The Graphic Novel

For an event that explicitly reduces the art of long-form prose fiction to the pursuit of a single number, National Novel Writing Month puts out really terrible graphs.

Note how neither axis has its single most important number labelled.

Problems with this graph

It’s a bar chart.

Look at those thirty little rectangles, clearly indicating exactly how many words I had at the end of each day. But except on day 30, we don’t really care about the total: we care about the change. We’re not interested in comparing the number of words on day 10 with the number on day 20. We know what the relationship there will be: day 20 will be higher. We’d be much better off with a line graph:

Now it’s super easy to see which parts are steep and which parts are shallow. Much more useful.

It’s the only one.

NaNoWriMo.org crashing is as much a part of the event as writing stories is, so I won’t hold this one against them. If they still had ridiculous fluff like big animated books for your extract, then I might, but they seem to have scrapped those.

Nonetheless, it would be nice to have graphs showing, say, how many words I produced per day.

The biggest bar, incidentally, is when my wife went away for the weekend. You should keep that in mind when performing your own interpretations of these statistics.

Or we could take the target line on the basic wordcount graph and make it horizontal, giving a nice clear picture of how far behind or ahead I was through the month. Not that I was ever ahead.

Or we could ask a more interesting question, and plot the relationship between how many words I wrote and how far behind I was at the start of that day.

The “behindness index” here is calculated by dividing the average number of words I need to write per day from that day onwards by 1666⅔, the average number words you need to write every day to hit 50,000 on the 30th. It therefore accounts for how many days are left to make up for past unproductivity.

I think there’s a hint of a positive correlation there. Of course, correlation isn’t the same as causation, and there may be a third factor at play here: the further behind I was, the more frequently my wife told me (in a nice, supportive way) that I probably wasn’t going to manage it this year and I should condsider downgrading my expectations. I suspect she was secretly doing this so that I would strengthen my resolve, just to be obtuse.

On reflection, perhaps they shouldn’t go this far. It’s a bit sad, isn’t it?

It’s inaccurate

Perhaps the silliest part of the National Novel Writing Month tradition is ‘validation’. This is the process of pasting your entire manuscript into a little edit box and letting its wordcount algorithm do its stuff. It automatically updates your wordcount for the day and, if you’re over the line, declares you a winner.

Because I wasn’t about to engage in this absurdity every day, the last bar on my chart is based on a different wordcount algorithm than the others. (A less generous one, as it happens.) And why? To make it infinitessimally harder to lie about your final wordcount? I find it hard to believe that any of the organisers care about that: it’s just a little end-of-the-month ritual. Why not make it optional? (If it is optional, why not make it clear that it is?)

What have we learned from these graphs?

• That whatever I was doing from the 20th to the 30th probably has the most to teach me about being productive.
• That targets are motivating, but being a long way away from them is apparently more motivating.
• That scrupulously recording your output pays dividends in fun statistical analysis – which is, after all, the only reason any of us ever do anything.