Statistics: What are you talking about?

The other day, Pachiko made her graph debut on TV.

Since I was told this in the newsletter,

So I gave it a try.

Everyone, do you think Pachiko is good at drawing?
That's not it; what I'm best at is information analysis and construction.

Things to prepare

1. Chatty

2. Google Sheets

3. Newsletter URL

Thinking is a hassle, so I'll leave it to Chatty and have them write theGasu for me.
Gasu refers to GAS, but once I read it asGasu I can't see it as anything else butGasu now. The official name is Gasu Kurobikari Academy, but I have to be careful only when talking to people. I had to speak in front of people the other day, so I asked Chatty how to pronounce it beforehand. Apparently, G-A-S is fine. It wasn't Gasu.
Huh?
You ask what GAS is?
Just think of it as a spell that lets you do whatever you want with Google stuff.

Extract the
link URLs from the articles

I pasted them all into a spreadsheet for that purpose, but I'll extract the source of each article from the newsletter URL list and pull out the URLs pasted within the newsletter.

Chatty keeps trying to be overly helpful, so I thought it might have been faster to do it myself... but I managed to set it up so that it only takes the main text, and if there's a table of contents, it splits the opening and the main body.

Classify them

It seems Stand.fm links end up as Stand.fm URLs (because they are iframes), so it looks like they can't be retrieved this way.
It's a hassle, so I'll put this matter on the back burner.

This time, I will only count note articles.

As for statistics, from Pachiko's hobby perspective,

1. number of published articles

2. number of accounts

3. number of published articles per account

of,

1. per newsletter

2. all

would be good to know.
If possible,

1. a directory of names featured in the opening

and such, but since it's troublesomedifficult to separate automatically without a table of contents, I've given up on this for now as well.

For now, I have retrieved the number of note articles and accounts for each newsletter.

Please take a look.

I have no idea what this is about!!!

what

it is

I have no idea!!!!!!

Total

Published in the 52nd issue of the newsletter,

Number of note articles:
total of 13,619 articles, 13,286 articles in total

Number of note accounts:
total of 5,553 accounts, 975 accounts in total

was the result.

975 accounts... that's almost a thousand people!!!

Data accumulation and graph development

There is a difference between accumulating and analyzing data, and presenting it in a graph.
Presenting it as a graph is meant to visualize the data and convey it to the viewer at a glance.

Therefore, the basic approach is to create a graph that focuses on what you want to convey.
For example, if you want to show that the number of articles increases with each issue of the newsletter, you might calculate the average for every 5 issues to create an impression of obvious growth.
With the current graph, it's jagged and hard to grasp the increasing trend. Here is what it looks like when calculated using the average over a period.

Above all, the width has narrowed, making it easier to understand, right?

Conversely, you can also extract only the recent trends to clearly show an increasing trend.
In that case, the key is not to start the graph axis from zero.

Isn't that sneaky? Isn't that a giant sequoia?
But still, don't you all do the same thing when writing essays?
You convey the important parts firmly. You cut out episodes that stray from the main topic or might cause misunderstandings among the trivial details, right?
Graphs are the same.

In other words, you shouldn't trust the graphs presented on the internet or TV.
I often think, 'Show me the source,' but most graphs are processed even more crudely.
For some reason, many people think data equals graphs, and if there is a graph, they believe it is correct information. You shouldn't believe it unless you see the raw data with your own eyes.
There is also a lot of data that is calculated and collected using statistically incorrect methods; furthermore, you must be more careful about the source of the data and the collection method.

#WhatAreWeTalkingAbout

Impressions

By the way, Pachiko joined the correspondence starting from the 30th letter.
Previously, Maiton-san wrote that they were moved by reading Konishi-san's commentary when they first appeared in the correspondence, and I thought, 'I get it, I was moved too.'
Since I had the chance, I looked back on it.

...I was... moved, wasn't I.
Huh...?
I was moved... because I was told it was strange to have parentheses...?

I am stunned.

However, while it was fun to gather the data, I really want to read the content, don't I? Don't I? Don't I?
Next time, I think I'll try reading from the first letter.

But you know, if I do weird things, the story won't progress at all. This is troublesome, haha.

Other statistical information

In addition to what I have posted, for example, I want to know how much I have appeared in the correspondence! I can also respond to requests like that.
If there is information you would like to know, please whisper it to Pachiko, and I can convey it to you directly or as an article to be collected, so please feel free to ask.

Also, I am accepting inquiries such as wanting other data acquired or analyzed, or wanting to know detailed data collection methods.
After all, I am QC Certification Level 2, so I am also accepting inquiries asking to teach me about probability and statistics.

#WhatAreWeTalkingAbout