How having a heartbeat can kill your writing

In a previous article, I introduced one of the tools that I hoped would provide me with a lot of useful data about my writing practices, that I would later be able to mine for hints about how to make better use of my time. This article is the first serious attempt to look at some of that data, in light of my recent highly productive writing retreat.

In it, I’ll evaluate the tools I’ve created and how they are holding up, and I’ll look at some of the stats they’re showing me, warts and all. Let’s take a look…

To refresh your memory, the core of this system is a simple python script I wrote, called the production-monitor. Its job is to wake up once per hour, count how many words of writing there are in the text files in my project directory, and then make an entry in a project log file if there has been any change in word-count from the last time it made an entry. What this gives me is a file that only has entries for the hours in which I actually wrote something, and it records all of them. So I can now produce a rather embarrassing graph like this one, that I call the heartbeat graph.

In this case, however, having a heartbeat is a bad thing. The thing that makes this line graph look like an ECG display is the fact that the blips of writing are so clearly distinct from each other. If I had been writing every day, this graph would look more like a stock market report.

Beginning in early January, when I first began measuring my productivity, this graph shows that I am clearly confused about the meaning of the word “productivity.” You can see, for example, that only once did I write more than about 1500 words in a single day, from Jan thru July, and I have to confess that on that particular day in late July, I didn’t actually write anything at all – I simply copied some deferred scenes from the Strange Places working files over into the Strange People folder. The fastest 3000 words I’ve ever “written.”

This highlights one drawback to the production-monitor script – it can’t tell the difference between writing and importing. In my experience so far, I haven’t found this to be a problem, but others might.

Another minor anomaly in the pre-August period is the unusual-looking negative writing day that immediately follows the copy-day blip. This simply reflects the fact that on that day, I reviewed the newly imported material and deleted some bits that were no longer relevant. Hence, the word-count in the directory went down, and so the graph plunged below the line of death. Again, I personally don’t find this troublesome, as it might actually be showing me an “edit day.” I suspect that when I finish the first draft of Strange People, I will begin to see more of those “edit day” downward blips as I compact and reduce the first draft into a tighter second draft. I promise to report back on that later.

Overall, the first section of this graph tells me something I already knew – I really wasn’t doing enough writing during the first 7 months of this year. I could explain that, but excuses are always boring to read, so let’s assume that I’ve given them, and you’ve rolled your eyes at their flabbiness, and just move on.

I wish I had been using this script during the writing of Strange Places. I wish I knew what my most productive day had been, how many productive days I had, how many consecutive days, what my average words/hr rate was, etc. I would love to be able to compare productivity now (on good days) versus productivity then, but all these questions shall have to wait until later as well.

What I can show you is what a good day looks like this year. Turning our attention to the happier part of this graph, one might ask, “Whoa! What happened on August 12?” The answer is, I went on retreat. Now, it isn’t as though I just hopped in the car one day with a sudden notion to go elsewhere to write. This trip had been planned for a couple of months. In fact, those small upward blips right around the beginning of August were me making some notes in the current draft about what areas needed to be fleshed out and what direction certain storylines would be taking once I got to Banff. (Yes, my retreat was at the wonderful Banff Centre – a place I now recommend highly for many, many reasons.)

I arrived in Banff on a Sunday evening at about 7:00. Even after getting checked in, finding my room, and taking care of all the other administrivia that travel entails, I still managed to put down 1672 words. I worried at the time that this might just be a flurry of activity that had been jammed up in anticipation, waiting for release, and that I wouldn’t be able to sustain such output. Remember, until that day, I hadn’t ever had a day recorded with that much actual writing. I was pretty sure I had actually produced more than that on some days of the Strange Places journey, but I didn’t know that I had. I didn’t have documentation and graphs to back it up. So, nervous about keeping up the pace, I launched myself into the first real day of writing.

And threw down 3866 words.

Now, I need to appeal here to the Swedish judge for a ruling. See, this reveals a previously unconsidered flaw in the system. I actually wrote 5037 words on that first day, but I was penalized on a technicality, because 1171 of those words happened between 11 pm and midnight, which means they didn’t get logged until 12:01, which was technically the next day. So even though I did in fact write them on the 13th, they were included on the system totals for the 14th. I intend to correct this timing gaff by simply having the logger run one minute before the hour, rather than one minute after, but I don’t want to queer the data I’m collecting by making that change in mid-project.

So, technicalities and Swedish judges aside, I put out a whopping 5037 words on the first full day of my retreat. More than 3 times the output from any other productive day in the data previously.

But wait. It gets better. So far on this retreat I have averaged 5224 words per day – and that includes that skeezy first day of only 1672. Over the ten full-length days of my stay, I logged an average of 5579 words per day, and a best day of 9503.

That number still boggles my mind: 9503. That’s more than two full chapters in a single day. Now, to be clear, this is first draft we’re talking about, but still it’s about six times as much output as the best days prior to my trip.

So, what does this all tell me? That I’m much better off writing in seclusion than at home? I don’t think that’s it. That I should expect to get 5K words or more on every writing day? No, that’s not it either. What it’s telling me is that I have to kill that heartbeat effect. I have to find make more time for writing on days when I write, and I have to put longer consecutive strings of such days together.

I’ve got more data to share, but that’s enough for this installment. I hope some will find it useful. Next up, I’ll be reporting on how this data is shaking some of my most strongly held beliefs about the times of day that I am at my most productive. It’s sure to be a shocker.


New Interview: Writing time is family time
The author chats - with his protagonist

About the author

Jefferson Smith is a Canadian fantasy author, as well as the founder, chief editor and resident proctologist of ImmerseOrDie. With a PhD in Computer Science and Creativity Systems compounded by a life spent exploring most art forms for fun and profit, he is underqualified in just about everything. That's why he writes.