As Big Data continues to make its push to control everything about our existence, we finally have a site aggregating what people think about individual TV episodes. Enter Graph TV.
According to his Twitter bio, Kevin Wu enjoys biology and writing code. What he hasn’t mentioned is that he has created the incredibly useful website Graph TV. After entering the name of a television series, the site creates a scatter plot of the average rating each individual episode has received on IMDB. The plot includes color-coding and a trend line for each season. This allows the user to suss out which episodes are the most and least popular and how show quality has changed season-to-season.
I spent at least an hour playing with the site and found some interesting tidbits about the least popular episodes of shows:
- Backdoor pilots do not do well. The least popular episode of The Golden Girls was the season two finale “Empty Nests,” which was intended to be a spin-off starring Rita Moreno. This was not the origin episode of the actual spin-off Empty Nest, which didn’t come along until a couple years later.
- Dream sequences do not do well either. One of the worst episodes of Roseanne according to Graph TV was the season two episode titled “Sweet Dreams.” Roseanne has an elaborate fantasy about trying to take a bath, but involves killing off her kids and an elaborate musical trial. This might have been the origin for Dancer in the Dark, but is quite dreadful as a comedy.
- Two clunkers for Little House on the Prairie include “The Godsister” (the only episode where Carrie is the main character) and “The Halloween Dream”—where Albert dreams he and Laura are pawns in a war between tribes.1
- Clip shows are never as good as the original. There are only three negative outliers in the first 13 seasons of The Simpsons according to the graph: 6-3 (“Another Simpsons Clip Show”), 9-11 (“All Singing, All Dancing”—also a clip show), and 13-17 (“Gump Roast”—yet another clip show). the least popular episode of season 4 was also a clip show, though it was not as negatively received as the aggregate.
There are some flaws with the methodology. Not all series are fully represented as the fanbases for Punky Brewster and Living Single have not evaluated each and every episode. Reality shows also do not measure quality, as some of the ratings may be based on outcomes rather than content. In other words, a popular contestant’s elimination may skew the results downward.
I find this kind of aggregation fascinating. Also, as we continue to cover Unfortunate Episodes, we now have quantifiable proof of which ones were bad ideas. Thank you Kevin Wu. Graph TV is doing the Lord’s work.
- Oh, and they are in Native-American-face. ↵