Wikipedia:Wiki Guides/0000Feb26DiscussionHours

[2011-02-25 18:58:00] =-= YOU are now known as Jalexander
[2011-02-25 18:58:16] =-= Jalexander has changed the topic to ``Next IRC Office Hours: See http://meta.wikimedia.org/wiki/IRC_office_hours - WikiGuide discussion hours 00:00 Feb 26th (Publicly logged)
[2011-02-25 18:59:00] <Jalexander> (changing my name just because possible some of those coming on don't connect me with Jamesofur)
[2011-02-25 18:59:38] <apergos> you have got to be kidding me
[2011-02-25 18:59:46] <Jalexander> why! :P
[2011-02-25 18:59:50] <apergos> I have hung around long enough to make the next round of this? oh noes
[2011-02-25 18:59:55] <Jalexander> lol
[2011-02-25 19:00:09] -->| Chzz (~Chzz@wikipedia/Chzz) has joined #wikimedia-office
[2011-02-25 19:00:11] <Jalexander> yeah aren't you back on the other side of the pond now too?
[2011-02-25 19:00:12] <Peter-C> Jalexander - wait... is it now?
[2011-02-25 19:00:17] <Jalexander> shouldn't you be Asleep!
[2011-02-25 19:00:24] <Chzz> o/
[2011-02-25 19:00:32] <Jalexander> Peter-C: yeees
[2011-02-25 19:00:37] <Jalexander> 00:00 UTC :)
[2011-02-25 19:00:38] <apergos> yes I am.
[2011-02-25 19:00:44] -->| Philippe (~Philippe@adsl-71-132-137-232.dsl.pltn13.pacbell.net) has joined #wikimedia-office
[2011-02-25 19:00:48] <Jalexander> uh oh
[2011-02-25 19:00:52] <Chzz> can someone outline the remit of this meeting, if possible?
[2011-02-25 19:00:59] <Peter-C> The boss is here! Hide the fun!
[2011-02-25 19:01:00] |<-- Philippe has left freenode (Changing host)
[2011-02-25 19:01:00] -->| Philippe (~Philippe@wikimedia/Philippe) has joined #wikimedia-office
[2011-02-25 19:01:02] <apergos> and yes I should. instead I am having a protracted discussion about xml dumps in the tech channel
[2011-02-25 19:01:08] <Philippe> there's fun? who authorized fun?
[2011-02-25 19:01:14] <Philippe> I only authorized booze.
[2011-02-25 19:01:26] <Jalexander> can't authorize booze without fun
[2011-02-25 19:01:31] <Jalexander> legal things
[2011-02-25 19:01:48] <Chzz> yes, lol, etc. - fun is nice. but is this a meet, or just lulz? james...oh fu...r, where is xe?
[2011-02-25 19:02:09] <Chzz> ah; Jalexander. right.
[2011-02-25 19:02:35] * Chzz firmly believes that "anyone who changes nick more than once a year should be shot until very dead" - calls for support
[2011-02-25 19:02:43] <Philippe> +1
[2011-02-25 19:02:45] <Jalexander> Chzz: Relatively open but mostly focusing on questions/thoughts about Wiki Guides, rolling out etc. I'm also interested in some thoughts on a new tool we have (not done but stable and usable)
[2011-02-25 19:03:00] <Chzz> ok; so; I can throw a q
[2011-02-25 19:03:07] <Chzz> ZQ1: What %age of new users give an email addy? I asked this on 03 feb; see https://jira.toolserver.org/browse/TS-935 - without that data, it's hard to judge the effectiveness
[2011-02-25 19:03:43] <Jalexander> given that people who KNOW me have issues connecting Jalexander to Jamesofur I changed it here so people saw me from all those emails :)
[2011-02-25 19:04:14] <Chzz> top tip: choose nick. Stick with it. Is all.
