Episodes Transcripts

In this section, we will tackle everything related with Natural Language Processing. We will talk about how we analyze song lyrics and episodes transcripts and then discuss the results!

Sentiment Analysis

The first type of analysis we ran through the episode transcripts, was a Sentiment Analysis. What is sentiment you may ask? Well when it comes to data science, the sentiment is a measure of how the word is perceived by people. Depending on the perception (positive or negative) each word is given a grade from 0 (very bad) to 9 (very good) a word is. For example "love" has a very high sentiment score, whereas "terror" has a very bad one.

So we used a labMT word data-base and their sentiment score to calculate the average sentiment score of each episode. This way we could have an idea of how Rick & Morty language would be perceived. We got the following results :

So it appears that overall the sentiment score of each episode and season is just around the average. But the general reviews of Rick & Morty often shows that it's a pretty fun and happy show, so how could that be explained ?

There is a few things to consider, the first one being swearing, there is a lot of swearing in the show, and most of the time the sentiment score of a swear word isn't that good. Whereas the R&M audience might find this overuse of swearing pretty funny. So this is something that might have decreased the overall scores. But we will discuss swearing a lot more later on !
The second element to consider is also the word creativity is Rick & Morty, new words appear every episode, new puns, etc. All those new words are not taken into consideration when calculating sentiment scores, even though they are also well perceived and are a reason R&M is a highly rated show.

Then we can notice another trend while looking at the last numbers, over time the score is decreasing when it comes to Rick & Morty.
It might a hint, that overall the choices of words are changing and not in a good way. The words chosen by the writers are getting rougher over the course of time. But does that mean that the episodes are less appreciated?

Comparing Sentiment to IMDB Ratings

To answer the last question we decided to fetch IMDB ratings of the show and to compare them to sentiment score to see if there were any correlation between the two.
And well, even though the sentiment score decreases over time, the ratings do not and they even reach their all-time high for the last season of the show. So maybe sentiment wasn't really the best way to analyze Rick & Morty language. But we'll give it one last chance :

The most positive character ?

We took a deeper look at the sentiment score for Rick & Morty especially. We ran the sentiment score algorithm only for their lines, to see who was contributing the most to the happiness of the show, and see who was the bummer.

First, we can see that Rick & Morty's lines are aligned with the previous results we got, when their sentiment score decreases so does the overall shows'. Apart from that, both characters' way of expressing themselves is full of ups and downs. The only clear conclusion here is that the second half of the last season was the least sentimental one for Rick and Morty.

It's all about the words!!

Enough about sentiment, we clearly just realized that sentiment could not really give us the full taste of Rick & Morty. So we'll dive deeper into swear words and words creativity.

Most used words for Rick & Morty

A nice thing to do when you have access to the transcript is to look for the most frequent/important words. We did that by parsing transcripts and applying TF-IDF to their respective lines over all the episodes transcripts. TF-IDF is "a statistical measure used to evaluate how important a word is to a document in a collection or corpus". So we used TF-IDF most important words for Rick & Morty that we snapped into some schwifty word clouds:

Morty's Top 3
ooh
poncho
bailed

As it turns out, the words with the highest TF-IDF scores describe quite well the characters. Morty's infamous 'ohh' is the most typical response of his - no matter if he's bored, happy, or terrified - his reaction usually comes down to emotionless 'ooh'. But then he apparently mentions "poncho" a lot too ?! And "bailed" isn't too surprising as Morty's word is often "Thanks for bailing me out Rick".

Rick's Top 3
burps
sauce
new

For Rick, it's even more iconic. Maybe it's not a word but still - Rick burps all the time, and here we see that it was always carefully noted down in the transcript files. Rick also loves mentioning sauce and the world "new". The later being less suprising as he also invents or discovers new stuffs.

Most used words per seasons

We applied the same method mentioned for the last part to get the top words per season:

Top 10
meeseeks
needful
snuffles
bloom
jacob
pluto
dream
pancakes
doofus
science

Top 10
testicle
gary
jerrys
feels
purge
deer
doctor
poopybutthole
nathan
summers

Top 10
pickle
toxic
ants
citadel
wong
candidate
mechanical
campaign
manager
cornvelious

For each season we can see that the top words are heavily influenced by the theme of some episodes, for the first season words like meeseeks, dogs, and snuffles appear. Or even for the third season, clearly with the presence of one of the most famous quote of the show: "I'm Pickle Riiiiiiiick !".

Swearing Analysis

Rick & Morty is known for its rough language, so we decided to dig a little deeper into this, to find the most used swears, who uses the most and also their importance in the show's success & flow.

A little bit earlier, we introduced you to sentiment score, now let's talk about profanity score! We used the profanity check library to calculate it but what does it do? When given a string it analyzes it and returns a probability (score between 0 and 1) of the string to be offensive. So "F*** you!" would return 1, whereas "go to hell, you scum!" would return something around 0.8. So using this library on the transcript can tell you if they are more or less offensive.

First we can notice that profanity score are pretty low on a scale to one, they are all around 0.12. This can be explained by the fact that we are calculating profanity score on average, and most of the transcript are not swears, so that would defenitely lower the score.
If we are now looking at scores over time they are all the same over the seasons, except for the last one who has slightly lower average, so that proably means less swear words. So if we think back at our sentiment analysis we can probably forget the hypothesis that stated that swear words would lower the score for the last season.

Comparing Profanity Scores & IMDB Ratings

As we did for the sentiment score let's compare profanity score and ratings to see if it actually matters. And it appears that it does, ratings and profanity score, for the most part, seems to be in sync, does that mean that the more swear words the better the episode? From our point of you, it's not the only parameters but yes it counts!

Swearing Contest !

It has become a tradition for us, to compare Rick & Morty's line. Let's do it again for swearing!

The plot above gives an insight into how the vulgarity of Rick's and Morty's language was changing over time. On the x-axis, there are seasons (divided into halves: a and b) while the y-axis represents the profanity score. Morty's style is more or less consistent through-out 3 seasons with the peak somewhere between the end of season 2 and the beginning of season 3. When it comes to Rick, here the analysis becomes a bit more interesting. From the very beginning of the series till the first half of season 2 his profanity score fluctuates around the same values as Morty did. But in the second half of the last season, the number of swear words skyrockets, while Morty's score declines sharply. One could put forward a thesis that Rick's language became much vulgar under the influence of Morty, while Morty's temperament melted down.

Made Up Words!!

Another fact about Rick & Morty is their creativity when it comes to inventing new funny words, so let's see what we can find in there :