Ok then, every TV needs 2 cameras

Folks, this would make an interesting doctoral dissertation in statistics.
Is anyone interested? If not, until you have the knowledge to refute current science, anecdotes won't do.
Remember, most people on line are not here, thus those here are very different than anyone else.

Mike from Morgantown

Confidence intervals!

The equation is always:

INT= Sample(x or p, in this case p) ± Z*(how confident you want to be, in a Z-score) x St. dev (in this case it'd be the square root of (p(1-p)n)

Bit of a statisics nerd, its the most valuable school of mathematics IMO

They should take into account DVR recordings for people who watch the show days after it's aired.

The advertisements can be skipped that way. There is no reason to assume anyone would see recorded adds. Those DVR recordings are pointless to Nielsen's bottom line.

It's all about the ads and most people skip the ads of shows they dvr.

They do. Up to seven days after the show has aired.

Those 5000 people obviously aren't chosen randomly, though. They specifically choose people who say they watch broadcast TV. So it's only accurate among Americans who watch broadcast TV. And since they don't know the percentage of Americans who actually watch broadcast TV, they can't accurately gauge anything off of that.

The second type of market:
Metered Market: These markets use a set top box to track what people are viewing. When you turn on the tv you enter your code so the machine know which person in the house is viewing the shows. Much better than a Diary market, but does not easily keep track of how many people in the house are viewing at once or if someone leaves the TV on and leaves the room.

The last type of Nielsen market:
Local People Metered Market: These markets use Nielsens newest system. The box in the house can keep track of who is watching the show, and if the person is in the room. As people walk in and out of the room the box tracks who is watching and which demographic they fit into without user input. This system was going to be spread across the who country, but since it was very expensive to implement, Nielsen pushed the cost onto the local TV stations in a market and the Companies that subscribe to Nielsen. With the huge increase in prices many stoped subscribing to Nielsen and the roll-out slowed to a halt.

Rentrek is the new tracking company that is slowly building up steam. They are much cheaper to subscribe to and have a larger sample base, because they track every cable box in a market. They do not have to pay their sample base either, because they do not put special equipment into the house. Though many media companies are afraid to switch over, because Rentrek only looks at houses that subscribe to cable but not houses that only receive over the air TV.

I think eventually live broadcasts will cease to exist for reasons like this and others. It should pave the way for more investment in TV that people are interested in (via some kind of per-demand system similar to Netflix), and less towards crappy filler shows (personally, I think the reality show craze needs to die a horrible death).

At least this would give a more accurate display of who's interested in a series, and whether it should progress.