry.

As far as analyzing the data and processing into relevant info for networks and advertisers, it would take an investment, but they could establish a small department within the company to do just that. In addition, the networks could do more with the somewhat raw data provided. They are the ones who have to sell advertising spots. They could analyze it further on their own dime.

My guess is that Nielsen probably owns some sort of patent on the business model. As far as the data goes, it isn’t just about gathering the data (that’s the “easy part”). The real trick is processing that data into information that subscribers can readily use to make business decisions about whether or not to cancel a show, or to but advertisment time.

And everyone always forgets that if every single TV in the country were monitored, you could just as easily see your favorite shows go DOWN in the ratings as up.

Of course you would still have to keep power to the systems that keep track of all of this

20,000 is a statistically significant sample size so I can see the current arguments for validity.

Now, 20,000,000 is about 16% of the adult population (~150,000,000). Let’s say that for a given hour, the ratings are as follows:

1 5 mil (CBS)
2 5 mil (FOX)
3 3 mil (NBC)
4 3 mil (ABC)
5 2 mil (CW)

with demos of 1.2/1.2/1/1/.8 (5)

How does this, especially in a technologically advanced age, represent anything? Why not have a transmitter in each TV that can record data when someone pushes the accompanying button on their remote? I’d guess (educated guess) that audiences and demos could jump anywhere from 3-5X. This would bring numbers more in line with the actual American population. Also, the jump in demos from this would generate ad revenue for the networks that could justify keeping certain shows that seem in constant danger of cancellation. [B]This might also encourage show-runners, as a larger potential audience might keep more pilot shows on air for at least a second season as other shows drop in ratings demos and viewers over the years[/B].

Nielsen and Arbitron tried an expanded ratings system, project Apollo by combining traditional, internet and alternative collection methods – the problem? COST. Advertisers wouldnt foot the bill. Google “apollo nielsen arbitron terminated” and see for yourself.

Again, its COST, COST, COST. Its why no successful attempt to produce a better ratings system has succeeded. I guess some people will never understand that it costs money to collect data.

Sure, when elections are close there are suprises but they are few and far between, especially when they are within the margin of error – usually +/- 3%. So if a candidate is down 6% the day before, the could still win in a squeaker and the poll would be accurate (the leader loses 3% and the trailing candidate gains 3%)

If some day someone figures out a way to give advertisers a superior ratings system at equal or lower cost that Nielsen, those advertisers will beat a path to their door and dump Nielsen. Problem is advertisers understand the methods Nielsen uses and know they are reliable.

I love the endless moaning about Nielsens sampling methods. If advertisers didnt believe the numbers were worth anything they wouldnt be paying Nielsen for them. Many have tried to and failed to produce numbers with larger sample sizes – problem is the methods are not cost effective. Once you get to a certain sample size, adding more people to the sample doesnt change the results. Then again, schools really dont teach much math these days, let alone statistical analysis and sampling, so its not suprising people dont understand how sampling could possibly work. Its the same reason people cant figure out why political polls with sample sizes around 1,000 can accurately predict election results the day before the election when millions of voters are involved.
———————————————————————

You say that and yet there have been more than one election that the networks have missed calls on certain states or elections based on small polling samples. It has happened. A bigger sample will and can bring about even better acurate results. Do I think that the tv nubmers are way off no however I am positive there are some that are off a significant amount.

Does neilsen account for events like sports where people gather together. I insist that as good as many sports ratings are good they would be even better if that were factored in.

I won’t post the link without an okie dokie from Rob or Bill.

I love the endless moaning about Nielsens sampling methods. If advertisers didnt believe the numbers were worth anything they wouldnt be paying Nielsen for them. Many have tried to and failed to produce numbers with larger sample sizes – problem is the methods are not cost effective. Once you get to a certain sample size, adding more people to the sample doesnt change the results. Then again, schools really dont teach much math these days, let alone statistical analysis and sampling, so its not suprising people dont understand how sampling could possibly work. Its the same reason people cant figure out why political polls with sample sizes around 1,000 can accurately predict election results the day before the election when millions of voters are involved.

who here has drained all of the power to make Nielsen ratings go?

The whedon fans?

The visitors impersonating some of the regular posters?

Automan? (obscure TV reference, I know.)

…so, if a man thinks that FOX News discusses how to shoot a fox or deer with an AK-47, he just might be a redneck :p

Evil – Ask the Brain – my guess is it helps him be better able to take over the world (nark!!)

If we don’t get them
We’ll have a cow!

We’re gonna shut you down
We’re gonna cut off your power
You’re going down the drain, losing half your brain,
through the courtesy of The Dollhouse Company!
And The Castle Company!
And Three Rivers Company!

ABC:
DWTS – 3.7
Castle – 2.5

CBS:
HIMYM – 3.6
AOP – 3.0
2.5M – 4.6
TBBT – 4.8
CSI: Miami – 3.9

NBC:
Heroes – 2.3
Trauma – 1.8
Leno – 1.3

Fox:
House – 5.2
Lie to Me – 2.8

CW:
OTH – 1.2
GG – 1.3

Tuesdays (Wednesdays in my country) is always the busiest day here at TVBTN so it feels weird. I feel empty LOL

DWTS GO UP!

http://popwatch.ew.com/2009/11/17/abcs-find-my-family-promos-are-so-eco-friendly/