[2011-02-25 19:04:40] <Peter-C> Jamesofur just confuzzles
[2011-02-25 19:04:40] <Jalexander> it's a good question, that unfortunately I don't have a good answer too... but actually that gives me a thought. Toolserver doesn't do the User_email_authenticated yet but it does do some pref stats
[2011-02-25 19:04:45] <Chzz> I am in 35ish channels on 3 IRC servers. Each channel has avg 50 users. If they all change nick, my brains asplode.
[2011-02-25 19:04:52] <Peter-C> "James o fur" - Oh, this guy is a furry!
[2011-02-25 19:04:53] <Jalexander> so we can actually probably look that up, I'll get back to you as soon as I can
[2011-02-25 19:05:01] <Philippe> Chzz... that would be messy.
[2011-02-25 19:05:08] <Chzz> Peter-C I prefer thinking of "James of Ur"
[2011-02-25 19:05:12] <Jalexander> Peter-C: Watch yourself, I speak soft but carry a big stick
[2011-02-25 19:05:17] <Chzz> jeje
[2011-02-25 19:05:21] <Chzz> but srsly
[2011-02-25 19:05:25] <Chzz> my q. stands
[2011-02-25 19:05:36] <Philippe> it's a great question.  :)
[2011-02-25 19:05:57] <Chzz> Philippe it is #1 in a series of over 9000. I am a real cunt for that sort of thing.
[2011-02-25 19:06:10] <Philippe> its gonna be a long meeting
[2011-02-25 19:06:14] <Chzz> :-)
[2011-02-25 19:06:16] <Chzz> TIND
[2011-02-25 19:06:59] <Jalexander> it is a good question, well pends on your definition :( Right now we have Peter who was at the other meeting and Chzz who isn't signed up YET /me stares (but is really interested in the idea obviously)
[2011-02-25 19:07:01] <Philippe> Can I ask a process question, James?
[2011-02-25 19:07:05] <Jalexander> of course
[2011-02-25 19:07:31] <Philippe> Does it make sense to answer one basic question, upfront? "What is this study supposed to do?"
[2011-02-25 19:07:37] <Philippe> I'd be willing to bet there are multiple answers.
[2011-02-25 19:07:41] <Chzz> Jalexander tell me what hoop to jump through to 'sign up' (whatever that is) and I will do it.
[2011-02-25 19:07:59] <Jalexander> Philippe: That probably does make sense
[2011-02-25 19:08:06] <Philippe> Chzz, first you must bring the tear of a Pheonix.
[2011-02-25 19:08:07] <Philippe> Then you can sign up
[2011-02-25 19:08:19] <Jalexander> Chzz: Wikipedia:Wiki Guides. Sign there and send me an email. I even give you a link :)
[2011-02-25 19:08:26] -->| killiondude (~Killer@wikimedia/Killiondude) has joined #wikimedia-office
[2011-02-25 19:08:29] <Philippe> We're into "high barrier to entry" here.
[2011-02-25 19:08:37] <Chzz> will a can of Special Brew suffice as a proxy, Philippe?
[2011-02-25 19:08:41] * Philippe glances at killiondude .... evidently not high enough.
[2011-02-25 19:08:44] <Philippe> sure.
[2011-02-25 19:09:02] <Chzz> omg, killiondude. well, there goes the neigborhood(sic)
[2011-02-25 19:09:22] <killiondude> ohai.
[2011-02-25 19:09:36] <Philippe> So James, what is this study supposed to do?
[2011-02-25 19:09:38] <Jalexander> There are definitely a couple answers to the "what is this supposed to do?" question. I think the first is to try and gage the effectiveness of actively reaching out to new users who just started
[2011-02-25 19:10:02] <Jalexander> trying to see if we can get them hooked and past the first couple edits
[2011-02-25 19:10:42] <Chzz> "point of order, mr. Chairman" - can we address ZQ1 please? yes/no/don't care all being acceptable answers
[2011-02-25 19:11:04] <Jalexander> That's of course relatively easy to start trying to actually do but harder to totally gage HOW it's being effective. So we also want to gather as much data as possible to help us with that and other questions
[2011-02-25 19:11:15] <Jalexander> Chzz: The one regarding how many people give an email?
