Log in

No account? Create an account
An author of no particular popularity

Jay Lake
Date: 2008-01-17 11:38
Subject: [process] More on accounting for productivity
Security: Public
Location:Nuevo Rancho Lake
Music:lunch break mambo
Tags:process, writing
Some good comments on my counting words post. I'm having a quiet lunch break by myself today, and am continuing to contemplate that feedback as I wrestle with this issue.

It comes down to conversion rates.

Basically, if I measure productivity via word count, I am normalizing for the varying qualities of my hours expended. I'm not paid by the hour, I'm paid by the word (more or less, it's a bit different with novels, but you can boil it out to the same thing). So a slow, distracted hour's writing which produces 600 words of draft and a fast, in the zone hour at 2,500 words of draft can be accounted for equivalently — words is words, and they are the raw material from which I produce my finished product.

If I'm measuring productivity by the hour, I can account for a number of activities which do not efficiently map to word count. The point of discussion is rewrites/redrafts, but in a large scale sense, I could also be measuring what autopope calls the epiphenomena of writing — copy edits, galleys, publishing-related correspondence, short story submittals, and so forth. These are my Writing-Related Program Activities.

With the exception of actual manuscript time expended, I simply don't account for the other epiphenomena. I don't feel a need to. Those are overhead, separate from the writing process itself. But for rewrites/redrafts what I need is a conversion rate, to express the relationship between word count and time. Then I can normalize all my productivity data and review it in toto.

Except that relationship between word count and time is far too elastic. The algorithm would be silly, and necessarily incomplete.

One solution is to assign some arbitrary value which reflects a median value derived from the set of the possible conversion values...as a practical matter ranging from 500 words-to-the-hour to 2,500 words-to-the-hour. I'd be inclined to peg this at 750 or 800 words per hour. But then I'd have to remember to time my rewriting efforts, which is a lot more trouble than pulling a word count out of MS Word at the end of a novel work session or the wrap point of a story.

I'd also like to say I'm just not that anal, but I can't do so with a straight face. If I had an LJ tag for /wank/, I'd use it on this post.

What I want is a simple solution to a (relatively) complex problem.

Maybe I'll work out the full algorithm, just to see what it looks like...

It's a damned good thing I don't think about this stuff while I'm writing, or I'd never get anything done.
Post A Comment | 4 Comments | | Flag | Link

S. Boyd Taylor
User: sboydtaylor
Date: 2008-01-17 20:28 (UTC)
Subject: (no subject)
Y'know, you could just make a spreadsheet with the following columns

Time Expended
Task Accomplished (Epiphenomena)
Word Count (if there is one)
Reply | Thread | Link

S. Boyd Taylor
User: sboydtaylor
Date: 2008-01-17 20:29 (UTC)
Subject: (no subject)
Make that Tasks Accomplished.
Reply | Parent | Thread | Link

Jay Lake
User: jaylake
Date: 2008-01-17 20:37 (UTC)
Subject: (no subject)
Yah. At least I'd have the raw metrics, whatever I decided to do with them later.
Reply | Parent | Thread | Link

S. Boyd Taylor
User: sboydtaylor
Date: 2008-01-17 20:56 (UTC)
Subject: (no subject)
Precisely. Metrics is da bomb.
Reply | Parent | Thread | Link

my journal
January 2014
2012 appearances