So one of my favorite shows, Community, just got pulled off the mid-season schedule on NBC. That means that after the winter break for shows on NBC, it’ll stay off the air for most of winter and come back around march or so. Naturally, when I first heard the news, I was under crisis alert. I had the Arrested development flashback moments and the firefly jitters (both are excellent shows, you should watch them). Community is kind of a cult show. It generally does not get good Nielsen ratings but, it has a small and devout group of fans. In addition, it’s on at the same time as The Big Bang Theory and whatever singing show is on fox at 8 on Thursdays. This essentially guarantees that most Nielsen box people will not watch the show, which is incredibly dissapointing.

But, how exactly do all of these executive producers and line-up managers decide what show goes on and when does it get cancelled? They use an outdated source of information called Nielsen ratings. Essentially, 25,000 Americans get set-top nielsen boxes that record information. They send back information saying that these people watched this show at this time, and these people watched an entire season and so on. After taking my AP stat class, it’s apparent that this sampling size is big enough to give producers the data they need, but, is it? Can a diverse “melting-pot” be accurately measured with these approaches? In addition, these boxes just give out pure, cold, hard statistics. Is that just enough to factor the viewings and popularity of a show?  For example, take the “wildly popular” new show on NBC, Whitney. This show has gotten fantastic ratings due to the fact that it leads in after the office. If people with the Nielsen boxes leave their tv on after the office, Whitney starts getting “views” and it’s rating starts going up. Isn’t that just fantastic? In comparison, Outsourced, another show that had a lead-in after The Office had fantastic ratings, even better than those of Whitney, but once the show’s timings were changed, it’s ratings dropped to such a low level that it was cancelled in a heart beat. In essence, these ratings are skewed and are killing shows that might actually be doing well. While 25,000 people is a great sampling size, there has to be a better way to control results to prevent bias and false results.