[2011-02-25 19:12:28] <Jalexander> that's a good question, that unfortunately I don't have a good answer too yet. But you made me think of a thought that might help answer it. Toolserver doesn't do the User_email_authenticated yet but it does do some pref stats
[2011-02-25 19:12:48] <Jalexander> which should allow me to get that. Probably not during the meeting but very shortly after
[2011-02-25 19:13:23] <Chzz> hold
[2011-02-25 19:13:35] <Chzz> -this stat would take 0.1 seconds to get
[2011-02-25 19:13:36] * Philippe plays hold music
[2011-02-25 19:13:40] -->| Orionist (5c60fd52@gateway/web/freenode/ip.92.96.253.82) has joined #wikimedia-office
[2011-02-25 19:13:48] <Chzz> any devs here?
[2011-02-25 19:14:05] <Philippe> ping apergos
[2011-02-25 19:14:07] * Chzz is nothing if not dynamic
[2011-02-25 19:14:32] * Chzz listens to "greensleeves"
[2011-02-25 19:14:40] <Philippe> anyway, let's go on
[2011-02-25 19:14:42] <Philippe> we can come back to this
[2011-02-25 19:14:56] <Jalexander> yeah apergos or someone may be able to help us with that. I just checked and it's not in that table. Definitely come back to it though
[2011-02-25 19:15:14] <apergos> ow
[2011-02-25 19:15:16] <apergos> ow ow ow
[2011-02-25 19:15:24] <Philippe> it didn't hurt that bad
[2011-02-25 19:15:25] <Philippe> drama queen
[2011-02-25 19:15:32] <apergos> it does at 2:15 am
[2011-02-25 19:15:34] <Chzz> apergos heh; kewl; the taser-ping-link works
[2011-02-25 19:15:49] <apergos> I don't have toolserver access
[2011-02-25 19:16:02] <Chzz> apergos q is What %age of new users give an email addy?
[2011-02-25 19:16:02] <Jalexander> I do, that doesn't give me the answer ;)
[2011-02-25 19:16:05] <Philippe> oh, what good are you to us then.
[2011-02-25 19:16:20] <Chzz> apergos I have toolserv. But that field is not visible
[2011-02-25 19:16:37] <Jalexander> we'll probably have to check with Erik or someone. My guess is it's actually already available or we can ask someone to do a quick SQL query for us.
[2011-02-25 19:16:39] <Chzz> apergos https://jira.toolserver.org/browse/TS-935
[2011-02-25 19:16:56] <Jalexander> I'm going to be asking for another query soon so I'll see if I can get a rough one for that as well which will give us a good number.
[2011-02-25 19:17:01] <Jalexander> Only another 3 lines
[2011-02-25 19:17:11] <Jalexander> but moving on with Philippe's question about what we're going for
[2011-02-25 19:17:11] <apergos> I'm no good... no good for you ( Philippe ) :-P
[2011-02-25 19:17:16] <Chzz> Jalexander yeah; and sorry to push the q - but...I suspect the % is so small, it directly conncerns all the rest of the points
[2011-02-25 19:17:17] <Philippe> let's not get distracted by this.  :)
[2011-02-25 19:17:21] <Philippe> (agreed, apergos )
[2011-02-25 19:17:36] <Philippe> Chzz: Regardless, it's the method we have for trial :)
[2011-02-25 19:17:39] <Philippe> Trials are often messy.
[2011-02-25 19:18:02] <Jalexander> aye, if we need to adjust we clearly will. Especially after the trial.
[2011-02-25 19:18:14] <Chzz> Philippe ok; if you take my point - about the relevence, - yeah; sure - happy to press ahead, given we can't get that stat *right now* (or, indeed, in 3 weeks, apparently)
[2011-02-25 19:18:26] <Philippe> So, then why are we here? :P James?
