My Last 10 Years in Books

For a while now, I've wanted to play around with visualizing data on the books I've read over the last few years, particularly to discover what genres or themes I lean more toward, and to see how my reading habits have changed over time. Recently, two things happened that made me decide to act on this curiosity.

• First, a friend asked me what my all-time-favourite book is. My answer was, to her disappointment, "it depends." Because it does! What makes a book impactful or transformative depends on when you read it, and what's going on in your life at that time. [1] This got me thinking about my reading journey over the past few years.
• Second, I discovered a new alternative to Goodreads called The StoryGraph, which offers more sophisticated and insightful reading stats. I wanted to transfer my reading records from Goodreads to StoryGraph and in the process, realized that I could export the data I needed to play around with visualizing my reading data myself.

The process of collecting and organizing the data was more onerous than I anticipated. Initially, I imported my data from from Goodreads to StoryGraph, but the import was rather messy. I had to go through each record on StoryGraph to ensure its accuracy and then export the updated data for my visualization purposes. Unfortunately, StoryGraph's supplementary categories which I was especially excited about (book genre, mood, pace, etc.) are proprietary and were therefore excluded from the export. [2] I figured I could then just use whatever categories StoryGraph allowed me to export, but I soon realized that my struggles with data were far from over. I kept finding new problems and discrepancies that I hadn't seen coming. Every time I fixed one, another popped up, and it became a seemingly never-ending cycle that spanned over 4 days:

• When I started cleaning the data export from StoryGraph, I noticed that for some books, I was missing data on the format (digital or print). I learned that this was because different editions had different formats, so I went through each record on StoryGraph, checking to ensure I captured the correct edition.
• When I exported the data again, I noticed that some essential categories, such as year of publication and number of pages, were missing. I learned that some editions didn't include this data because StoryGraph relies on user-generated data for book editions, which is sometimes incomplete. I went through the incomplete records and where possible, added missing data to the StoryGraph database. And although my records were now up to date, I still couldn't export data for these crucial categories.
• When I exported the data again, I noticed that ... you get the idea. After several such cycles, I decided to switch to using Goodreads data.
• Since my StoryGraph records were now more accurate than the ones on Goodreads, I thought the most efficient way of updating my Goodreads records was to import the StoryGraph data directly into Goodreads. That done, I now faced the problem of double entries, so I had to go through all the the records and delete redundant ones while adding any other missing data.
• When I exported the Goodreads data (last one, I promise), I noticed a few more missing data points, for example some editions did not include data on number of pages, or original year of publication. At this point, I decided to simply update the exported csv file manually.

After struggling through the tedious process of organizing and cleaning the data, and feeling like I had achieved a decent level of accuracy (not perfect, but good enough), I finally got to the fun part: visualizing the data. I began with browsing the net for some inspiration, and am particularly grateful to:

• An Ocean of Books - a beautiful mapping of books that I so thoroughly enjoyed exploring, and was reminded of what a vast world of books we're lucky to have. I especially loved the colour palette and decided to adapt it for my own visualizations.
• Books I've read in the Past 10 Years - a Tableau dashboard by Judit Bekker. I liked how this dashboard presented the key takeaways clearly, and wanted my final visualization to do the same.
• Books I've Read (2011-2021) - a Tableau sunburst chart by Alex Chapman. I loved how inviting and captivating the visual of the sunburst visual was, and wanted to adapt it in one of my visualizations.

Now, take a look at the trends and insights from my last 10 years in books below:

• My favourite authors matched up with my expectations. I was not surprised to see Elena Ferrante's books consistently appear in my top-read authors. [3]
• I was surprised to find that I'm currently reading almost as many print books (paperbacks and hardcovers) as I am Kindle editions, as I thought I read a lot more on Kindle these days.
• I didn't include this in the charts, but I noticed a slight upward trend in my ratings over time. My guess is that I'm getting better at abandoning books I don't like. As StoryGraph summarizes it in their tagline "because life's too short for a book you're not in the mood for"
• Finally, I was also surprised that most of the books I've read were written in the 2010s, which may be a good thing for non-fiction books, as they offer up-to-date information.

You can interact with the full story on Tableau Public here.

Overall, while my initial plans for the visualizations were more ambitious than what the data allowed, I'm still pleased about the intermediate deliverable I ended up with. Who knows what version 3, or 4, or 5 of this will look like in the future when I have access to more sophisticated data!

Notes

[1] For example, two of the most impactful non-fiction books I read last year were Four Thousand Weeks and The Pathless Path. I read the first at a particularly tumultuous point in my life when I was chronically busy and overwhelmed with all the things I wanted or needed to do. It was precisely the right time for me to step back engage with Oliver Burkeman's ideas about re-thinking our relationship to time. And a few months afterwards I read The Pathless Path around when I was thinking of leaving my job to go on a sabbatical. That too felt like exactly the right time to read this book. Reading about Paul Millerd's journey gave me the confidence/permission-slip to step off the treadmill and carve out a bold new path. Contrast this with Americanah, a book I've read twice. First, in my sophomore year in college when the themes around race and identity felt especially relevant to my own experience as a new immigrant in the U.S. When I re-read it in early 2021, it didn't quite hit the mark.

[2] Below is a chart that breaks down the books I've read by genre since 2014, which illustrates some of the additional data categories available on StoryGraph. I was a little surprised (okay, maybe even a tad embarrassed) to see the "self help" genre up there in the top 3, with a total of 14 books, most of which I've read in the past two years. But I suppose there’s nothing wrong with a little self-improvement, after all. Some of the books in this category have really transformed how I approach life, like The Pathless Path, Four Thousand Weeks, Focusing, Creativity: A Short and Cheerful guide, and Steal Like an Artist.

[3] Ever since I discovered Ferrante through the My Brilliant Friend series, I’ve been truly enamoured by her writing, and I unreservedly recommend her books to anyone interested in sharp, candid, piercing exploration of human relationships.