[WikiEN-l] Featured churn

[Sage Ross]

On Tue, Jul 14, 2009 at 10:07 AM, Ian Woollard<ian.woollard at gmail.com> wrote:

>
> It's looking to me like 3.5 million is about the plateau, since the
> curve is bang on that, but we might make 4 million *eventually*.
>
> http://en.wikipedia.org/wiki/Wikipedia:Modelling_Wikipedia%27s_growth#Logistic_model_for_growth_in_article_count_of_Wikipedia
>

I don't think the bell-shaped articles/day curve of the logistic model
is a good description of the trends.  Since article creation peaked in
2007, the falloff in article creation has been much slower than than
ramp-up.  Rather than falling back to close to zero articles/day over
the next 5 years or so (as the logistic model predicts), it looks like
we're heading to an asymptote of (I'm eyeballing it here) around 1000
articles/day.  I expect 4 million articles a lot sooner than
*eventually*.  ;)

The significant non-zero asymptote makes sense to me in terms of what
kinds of articles remain to be written:
*High-profile topics that aren't notable yet.  The news generates a
steady stream of newly notable topics that a wide variety of people
want to write about.
*The vast pool of notable topics (tens of millions, according to some
back-of-the-envelope estimates that have made the rounds here before)
for which article creation rate is limited by the number of
interested/knowledgeable writers.

Between those classes, I'd be surprised if yearly growth fell below
300,000 any time soon.

-Sage