[2011-02-25 19:18:32] <Chzz> ZQ2: STATS - c'mon. We need to know. How many new users (1 contrib) join each day? how many stay >1 week, month, year?
[2011-02-25 19:18:33] <Chzz> how many edit >1 article? No idea? This is easy to find; why don't we know? (I have asked S Walling about this, previously)
[2011-02-25 19:18:36] <jorm> is this an office hour?
[2011-02-25 19:19:20] <Philippe> not really, no. :)
[2011-02-25 19:19:24] <Philippe> but it's a meeting, and you're welcome
[2011-02-25 19:19:36] <jorm> i'm never welcome anywhere.
[2011-02-25 19:19:40] * Chzz highlights: feel free to tell me to STFU if you've got better ideas to discuss
[2011-02-25 19:19:47] <Philippe> I just asked jorm by IM about Chzz's question.
[2011-02-25 19:19:49] <Philippe> He's gonna poke.
[2011-02-25 19:19:57] <Chzz> ta
[2011-02-25 19:20:15] <Jalexander> so, 1st thing trying to gage the effectiveness of actively outreaching to get people to continue editing. 2nd is to try and get data. Both about the users in general (how many are getting blocked etc) and about how we're reaching out
[2011-02-25 19:20:31] <Philippe> gauge*
[2011-02-25 19:20:41] <jorm> The number of users who provide email addresses and then make edits . . .that's not a high value statistic.
[2011-02-25 19:20:48] <Philippe> James, may I add a third thing?
[2011-02-25 19:20:50] <Jalexander> how many are getting back to us? How many are editing? Do the users we don't reach out too (the control) get blocked more often and edit more?
[2011-02-25 19:20:53] <jorm> email address is actually pretty low value.
[2011-02-25 19:21:06] |<-- Jan_eissfeldt has left freenode (Quit: Leaving.)
[2011-02-25 19:21:11] <Jalexander> Philippe: Of course, please
[2011-02-25 19:21:23] <Chzz> ZQ3: Guide selection: who chooses? Like it or not, it is a 'position of trust' likely involving private data. What is the vetting process? < prob mostly @ Jalexander
[2011-02-25 19:21:24] <Philippe> Third thing is to develop and grow the WMF's in house capacity to iterate quickly on a series of tests and trials.
[2011-02-25 19:22:05] <jorm> We have the user stats, though, that you're asking about. And they're pretty . . . not pretty.
[2011-02-25 19:22:11] <Jalexander> and I prefer discussion hour. Yea jorm I expect it's relatively low . I know for example yesterday we had around 2200 users who made an edit. My guess is no more then half of them authenticated an email but it may even be much less.
[2011-02-25 19:22:24] <Chzz> Philippe re. "how many" - that's my point, really; we have no fucking idea. If we'd kinda decide *before* marching ahead what stats would show if we're spending our money wisely...etc
[2011-02-25 19:22:59] <Philippe> Actually not really.  :-) Because this is, like I said, a proof of concept. It's about determining whether interventions make sense....
[2011-02-25 19:23:12] <Chzz> no
[2011-02-25 19:23:15] <Philippe> not particularly to this subset, but do interventions at ALL make sense.
[2011-02-25 19:23:26] <Chzz> just 'coz it is 'proof of concept' does not mean it shouldn't be measurable
[2011-02-25 19:23:31] <Jalexander> aye, that's a very good point. In things like fund-raising those tests did very very well. Trying to find ways to do that on a broader skill (and learning how to do that well) is much better.
[2011-02-25 19:23:43] <Chzz> by trying to measure what is happening, we might make progress
[2011-02-25 19:23:50] <Philippe> Chzz, it absolutely must be measurable. But what I'm saying is, don't pin the program to this one particular factor.
[2011-02-25 19:24:06] <Jalexander> it is measurable. Even failure is a very important answer to get
[2011-02-25 19:24:16] <Philippe> There's nothing saying it will continue to be hinged on this way of finding people.
