I thought it had to do with Nielson ratings? It's like a poll. 100 people are polled "do you like chicken", 10 yes, 90 no. For some reason people seem to think this holds true if you expand it across the population, so 10 becomes 10 million, 90 becomes 90 million etc. In reality, you could have simply just polled the only 10 people in Britain that like chicken, and actually 100 million minus just a measly 10 dislike chicken. And that's how Nielson works, with what you watch being your answer in your poll. Of course, not everyone watching TV has a Nielson box, so you can imagine how wildly innacurate it is, despite its system being treated as if it's infallible and 100% accurate. It's basically useless. 90 million people could watch Glee, but if no one with a Nielson box did...guess what? Just think how many great shows got axed due to this inaccuracy.
^Of course I doubt they still use this method. Since theres many more methods way more accurate.
Comments
Since 2001 only sporting events been past 20m viewers. 2013 was particular bad with NYE fireworks being most watched program with just over 13.5m.
http://en.wikipedia.org/wiki/List_of_most_watched_television_broadcasts#Most_watched_programmes_per_year_2
Almost everyone here does as well, the difference being the U.S. have had it much longer.
^Of course I doubt they still use this method. Since theres many more methods way more accurate.