Breaking Bad is one of the most critically acclaimed series of all time. Don’t take my word for it! According to IMDb, everyone agrees! I recently came across the website http://graphtv.kevinformatics.com/, which allows you to type in the name of a show and plot the average IMDb rating for each episode, along with a regression line per season that shows how the season progressed. I typed in “Breaking Bad” and the resulting graph showed a series that started good and got better. The site is very impressive, but being a Tableau junkie made me want to slice and dice the data to see what else I could come up with.
First, I wanted to play around with the layout and add some cool images into my visualization. Adding an image to a dashboard is pretty easy, but it gets trickier when you want to do so on a worksheet. If you’re getting frustrated, remember that within the Background Image Dialog Box the measure on your columns shelf corresponds to the X axis, and rows measure is the Y axis. Both need to be on the worksheet before the background image will appear.
I decided that for this view I like the general idea of the kevinformatics.com view, so I put Episode Number on the columns shelf and the IMDb Rating as the rows. However, I wanted to see an overall trend line of the series, which I plotted in yellow. Yep, the series got better as it went on!
Next I added a reference line of the average episode rating so I could quickly see which episodes were above or below average. Note that I computed the average rating as 8.93, while kevinformatics.com came up with 9.5 out of 10. What gives? My average is an arithmetic mean of the rating of each show, meaning the rating of each episode has equal weight. The kevinformatics.com average is weighted by the total number of reviews. What this means is that the series finale, which had over 26,865 reviews (10 million viewers) weighs more heavily on the average than the pilot, which had 5,207 reviews (and a bit less than 2 million viewers). Which average is better? As with all good answers, it depends. For the sake of my viz the simple average is fine, so we’ll go with that.
On the next worksheet I wanted to try a different way of comparing seasons. By making each column a season plotting each episode I can easily see the distribution of ratings within each season. Then I plotted the average, which allows me to see how each season stacks up.
Finally, I added a worksheet that displays the synopsis of whichever episode beaker is selected. It’s great to see the episode number and title elsewhere, but many times I need something to jog my memory about what was actually in the episode.