Countering active admin flight
Variously, what it really means, what it says, what it's showing... and also in some contexts, its salience, "what matters and why it matters", and so on.
In this case I'm not sure what it really shows. The raw data's superficially obvious but what would this information really mean if examined carefully, and how "key" a fact is it?
That sort of thing.
It shows a very sharp, consistent, and accelerating decline in the number of active administrators which began two years ago; that much is obvious. By any statistical measure of significance, it is troubling.
I do not believe that all admins will be gone in five years, but unless I see some sign that anyone is willing to take this problem seriously, I will not have any reason to doubt what I have experienced first-hand will continue: that administative stress, decision making, queue length, aggression, and talent will all become substantially worse; even in the near term, because of the second derivative of the statistic.
Is there any support for taking closer and more careful measurements of this phenomenon?
We have the same problem on my home project: the number of administrators is roughly stable (about 70) for more than 2 years. Frankly speaking, I do not think we can suggest anything there: the communities will resist any changes until they are ripe for these changes to happen. And when they are right they will design smth themselves, not just accept what we suggest.
Do you think that means we should refrain from making any suggestions on the topic? Will the communities know about the problem before it is too late? It may be useful to 'prime the pump' by trying to think of some solutions for what appear to be very serious problems.
I would say the same thing about "Article Wizard" interfaces: Easier said than done. If such an idea is to be taken seriously by implementers, we need to know exactly what kind of "Wizards" are being proposed, and we need to measure whether they will help or hinder before they are deployed on a wide-spread basis.
But what can we recommend? Electing more admins? This only worls if there are more reasonable candidates. Easing the requirements for an admin? I am all for discussing and recommending smth, I just honestly do not see what we can do.
We can look and see what other large volunteer organizations do to attract and retain the best and brightest as their top line volunteers.
Cash, gifts, real-life community recognition (if they want it, and anonymity if they do not!) are some possibilities. Which of those possibilities do you think has the greatest chance to make the most positive difference?
What are the upsides and downsides to each?
Obviously real-life community recognition.
Cash and gifts in my opinion are more likely to attract people who want cash and not creating encyclopedic content/
Do you suppose we have any admins who would be embarrassed if their employer discovered their hobby? More importantly, how many admins want to remain anonymous because they were involved in a content dispute in which they were threatened? I suspect, given the extent to which both occur in practice, that real-life community recognition would not make as much positive difference as anonymous gifts or payments.
(Regarding Wizards and the like, a range of concrete ideas have been proposed and gained significant support early on. It's mainly (but not all) in Archive #1.)
I would recommend tangible incentives for top-line volunteers at least at the level the Red Cross provides before I would agree to asking people who want to create an article to fill out a form. We tried complicated forms for copyright concerns on Commons, but a lot of people just click around them and choose the license from the big pull-down menu (HTML form INPUT TYPE=SELECT) instead of going through the forms.
I'm not opposed to them, I'm just saying there are a lot more tangible things that need to be done immediately to, for example, reduce the number of contested prods which come back to undeletion requests. Treating admins like respectable volunteer organizations treat their top volunteers is one of those things.
There's a whole thread (several actually by now) on wizards and interfaces, and ways we can use both. You're missing the point on all of these, best look up those threads (on this wiki) and see what we have been thinking as an overall approach. What you're thinking is crude and indeed gets in the way as much as it helps.
In what way do you find those thoughts crude? What are they getting in the way of?
If there are more developed examples of "Wizard" interfaces, please cite them.
"Crude" relates to your point:
- "We tried complicated forms for copyright concerns on Commons, but a lot of people just click around them and choose the license from the big pull-down menu (HTML form INPUT TYPE=SELECT) instead of going through the forms."
I would say that's very crude (ie, agreeing with you), and that kind of approach would not be too productive.
The broad concept that task force members have in mind is very different. I'll look up the cites in a bit (sorry for the delay, not much time free the last day or so), but you can find them on this taskforce's threads and archived discussions. That said, it's a bit of a tangent here.