What's the highest possible rating?

[Mister Oragahn]

What's the highest possible rating?

When God himselfs says he's a fan of Stargate?

[Morrolan]

It's technically possible, but highly improbable because Neilsen will not want to include that inaccuracy in to the sample.

The question was what is the highest possible rating, not the likelihood of it. The highest rating possible is 100.
I expect anything over that would result a fair bit of crosschecking and statistical deletion of the offending household. Not that it would matter, since it would still mean 100% of the audience watched the episode.

But the highest possible rating is not 100, it can be higher. Back when the ratings were created, you could only have 100% viewership. With multiple timezone feeds today, due to cable and satellite, it is possible to have greater than 100% viewership as far as ratings go.

You're wrong about statistical accuracy. Neilsen uses raw data and it is interpreted in many different ways. The numbers are never real. Ratings are not scientific in any manner.

[smurf]

But the highest possible rating is not 100, it can be higher. Back when the ratings were created, you could only have 100% viewership. With multiple timezone feeds today, due to cable and satellite, it is possible to have greater than 100% viewership as far as ratings go.

You're wrong about statistical accuracy. Neilsen uses raw data and it is interpreted in many different ways. The numbers are never real. Ratings are not scientific in any manner.

You are going to have to back up where you are getting this information from because I'm getting lost as to where Neilsen (or its paymasters) are ever likely to accept a statistic which gives a higher than the total possible audience.

In fact I'm being daft, it is not possible to get a higher than 100% viewership. If Channel X has 50 households in the east, and 50 households in the west, it needs all 100 households to watch to get 100% viewership.
If one of those households straddles the east/west line then to get 100% viewership it needs 49 households, and the messed up one to watch it in the east, and then 50 households, and the messed up one to watch it in the west in order to gain 100%. Why? because even though there are only 100 households it would essentially mean that Channel X has 50 households in the east and 51 in the west (or vice versa, since one falls in both areas), and so now needs 101 households to gain 100%.
And seeing as percentage equals points, you can't have over 100 points.

I'm sorry, but what the heck is the raw data if it isn't the numbers collected through statistical sampling?
I'm not arguing Neilsen's absolute accuracy, but they are hardly likely to deliberately add inaccuracy to the numbers are they?

[Morrolan]

You are going to have to back up where you are getting this information from because I'm getting lost as to where Neilsen (or its paymasters) are ever likely to accept a statistic which gives a higher than the total possible audience.

In fact I'm being daft, it is not possible to get a higher than 100% viewership. If Channel X has 50 households in the east, and 50 households in the west, it needs all 100 households to watch to get 100% viewership.
If one of those households straddles the east/west line then to get 100% viewership it needs 49 households, and the messed up one to watch it in the east, and then 50 households, and the messed up one to watch it in the west in order to gain 100%. Why? because even though there are only 100 households it would essentially mean that Channel X has 50 households in the east and 51 in the west (or vice versa, since one falls in both areas), and so now needs 101 households to gain 100%.
And seeing as percentage equals points, you can't have over 100 points.

I'm sorry, but what the heck is the raw data if it isn't the numbers collected through statistical sampling?
I'm not arguing Neilsen's absolute accuracy, but they are hardly likely to deliberately add inaccuracy to the numbers are they?

If there are 50 household in the east and 50 households in the west, and 10 of those households receive both east coast and west coast feeds, you have 110 households watching the program. The data is desceminated that way. Neilsen doesn't split regions equally.

The Neilsen numbers are constantly manipulated to reflect higher ratings for a show a network wants to push, and lower for a show that a network wants to kill. Neilsen collects the raw data and run it through a changing formula to compute the ratings. Neilsen isn't interested in accuracy, it's about the dollars. The same goes for movie grosses.

[smurf]

If there are 50 household in the east and 50 households in the west, and 10 of those households receive both east coast and west coast feeds, you have 110 households watching the program. The data is desceminated that way. Neilsen doesn't split regions equally.

