How are TV Ratings calculated?

[Camp Anawanna]

This has bothered me for ages, and may have been addressed here in the past at some point but I don't think in a direct thread.

TV people...Fishman...HOW ARE TELEVISON RATINGS GATHERED?

Do they actually know that I'm watching the channel...that 100,000's of people are watching it and they can see everyone?

I've also heard that there are certain 'elite' people who have a special box hooked up to their televisions and they use that data to generalize the ratings. Who are these elite people? Are there enough of them to actually make it representative?

The whole TV ratings thing is a mystery to me.

[Oodus]

TV people...Fishman...HOW ARE TELEVISON RATINGS GATHERED?

In a very primitive and outdated way that is not accurate.

Basically, random people get a letter that asks them if they want to be a "Nielsen family," which means they will get either A) a box put on their TV that calculates what they watch or some shit (they have to input info themselves, like how old they are and other demographic crap. What fun) or B) they fill out a booklet. I believe the booklet has been fazed out. Nielsen then calculates the info using a stupid math formula, which determines that those few families represent the entire market population.

You get paid for the trouble, forget how much. Not a lot.

It is not accurate (a gripe about Nielsen since the beginning), but it's probably the closest we will ever get. Of course, it doesn't take into consideration many people watching one TV (say, a Lost party) or bars.

[Fresh428]

i was asked to do it while in college, but it was like 3 months before the end of the semester and it was going to be a pain to have them install it and such.

its not accurate at all like fish said. i think they compare it to census information for the area that you live in and just assume that like-households are also watching it.

you would think by now that they would get the cable/satellite companies involved and install something into the boxes that only transmits programming information along with the demographic information anonymously.

[Papa Shango]

ratings are over- RATEd. get it?

no but seriously...although many cases a selection can represent the general population, but with this, there are far too many variables involved.

i feel like cable boxes have HAD to reach the point where they know what we are watching. I mean, come one, they know when we order ppv, why cant they tell what we are watching on cable? however, the stat of how any people are watching in one place is something hat will never be reliable.

[NoVA]

I can never quite wrap my brain around ratings points and shares.

Help?

[Oodus]

Rating = The percent of households who have a television in the house that were watching a show. Since by now virtually every household in America has at least one TV, ratings are considered the percent of households in the market, period.

Share = The percent of households who actually had their television on.

So let's say Lost earned a 10.4/15, that means 10.4 rating and 15 share. 10.4% of households in the market were watching Lost, and 15% percent of households that actually had their televisions on were watching Lost. That's why the share is usually higher. During the weekday, for a show like Maury Povich, ratings could be 1.9/14. It has a low rating because fewer people are at home during the day, but it has a high share because it's going by the people who were actually watching TV at that time.

It sounds confusing, I know, but it isn't.

[Jimbuktu]

i feel like cable boxes have HAD to reach the point where they know what we are watching. I mean, come one, they know when we order ppv, why cant they tell what we are watching on cable? however, the stat of how any people are watching in one place is something hat will never be reliable.

seriously it's fucking BULLSHIT.

can't believe the nielsen system still exists. is this fucking 1965 or something? get fucking real.

[gAnonsStower]

oh, they definitely KNOW what we are watching. They just dont want to tell us

[NoVA]

Rating = The percent of households who have a television in the house that were watching a show. Since by now virtually every household in America has at least one TV, ratings are considered the percent of households in the market, period.

Share = The percent of households who actually had their television on.

So let's say Lost earned a 10.4/15, that means 10.4 rating and 15 share. 10.4% of households in the market were watching Lost, and 15% percent of households that actually had their televisions on were watching Lost. That's why the share is usually higher. During the weekday, for a show like Maury Povich, ratings could be 1.9/14. It has a low rating because fewer people are at home during the day, but it has a high share because it's going by the people who were actually watching TV at that time.

It sounds confusing, I know, but it isn't.

So, if "Lost" gets a 10.4/15 in Baltimore, that means 10.4% of all houses in Baltimore with a TV are watching "Lost" (whether that television is turned on or not).

But of all houses in Baltimore watching TV at the moment, 15% are watching Lost.

So share is the better number, right? Because you can't control folks who aren't home/not watching TV. If they are home, you can control what they watch by making content that draws eyeballs.

[Oodus]

Exactly.

And yeah, many people prefer to see the share, especially now that more markets are going LPM (meaning demos). So CW, which is weak as fuck, now looks good with young women. Crap shows like Melrose Place get a high share with W18-34.