Bionic Punchline

Statistical Analysis

Well, here we are. You’ve (presumably) read all the fun stuff, and now you’re stuck with this statistical analysis of the event. You’ve gobbled down all the nice chocolates in the box—and they were sweet—but now you’re stuck with nothing but coconut. Dry, mathematical coconut. Sure, you could just skip this section. You could just chuck all those coconut chocolates in the bin, still lovingly encased in their shiny blue wrappers. But you won’t. You’ll wolf down this unappealing block of graphs, and do you know why? Because I heard some guy say you couldn’t.

Let’s get started.

(Wattpad doesn't allow images in chapters. Just imagine that this bold text is a really rockin' line graph.)

Fig. 1: As you can see, this graph wasn’t quite ready for its moment in the spotlight and is currently having a mini freak-out.

Fig. 1 is a style of graph I’ve been using since I first started doing this in 2012. While the word count across individual days (red line, square markers) makes trends hard to interpret because it forms such a spiky line, taking an average across five days (green line, triangular markers) smoothes things out and makes it much easier to see what—if anything—is going on.

What appears to be going on is that word counts across the month remained reasonably steady (or steadily erratic, if you want to look at it like that) with a bit of a slump in the middle. Besides a fairly impressive peak at around 900 words on Day 10, there’s not a lot else going on. Compare that to previous years:

 (Imagine a different line graph. More Middle Earth-y this time.)

Fig. 2: This graph illustrates variations in topography across Middle Earth—from the Shire to the Misty Mountains to Mordor. Just kidding, it’s more average word counts.

I’ll admit, this one’s kind of confusing. However, that’s partly because the average word counts for all three years start off so similar. For about a week near the beginning, there’s not more than about 200 words difference between the three years. It’s only towards the end that they really start to separate out, with my first attempt at Flash Fiction Month finishing off with a long string of short stories, my second attempt ending on a high note, and this year’s work wrapping up pretty solidly in between the two. Which sort of screws up my analysis from last year in which I concluded that the first Flash Fiction Month, combined with eleven months of miscellaneous writing experience, had made me better able to produce work more consistently on demand.

So yeah. I’m starting to think that my B in GCSE Maths hasn’t really set me up to do this kind of thing in a statistically rigorous fashion.

(This line graph appears at the very top of the page, because it's so awesome.)

Fig. 3: I may as well just have been sticking pictures of cats in here all this time.

But it’s not all doom and gloom. Having now participated in Flash Fiction Month three years in a row—producing nearly a hundred individual stories in the process—I’ve collected enough data to do some things that I couldn’t before. For example, up until now I’ve been relying on averages across five days to explore trends over the course of each month. But now that I’ve got three events to work from, I can work out an average for each day across all three years!!!

Okay, it’s probably not exciting enough to warrant three exclamation marks, but here it is anyway:

(Imagine a smoothed-out version of the second graph you saw.)

Fig. 4: This graph is smooth and mellow, like the voice of James Earl Jones when he’s not playing an asthmatic cyborg.

This graph shows a word count for each day of Flash Fiction Month (blue line, diamond markers) averaged across the 2012, 2013 and 2014 events. Obviously this is going to even out some of the extremes since I’m unlikely to produce a 1,000 or 55 word story on the same day each year. Day 11 is notable for its average word count of 979, as is Day 31 (the final challenge in both 2012 and 2014 demanded a 369er, requiring a word count of exactly 207).

To make any trends across the month clearer, I’ve also worked out an average across every five days of the three year average (red line, square markers). Put like that, it sounds anything but clear, but it does exactly the same thing as the five-day averages in Fig. 1 and 2: instead of taking into account every individual day, when I might have ended up with only twenty minutes to mash a story out of my keyboard or been handed a 999 word challenge, it averages out how I was doing on any given five days.

Changes in this five-day average are likely due to my personal performance varying over the course of the month: either a period of getting worn down or gaining a boost, or a run of unusually long or short stories. With the exception of the dip at the very end—caused entirely by Day 31’s challenges bringing down the average for the previous two days—it looks as though I had a fairly steady run. There’s a gentle peak around Day 10, perhaps due to me getting into the swing of things, followed by a slow decline up until about Day 23, then a boost during the last week (ignoring that anomalous Day 31).

(Imagine...you know what? Imagine whatever you like. I seriously doubt that made-up graphs are really going to help illustrate what I'm saying here anyway.)

Fig. 5: A man walks into a bar graph and...oh, you’ve heard that one?

While I’m basically making the assumption that “more words” equals “doing better,” which is absolutely not true, it’s...well, actually kind of true. It may be possible to write a brilliant, Baby Shoes style masterpiece in only a handful of words, but during this kind of event a string of long, complex stories is naturally more of an achievement than a string of snippets thrown together just to get through the week. With that in mind, looking at the average words per day in any given year (alongside the minimum and maximum numbers of words possible, for perspective) should give some idea how it went. This year is middly. The stories have definitely been longer on average than they were for the first event, but it’s nowhere near matching the average from last year’s, which was pretty astonishing.

Still, it’s not all about length. While I definitely feel like I had a harder time keeping up for portions of this year’s event, I also think the quality has stayed reasonably consistent. I didn’t write anything I really wasn’t proud of, and I think even the stories that were written in less than ideal conditions—with very little time, on the road, in hotel rooms, that sort of thing—were inventive and entertaining. Overall, despite having now produced a hefty number of stories for Flash Fiction Month alone, I don’t feel as though I’m running out of new ideas. In fact, I feel like I’m coming up with more than ever. And really, I think that’s what this event is all about.