The Neilsen numbers are constantly manipulated to reflect higher ratings for a show a network wants to push, and lower for a show that a network wants to kill. Neilsen collects the raw data and run it through a changing formula to compute the ratings. Neilsen isn't interested in accuracy, it's about the dollars. The same goes for movie grosses.

But that's what I said.
Although there are actually only 100 households, they would have to double count those 10 households to get their 100%. Thus instead of 100 households equalling 100%, 110 households equals 100%. It doesn't equal 110%. This isn't ratings, this is mathematics.

I'll let someone else give an in depth reply to the Neilsen conspirancies.
Surfice to say that you need to know that the main people who pay for the ratings information are the advertising companies and not the networks. So it is not in their best interest to massage the ratings and have to pay over the odds for a low rating show.

[MediaSavant]

Having worked in statistics, collecting statistical data, I have yet to believe that you can actually get accurate numbers of samplings of this size. And no one has been able to convince me how.

They don't have to convince you. You don't work at Nielsen or any of the networks or advertising agencies that have higher access to the data than you do.

All that matters is that they are accepting of the statistics. After all, to a network, the numbers reflect income. To the ad agencies, it reflects what their clients have to pay. Rest assured that there is oversight coming from both sides.

[Morrolan]

But that's what I said.
Although there are actually only 100 households, they would have to double count those 10 households to get their 100%. Thus instead of 100 households equalling 100%, 110 households equals 100%. It doesn't equal 110%. This isn't ratings, this is mathematics.

I'll let someone else give an in depth reply to the Neilsen conspirancies.
Surfice to say that you need to know that the main people who pay for the ratings information are the advertising companies and not the networks. So it is not in their best interest to massage the ratings and have to pay over the odds for a low rating show.

There are 100 households! 90 of them watch the show at their given timezone time. That's 90% of the viewing audience. The additional 10 are counted twice, because they watch both feeds. 100 households, 110 households watching the show. Neilsen, to my knowledge, still has not conpensated for multiple feed cable and satellite feeds. I do know that as of a few years ago, the Neilsen data was extremely raw.

The whole ratings business is a sham. Yes, I know that advertisers contribute a large sum to Neilsen, but networks have heavy influence. Advertising rates are not solid for each advertiser, as backroom deals are made all the time. I will say that I have seen more manipulation when it comes to movies than television, but the numbers are not accurate at all, nor are they intended to be.

[smurf]

There are 100 households! 90 of them watch the show at their given timezone time. That's 90% of the viewing audience. The additional 10 are counted twice, because they watch both feeds. 100 households, 110 households watching the show. Neilsen, to my knowledge, still has not conpensated for multiple feed cable and satellite feeds. I do know that as of a few years ago, the Neilsen data was extremely raw.

The whole ratings business is a sham. Yes, I know that advertisers contribute a large sum to Neilsen, but networks have heavy influence. Advertising rates are not solid for each advertiser, as backroom deals are made all the time. I will say that I have seen more manipulation when it comes to movies than television, but the numbers are not accurate at all, nor are they intended to be.

In this scenario the percentage is not a representation of growth (difference from a given base number) it is a representation of the whole (difference from a to be determined maximum). The whole cannot be greater than 100%, because 100% is the maximum possible - ie. everything.

If you went out and did a yes/no survey of 400 people, and 400 people said yes then you would have the maximum result out of the determined maximum - 400 out of 400, or 100%.
If you get back to your desk and find you somehow have 500 sheets of paper with 500 yes votes, you do not have result of 125% you instead have a new maximum, 500. 500 out of 500 is 100%.

The 100% point is determined once the data is collected. Otherwise if I went out and did a survey of 800 agreeable people to your 400, I'd get a result of 200%. Which is daft.

Of course if you get 500 results when you expect 400 then you should review it very very carefully, or dump the entire survey and do it again.


ETA: If Neilsen can tell what you are watching minute to minute, then they can tell that you are watching the same thing twice even if they are on different feeds. To the best of my knowledge they put a box on everything which can receive a signal.