Hello.
I'm the developer and fishing king WEST from UniqueGood Company!
I was curious about what kind of reviews people who enjoyed Realworld games had written,
and what parts they got stuck on and left help requests for.
I thought those reading this post would be curious too, so I'm sharing this haha..
(I trust that you'll be curious about it)
So instead of just reading through everything, I thought as a developer I should at least write some code, so I ended up...
Before fully diving into the developer's path, I decided to try some 'text analysis' that I had briefly encountered!!
Data analysis that I've been interested in for a while!!! AI!! The Fourth Industrial Revolution!!
Shall we analyze the comments together?!
It sounds grand and like it would involve lots of theories and complex code,
but as the icons of faith and trust above suggest, it's really nothing special haha...
Just enjoy it for fun!
The game I wanted to analyze was 'Dad, Where Did You Go?!'
I also enjoyed this spin-off story of the Tale of the Rabbit and the Turtle~!
It's a very, very interesting game that makes you think about 'life is all about that one shot' and 'the environment' haha
From March 14th to March 17th, 272 posts were written!
There were 202 help posts and 70 review posts, about a 3:1 ratio.
(Comments on review posts were categorized and counted as help posts)
I extracted only nouns from all posts and checked their frequency.
Rabbit, problem, note, turtle, and hint appeared as the most frequent words in that order.
Just looking at bar charts feels a bit bland haha
For frequency checking, nothing beats the word cloud that everyone tries at least once!
If you've played the game, you should be able to see roughly what kind of reviews people left,
and which problems everyone struggled with :)
Let me check just the help posts separately.
It seems like people had difficulties with problems related to notes and rabbits haha (me too, actually...)
Everyone struggled with the parts where I got stuck when I was solving it.
Then shall we look at the review posts separately too??
Looking at the frequency roughly, it seems there were posts about thoughts on the environment and matters related to problem difficulty.
I'm curious what they were. Let's dig deeper!
The data isn't much, but I'll try topic modeling~
※Topic modeling is an analysis method that analyzes the distribution of word counts found in text to identify and predict what topics the text covers. (Please Google for more details ㅠ) (I set it to 5 topics.)
First, I got a list of the most frequent words and the distribution of topics.
Looking at the topic distribution, Topic 1 seems to be separate, Topics 2 and 5 appear similar to each other, and Topics 3 and 4 seem similar to each other.
Let me check them one by one starting with Topic 1.
In Topic 1, the most frequent words were rabbit, note, and problem with high frequencies.
It seems like help posts about rabbit note problems were clustered together.
Topic 2 seems to be a cluster of questions about problems related to numbers, KakaoTalk, and keys.
Topic 3 has content about turtles, KakaoTalk, etc., and also difficulty, emotion, etc., so it seems like review posts were clustered together.
It looks like people wrote reviews about the difficult parts, the touching parts, and the problem difficulty.
Topic 4 shows words like thoughts, problems, difficulty, etc., so like Topic 3, it seems to be clustered review posts.
It looks like people also wrote about difficult content within their reviews.
Topic 5 shows content like fishing spot, terminal, hints, etc., so it seems like help posts asking about difficult problems were clustered together.
Topics 1, 2, and 5 all seem to be help posts, but I was curious why they were separated, so I checked and found that someone had asked questions multiple times as comments on review posts.
That content was reflected and shown in the graph! How interesting :)
Up to here, I conducted a shallow text analysis using various Python data analysis libraries.
I'm embarrassed to even call it text analysis haha
I started with curiosity about what kind of posts game players left,
but ended up gaining relief that I wasn't the only one who struggled with certain problems. :)
It feels like a post where only I gained something.
I hope those reading this who completed 'Dad, Where Did You Go?!' also feel that same sense of relief.
This ending feels weird...;;