Rohit Nalluri

For as long as I remember, I have been in love with Excel. I like to find patterns in the data, and I’ve often thought of my brain as a massive data processing engine. If you leave me alone with a bunch of excel reports to create, I’m the happiest and I often find myself in a state of flow.

Only lately did I realize that it was not Excel I loved, but data, and working with data. Once I understood this, I became curious about other data analysis techniques, and began learning R, a programming language with strong emphasis on statistics. R is well known as the language made by statisticians for statisticians.

I’m no statistician, but over the months I’ve become somewhat proficient at using R and its various packages, such as R Shiny, a lovely package that allows you to quickly design and build online dashboards to display your data. I am working on a personal Shiny dashboard using data about myself that I’ve collected since 2010 [See? I love data. Also, I’m somewhat narcissistic], and that will be part of the next post. But in the meantime, I wanted to flex these new-born data analysis muscles.

I was shopping around for a simple dataset to play with, when I stumbled upon the wikipedia page for one of my favorite shows right now – Impractical Jokers. If you love to laugh and enjoy improv comedy, this show is for you. Go watch it, if you haven’t done so already.

If you know the show, then you’ll enjoy the infographic I’ve made using data skimmed from wikipedia. The link to the interactive infographic is here; feel free to jump to it while I ramble a little about the data sourcing, cleansing and analysis steps I took to complete the infographic.

I used the rvest package to pull the wiki page and extract the tables. I then used gsub, lubridate and other data cleansing/manipulation functions to cleanup the data. There were a lot of square brackets [wikipedia uses square brackets to hyperlink their sources] that had to be removed, seasonal and episodic data had to be streamlined and ‘factored’, and episode ‘aired’ date had to be converted into the proper data format [lubridate is your friend].

After prodding the data into shape, it was time to visualize the data. I used summary functions to quickly check the stats [where I found Sal was the biggest loser], and used the ggplot2 package to create a line chart to find that the popularity of Impractical Jokers has been trending down, from a peak of 2.08 million viewers in Season 2 to a low of .58 million viewers in Season 4. In Season 5, they have experimented with new pranks and upgraded their bag of tricks, and this has evidently helped their viewership, which slid back up to 1.23 million for the Season 5 finale. This maybe an outlier, as the average viewership for Season 5 is 0.83 million, which is lesser than the show’s overall average viewership of 1.18 million viewers.

As I observe in the infographic, the Jokers need to explore new tricks/pranks and new locations [their Miami and London episodes spiked the viewership]. Intuitively [and psychologically], it seems to me that new locations must have a positive impact on the Jokers’ special kind of improvisational comedy, but I don’t have the data to back it up. 🙂

Keep laughing.

*

Image of the Impractical Jokers Infograph. The Link is more interactive: