All (cross-posted to cz-l and cz-editors),

It occurred to me that this would be an excellent time to answer a question
always important to ask about new projects: "What would success look like?"
Or, more particularly, for those positivists among you: "What should count
as *evidence* of success?"

And actually, since this will be a perpetual project, anything like
"success" is surely years away, no matter how quickly we add people, edits,
and articles we add.  The more interesting questions are: "What does
progress look like?" or "What counts as evidence of progress?"  We can think
about progress in the present and even so far, not just about in the future.

There are some obvious metrics, such as number of

* live articles, total
* live articles started per day
* contributors, total
* editors
* authors
* contributors/editors/authors who have contributed ever
* contributors/editors/authors who have contributed within the past week
* edits per day

Also, the rate of increase (if any) in all of the above.

I consider our edits per day, the number of which bounces anywhere from 100
to 300 (I estimate), the growing number of contributors and live articles,
and the steadily increasing number of live articles.  I consider all this
and it seems to me that none of it *really* seems to matter very much.  I
mean, don't get me wrong, it's *nice* to have more people, edits, and
articles.  But ultimately, what we're after is something that distinguishes
us from Wikipedia, and *simply* adding people, edits, and articles doesn't
distinguish us from Wikipedia one bit.  After all, we may never equal
Wikipedia in sheer quantitative terms (though I think we should try!).

What I would count as distinctive progress, and thus noteworthy progress, is
the increase of our distinctive metrics, namely: number of active (expert)
editors and number of approved articles.

Unfortunately, there's no easy way to measure the former (right now), and
the latter, well, is going to be zero for some time.  Before we approve any
articles, editors need to start a discussion about what the standards of
approval ought to be, and then start actually applying them.  So I would say
that substantive progress will occur when we are working our way through
that discussion.  Another milestone will occur when we approve our first
article.  Then I think things will get very interesting, because number of
approved articles will probably come to be regarded as *the* main metric of
progress.

But that's not really right, either.  In fact, I propose that the most
interesting metric of progress is the number of *important* articles
approved, out of some list of --who knows?--10,000 articles.

This is, by the way, something that we could take as a goal and make into a
useful sort of game.  We have already got a list of "high priority"
articles: http://tinyurl.com/vkt3l

It's getting quite long.  Really, we need to break it down according to
disciplines, such as the list here: http://tinyurl.com/y4nftw

Then we have each discipline workgroup "nominate" approximately 250 (maybe
up to 1000, for "big" areas like biology and history) articles as top
priority.  We can then measure our progress at approving *really important*
articles by counting how many of those articles (i.e., at least one
*version* of each) had been approved.

Then we might expand the list to 100,000; and after that, well, it won't
matter, as we'll be very well on our way, and we can take sheer numbers of
approved articles as our main metric.

My point, then, is that, here in the early days, even the number of
*approved* articles isn't the interesting metric.  It's the number of
*important and approved* articles.

The practical implication of this ramble is: one of the first things
discipline workgroups should do is designate "important" or "high priority"
articles in their area.  Just please don't misinterpret me as saying that
the workgroups should then proceed to *assign* articles to people, because I
am strongly opposed to that.

--Larry


_______________________________________________
Citizendium-l mailing list
[email protected]
https://lists.purdue.edu/mailman/listinfo/citizendium-l

Reply via email to