Data Analysis: Am I a Discovery Writer?

Last week, I published a longer piece on the discovery vs outline writing debate. In that article, I mentioned that I “lean towards” discovery writing.

That got me thinking: what does “lean towards” mean?

If you’ve been keeping up with my newsletters, then you know that for the past 394 days I’ve been keeping detailed data on my writing habits (That’s right, I’m an even bigger nerd that you thought. And this isn’t even my final form.). Every time I write something for a personal project (for instance, this blog), I record the time of day, the number of pages/words, the amount of time I spend writing, whether I am writing, editing, or planning, etc.. And yes, I write whether or not that particular writing session was planned, discovered, or inspired.

In the past 394 days, I’ve done plenty of writing, editing, and planning for fiction, poetry, and nonfiction, adding up to 913 “pages” (“1 page” in editing is about 3 pages or half an hour. It gets iffy when you put things like “planning” into terms of pages.). But for these purposes, I just focused on writing, specifically writing in the fiction category. This dropped the overall page count to 416.6.

Of these 417 pages, 275 were discovery written, 21.6 were inspiration written, and 120 were planned. 

That seems pretty cut and dry. But my writing habits change all the time. How have those proportions changed over time? Is discovery writing always more than 50% of my writing? Have I been planning more often or less often in the past year?

That question is why I spent several hours making the following plot:

Discovery vs Outline Area
Percentage of pages I discovery, outline, or inspiration write over time. I’m not sure why the inspiration category seems to disappear just short of 250 writing instances. 

The above plot is pretty typical for averages over time. They start off really sharp because, well, if you’ve only written one page, then 100% of your pages fall into that one category. But the more instances that are collected, the more we can see trends. 

My discovery writing rate appears to hover around 75%; planned around 20 – 25%; and inspiration down at 5%. I faintly remember going through a particularly strong inspiration writing phase when I first started recording data on this, which explains why inspiration started off so strong. Interestingly, it does look like I’ve been planning more and more over time.

Some Notes on the Data:

Note: an “instance of writing” is a writing session. This can last anywhere from one minute to several hours. A day can have multiple instances, and some days may have no instances. Each instance’s page count contributes to the total proportion for its particular category.

Many writing instances in the first days of data recording were removed, because at the time I wasn’t keeping track of whether each session was discovery, planned, or inspired, and they would just show up on the graph as a big, ugly “NA.”

Please note that “planned writing” should not be mixed up with “planning.” “Planning” involves no writing at all, but rather things like research, character design, and, of course, outlining. “Planned writing” is writing that resulted from a miniature outline of sorts, or scenes that I had thought out ahead of time.

Variation from Project to Project:


Now, in my last blog, I also claimed that my writing habits changed from project to project. Did the statistics agree with this? Yes, yes they did:

Discovery vs Outline Alloland

Above is the same graph, but only for instances when I was writing my book Alloland. As you can see, the instances number dropped from about 300 to about 100. Alloland‘s trends reflect the main trend the most because, well, it’s the project I’ve written the most in. The main difference is that some of the planned writing appears to have been traded for inspiration writing., which now hovers at a steady 10% or so, twice its overall average.

Rhyming Reason (A Working Title):

Then, we have my other big projects. This graph is for a novel I’m writing as one of my undergraduate theses:

Discovery vs Outline Reason
Note: In this graph, “Planned” is now teal instead of purple. The program I made these graphs in, called RStudio, automatically assigns categories colors based on alphabetical order. The second category automatically gets teal, unless I manually change it. Since there is no inspiration writing on this graph, it made planned teal instead.

Boy, is that different from the main trend! I just started writing this one halfway through the semester, so there are only about 25 writing instances. As you can see, in the beginning, nearly all of this project was planned, but discovery writing quickly took over. I was actually surprised by how low the planned writing got on this one, since my thesis should ideally, involve a plan for getting across a very particular message.

The Domesticated Universe:

Then, we have a short story I’m currently working on called “The Domesticated Universe.” I haven’t mentioned it before, but when first setting out to write this, I thought to myself, “It’s been a long time since I’ve completely discovery written something. Let’s see where this goes.”

I think it turned out great, and it was a lot of fun! As I was writing, ideas about where the story would go started popping into my head, leading to the distribution you see below:

Discovery vs Outline DU

I’m not sure why the “Planned” category just disappeared around the end. Maybe an “NA” snuck in?

Anyways, as you can see, the story started off being purely discovery written, and planning became more and more important around the end.


Lastly, we have a project that I’ll be publishing under a separate pen name to remain anonymous:

Discovery vs Outline Anonymous

Where is the inspiration writing on the graph? Beats me, but RStudio sure seems to think it’s there. This is a longer project again, with over 100 instances. Apart from the completely missing inspiration writing, it seems to follow the overall trend pretty closely: 75% discovered, 25% planned.

So all in all, these things do seem to vary from project to project. However, it’s also possible that longer projects will also eventually fall into the norm. I haven’t been working on Rhyming Reason or The Domesticated Universe for very long, so it’s possible that they’re still in their volatile early stages, before the data settles down and starts seeing real trends.


Data tracking can be a good way to confirm or refute your perceptions. In this case, the data kind of did both. Going into this, I thought that I was 60% discovery, 40% planning, but found that the numbers were skewed much further. I didn’t realize that I inspiration wrote so much, either.

My perceptions of individual projects were a lot more qualitative, so they had a better time matching the data. My thesis had more planned writing because, well, it’s a fucking thesis. My short story had more discovery writing because, well, I set out to discovery write the story as much as possible.

All in all, it’s a victory for my memory/perceptions, which is normally 50/50 at best, but also a victory for my data collection, because these are just the kinds of cool things I wanted to visualize in the first place.

3 thoughts on “Data Analysis: Am I a Discovery Writer?


Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s