Not Quite Nate Silver: Predicting My Monthly Writing

35750854_241aaf618c_m
After all the interest in statistics and prediction from this year’s presidential election, this post feels both mundane and anticlimactic, written by a novice. So be it; I am no statistician. I’m just seeing if graphs can motivate my writing. So far, so good.



I’m participating in
Digital Writing Month, which sets a goal for its participants to write 50,000 words in November 2012. I joined for a number of reasons, mostly to force myself to write regularly and at greater length. I have comp exams coming up, and I’ve fallen off the proverbial wagon of regular academic writing. Asking myself to write 50,000 words in a month sounded like a good way to get back on that wagon.

50,000 words breaks down to 1,667 words a day. Knowing I’m not going to hit that mark every day, I did something a bit counter-intuitive: I set the bar higher but gave myself permission to miss it. I’ve been aiming for 2,000 words per day, but requiring myself to write 750/day as an absolute bare minimum. That minimum goal comes from
750words.com, which tracks word counts for text typed on the site and rewards points for continued production. 750words also hosts monthly challenges, where participants sign up for a commitment to write every day for a month, determining rewards and consequences for good and bad behavior. (For the record, I’m treating myself to a nice dinner out with a good beer upon successful completion. If I miss a day, I’m forcing myself to go for a run the day after to clear my head and get going again. Simple.)

The points system from that site is fun for checking off the days, but I wasn’t getting a sense of accumulation that I would need to hit the 50,000 mark. It also sets the bar lower than I needed; 750 words/day wouldn’t come close. I needed a way to see how many words were racking up, how far I had to go, and where I fell between my minimum and target daily counts.

I built a quick spreadsheet to help. The first graph I made has a line showing the 750 mark and another at the 2,000 goal. On top of those thresholds, I track my fluctuating daily counts and display a moving average to get a better sense of how a single day affects my overall pace. Notes in certain cells helped me recall why peaks and troughs existed: on 11/2, I helped with the collaborative novel-in-a-day project; on 11/6, America did this silly thing called an election. I was a bit distracted*. Here’s what I saw developing:

Screen Shot 2012-11-08 at 8 November | 10.08.14

My fluctuating pattern of output intrigued me. It looks like I can’t write two days in a row. More troublesome, though, was where the totals fell: After a week of writing, I had yet to hit my 2,000-word target. Apparently, the 7-day mark triggers my discouragement. I began to lose confidence in my ability to keep up with my word-count hopes for the month. Indeed, my daily average was trending toward being 7,500 words short by the end of the month.

Then I thought about Twitter.

The second major dip in my word count, seen on 11/6, happened consciously and willingly: the US presidential election happened that day, and I allowed myself to be distracted by the progress of the results. I knew my counts would be low that day because I spent more time reading websites and my Twitter feed than I did with any kind of document open to write in. With Twitter, though, I write as I read; I interact with the people I follow. And the night of the election, I interacted
a lot. I felt like not counting that writing toward my word count was very literally discounting Twitter as an avenue for digital writing, complete with its own expectations and constraints. So I extracted my tweets for the month and did a word count on those by day. Here’s what happened:

Screen Shot 2012-11-08 at 8 November | 12.41.23

By adding tweets to my word counts, I’ve been
above my target on all days but one. The peaks-and-valleys shape to my writing habits vanished. On days when I write less content that is serious or extensive, I write more that is short and trite, almost balancing the perceived decline in written production.

With these figures, my month-end total count is on-track to be nearly 10,000 words
over my goal. I’m honestly torn by this revelation. On one hand, I write more than I thought. On the other hand, what I initially saw as a serious challenge suddenly lost its intimidating edge. Writing 2,000 words a day now almost seems normal. I need to find a way to make me feel the pressure to write while also acknowledging that my writing takes different forms on different days.

Or maybe I just need to tweet less.

Anyway, the spreadsheet I made has another graph tracking total words written/remaining; I’ll definitely tweet about the day I cross the halfway point. Should be sometime this coming week. And because I joined the 750words.com November Writing Challenge (to write every day), I get daily emails telling me how many people are still in the running. I’m graphing that change over time, as well, to remind me that this
is a challenge, and that it is possible to fail at it.

Overall, I’m hoping these graphs will provide that extra nudge to motivate me to keep going, to work harder, and to write more. So far, they seem to be working. If you’d like to use them yourself, have at it:
Files updated 12 Nov 2012 to correct a problem with the third graph figuring totals incorrectly.

Do graphs like these help keep you motivated? Are there other statistics I should be tracking?

[Photo credit: scottwills on Flickr.]
blog comments powered by Disqus