The Nielson Ratings system is the most flawed form of influential statistics there is. Ok, maybe not, but it is up there. I attribute the sheer dumbassness of it to the cancellation of some of the best TV shows that have ever been on the air. Here are some of the shows that have been canceled due to low Nielson ratings: Firefly, Arrested Development, Freaks & Geeks, Dark Angel, and my favorite...Futurama.
If you're not familiar with how the Nielson Ratings System works, I'll clue you in; if you are, I guess you can skip this paragraph, but you should read it anyway. The way it works is that Nielson picks families and monitors what TV shows they watch. The key word in that last sentence is families, I'll touch more on that later. They monitor what families watch two different ways; one is by TV journal, and the other is by a Set Meter. The TV journal is simply a journal where the family writes down everything they watch and they send it in at the end of the week. The Set Meter is a box hooked up to your TV and the phone line, that lets Nielson directly know what you're watching at all times.
My gripe is that they have the same flaw as all other statistical samples, it's only a sample. Nielson doesn't disclose their sample size, which means it could be 1000, or 100,000 of the 114.5 million television households in the United States. Apparently the "public" doesn't need to know their sample size because it's not important to us. Let's say that their sample size is 10% of the total population of TV owners. If 3% of their viewers watch Lost on Thursday, then they assume that 30% of TV owners watched Lost (3%/10% = 30%). Does this seem wrong to anyone? Each TV network has a line. If a current show's ratings are above that line, they renew the series, if they're below, they cancel the series. There are rare cases where is line rule doesn't apply...rare. What if a show has a popularity with families that aren't a Nielson family? If Nielson reports that Terminator: Sarah Connor Chronicles has a .05% share of their 10% sample, this means according to Nielson, 5% of TV owners watched the show. But what if, in actuality, 20% of TV owners watched the show, and it gets canceled (let's hope not), and leaves a lot of fans really pissed off (which I will be if they do). I hope you see the flaw now.
But guess what? That's just the half of it! Earlier I mentioned how families was the key word. Well Nielson only samples families; households that have Dad, Mom, Billy, and Cindy. The Nielson Ratings System doesn't account for TVs outside the "conventional" home, such as colleges, bars, and other public places where you can watch TV. I am a college student and there are shows that college students love that many adults/families wouldn't, and in turn the show gets canceled. If they offered a way for students, or other "non-families," to record what they watched, maybe online, I would so be game to sign up for that program. Would it really damper their results? No.
They also don't account for the viewing of TV shows online. How many people who miss their favorite show, and don't own DVRs (TiVos), go online the next day to watch their show? I bet a lot. Networks are trying to factor in which shows are recorded by a DVR, but this is being done on the network end, meaning Nielson has nothing to do with it, which sadly means it won't matter. The fact that they haven't changed their 1950s way of measuring ratings is a joke. If they don't change their ratings system to add more to their sample size, or measure into other forms of media, then all of our favorite shows could be canceled at will.
Additional Links:
Save Sara Connor Chronicles!
Save Dollhouse!
List of top-rated shows since 1951.