Wednesday, July 1, 2015

Impractical Jokers and Shared Punishments

There are data everywhere.

I'm constantly looking for it, and I'm constantly compiling it. A lot of this compilation doesn't really go anywhere, other than having some data, and maybe having a slightly different way to think about something. 

Today, let's talk about one of these random data compilations (not a compilation of random data, though), because why not. 

I watch a lot of weird stuff on TV. A lot that is weirder than today's focus. Today's focus is on a show called Impractical Jokers:



If you haven't seen this show, check it out. I've never been a fan of shows that just capitalize on making a joke out of random people, let's make that clear. It's too easy to just feel for the person being made fun of.

Instead of making a joke out of random people, the four guys on this show make a joke out of each other. It's the difference between laughing at random unsuspecting people and laughing at people who have placed themselves in that situation. It's still awkward, and still weird, but usually pretty funny. It is a subtle difference.

The format of each show is a series of small competitions, with a clear loser at the end of each show that has to then handle a punishment put together by the other three guys. 

All it takes is the word competition for me to start to thinking about how to capture and track these results over time. I decided to give the first season a go, to see how the numbers play out.  

In terms of the first season, here's how the end-of-show losses broke down:


You can see that Q did the best during the first season, though not by much. Over the course of 15 episodes, losses were really pretty evenly distributed. Take note that I'm not including the final episode of the season, where there was a 4-way tie. If you wanted to include it you could just add 1 to each bar.

Frankly, this graph is almost suspiciously even, raising the question of how orchestrated (easily done through editing) this first season was, say, by producers. 

There hasn't only been one season of the show, though, and we can look at the second and third seasons:



A 4th season is currently ongoing, but we can look at what's happened so far:


These later seasons do seem to be a little less balanced than that first one, which also becomes more obvious if we just look at the trend overall:


It's fairly close, within about a half a dozen or so between the higest (Sal) and lowest (Joe), but there's also some clear separation going on. Murr and Sal do seem to lose more frequently than Q and Joe. 

We can also take a look at how this has played out over time:


Again, there's not a huge effect here, but you can see that Sal and Murr are both above what would be expected evenly (25%), and Q and Joe are both below. A quick chi-square goodness of fit for equal proportions fails to show that these differences in proportion are significant (3, N = 89) = 1.64, p = .6494.

That said, I would imagine that six more (or fewer) punishments at the end of these episodes are large enough effects to those receiving (or giving) them. Good job, Joe and Q!