statistics


Pencil Blogs – More Revolutions 2

Work is keeping me busy, so there haven’t been many blog posts recently. To prove that I’m still alive here’s just a quick add-on to my first blog post in my Pencil Blog Series.

Let’s have a quick look at the word count1 of Pencil Revolution blog posts over time.

Word Count

The longest Pencil Revolution blog post had nearly 2500 words. The shortest blog posts didn’t have any words and just consisted of a picture.

Pencil Revolution Word Count over Time

Let’s have a look at the word count plot. The x-axis (horizontal axis) is used for the dates of the blog posts. The y-axis (vertical axis) is for the word counts of the blog posts.

Each dot represents a blog post. Early posts are on the left, recent posts are on the right. Short posts are near the bottom, long posts are higher up.

The black line is a straight linear regression line. The blue curve is a LOESS (local regression) curve, which is a bit more ‘flexible’ than the straight line. You can see that in the first years, posts got longer and longer, while in recent years posts got shorter and shorter.

Cumulative Word Count

Looking at the cumulative word count we see that all Pencil Revolution blog posts together reached a word count of nearly 170,000 by now.

Pencil Revolution Posts and Cumulative Word Count

In this chart, the solid line shows the number of cumulative blog posts, as seen in the previous blog post, while the dotted line shows the cumulative word count. The average word count per post is 266, both lines are aligned accordingly. You can see that up to 2011 the posts were longer than the current word count average, while the posts since 2016 have been shorter than the current word count average.

That’s it for now like I said – just a quick blog post to show I’m still here. I hope to post more once work is less busy.

  1. I guess it’s not always clear what counts as a word and what doesn’t, so your word count might vary. []

Pencil Blogs – It’s a Pencil Revolution 2

It’s a Pencil Revolution.

No, I’m not talking about the song by Schradinova.

A New Series

I’m talking about Johnny’s pencil blog – because this is another one of those R / statistics blog posts – as threatened previously. I actually started to work on this in 2015. Originally I wanted to present all the data in one blog post, but because there are old pencil blogs disappearing and new ones appearing regularly this somehow hasn’t worked out yet.

That’s why I’ve now decided to switch to looking at the pencil blogs one by one, always adding to the previous stats and since I haven’t yet figured out a better way of ordering them I decided to start chronologically. The thing is, though, if pencil shops have been around for decades, there may be older pencil blogs I’m not aware of. If you know of any pencil blogs please let me know in the comments, so that I can add it to my list of blogs to be covered. At the moment I have around 30 pencil blogs on my list. Some of the next blogs to be covered will be Pencil Talk, Dave’s Mechanical Pencils and Lexikaliker.

I decided to stick with pencil related blogs, but if there is any interest (usually there isn’t much interest in the statistics related blog posts here) I’m happy to expand to more general stationery blogs.

Also, if the blog owner contacts me to object I will take the blog post about their blog offline, so there is a chance that these stats posts won’t stay.

Pencil Revolution

Here’s the first plot about the oldest pencil blog I know: Pencil Revolution. If there are no objections from Johnny other stats about Pencil Revolution will follow soon.

So far there have been 635 blog posts on Pencil Revolution. Congratulations!

Pencil Revolution Blog Posts Over TIme

Pencil Revolution Blog Posts Over TIme

The plot above shows the blog posts over time. Horizontal lines are an indication for periods of inactivity while ‘more vertical’ lines show periods with many blog posts. In the case of Pencil Revolution, you can see that Johnny stopped revolutionising between the Summer of 2006 and the Summer 2010 and between Spring 2011 and Spring 2012 before picking up again. The latest spurt started in the Summer of 2017. I won’t comment further as the graph is fairly straightforward.

Congratulations on more than ten years and more than 600 blog posts.


Scraping pencils 1

Pencils not only seem to be reliable writers, they also seem to be reliable in terms of keeping their prices affordable.

In June this year, I used R to do some web scraping. The data collected was supposed to be for a follow-up of my Why did stationery become so expensive? blog post. Well, this is the follow up blog post and there isn’t much to show, maybe because we are talking about wood-cased pencils. Fountain pens and mechanical pencils had a more extreme price development.

I did the R code not only for the follow up, but also to teach myself a bit more R, I would have normally done this in PHP, which I am more familiar with, having used it since the 90s and having used it for similar tasks in the past (even though people didn’t call these tasks web scraping at that time, or if they did I didn’t know).

chart showing how the Koh-I-Noor PolyColor Art Pencil prices have changed.So what exactly did I do? I was looking at how the prices for pencils have changed on the Cultpens web site over time. My code that will get historic prices from archived versions of the Cult Pens pencil web page from archive.org is available on GitHub if you want to try it or change it for your own purposes. If you do please bear in mind that my code is not very good, I realise this, but as I am new to R I don’t know how to improve it (at the moment – I hope to find more time to learn R in the future). Also, if you try this please try to minimise any strain on the server you get your data from.

The findings were less exciting than expected. VAT has changed a few times over the last few years and there are obviously currency exchange rate fluctuations to be taken into account, as well as inflation and other factors.

The most interesting changes I could see was a > 10% price increase for some Graf von Faber-Castell products between 2010 and 2011.

The  Tomboy Mono 100 got cheaper over the years. Between 2010 and 2013 it got > 20% cheaper.

The most extreme price rise was for Koh-I-Noor products. At the beginning of 2014 the big Polycolor Art Pencil tin got nearly 40% more expensive, smaller tins got more expensive, too.

Overall prices seem remarkably stable.

As mentioned before I assume the price stability is also linked to this being pencils. As shown in my previous blog post fountain pens seem to attract more extreme price hikes.


Picture: Mongol plus Pilot Color Eno pencils on Brunnen Der Grüne Block.