[2011-02-25 19:24:17] <Philippe> We may find participants another method for a phase 2
[2011-02-25 19:24:19] <jorm> we don't have many ways to measure what's happening. very few, actually.
[2011-02-25 19:24:23] <Philippe> We're testing the intervention here, not the method for choosing the contributors. :)
[2011-02-25 19:24:32] <Chzz> if we fire off random-emails, and get random-responses, and then discuss it in a further (sorry, English, apolos) *bullshit* meeting that concludes nothing...then we waste *our* money
[2011-02-25 19:24:51] <Philippe> Chzz, we can very very easily measure those things.
[2011-02-25 19:24:54] <Chzz> targets, people. key measures; milestones. Show us why it works.
[2011-02-25 19:25:04] <Philippe> We measure whether these metrics grow against the norm.
[2011-02-25 19:25:15] <Philippe> Do the users stay around because we explained why their edit was reverted?
[2011-02-25 19:25:24] <Philippe> Do the users stay around because they know a person to ask a question to?
[2011-02-25 19:25:28] <Philippe> These are things we can measure.
[2011-02-25 19:25:33] <Jalexander> yup exactly
[2011-02-25 19:26:08] <Chzz> Philippe fuck that; that's detailed shit. Just ask, "how many new articles in one day are deleted? how many are shit? how many of 'em never edit again?" - surely, this is *basic* stuff
[2011-02-25 19:26:26] <Chzz> I'm *begging* WMF to give us those stats
[2011-02-25 19:26:27] <Philippe> Ah, see, this is why I wanted to ansewr that basic question at the beginning of the meeting.
[2011-02-25 19:26:28] <Chzz> and it ain't hard
[2011-02-25 19:26:46] <killiondude> that's not what she said.
[2011-02-25 19:26:58] * jorm ^5 killiondude
[2011-02-25 19:27:02] * Chzz inserts a large gnome into killiondude
[2011-02-25 19:27:07] <killiondude> :-D
[2011-02-25 19:27:08] =-= Laura|Away is now known as LauraHale
[2011-02-25 19:27:10] <killiondude> Couldn't resist.
[2011-02-25 19:27:13] <Chzz> heh
[2011-02-25 19:27:22] <Philippe> Chzz, remember the purpose of THIS study: to measure whether particular types of intervention result in editor stick-iness.
[2011-02-25 19:27:33] <Jalexander> Chzz: To answer your question about how the guides are being selected. They are basically self selected. If I have someone who applied and we knew we could not trust them then we would obviously either watch much closer or ask them to step aside but I'm not sure how much use we have to actively cull them. There is no one who has signed up so far for example that I have a concern to a level...
[2011-02-25 19:27:33] <Jalexander> ...I would want to get rid of them.
[2011-02-25 19:28:29] <Jalexander> The private information they have access to is effectively null. The only thing they could get from a name is that the person has an email set up and they could figure that out by going to their user page (even if it doesn't exist) and pressing the 'email this user' button (if it exists)
[2011-02-25 19:28:42] <Chzz> shit, killiondude, we joined at same time; we've been through the wiki-mill; we're part of things now - in different ways. so yeah; we're all 'BTDT FUCK tl;dr AN=DRAMA WMF=fun lulz timewaste" and so forth. And admins smell of cabbage, obv
[2011-02-25 19:29:33] <Chzz> Jalexander your selection process sounds too much like the start of a cabal, for my tastes.
[2011-02-25 19:29:47] * jorm smirks.
[2011-02-25 19:30:10] <Philippe> self-selection is a cabal?
[2011-02-25 19:30:15] <Chzz> Philippe I sincerely believe, to "measure whether particular types of intervention result in editor stick-iness." first requires quite basic info - the stats I mentioned
[2011-02-25 19:30:18] <Jalexander> to be honest I don't know what to tell you Chzz. I think you know I work hard to try and get the answers I can and to do what's best. I may be wrong at times obviously and I think most of the the WMF (and for that matter the wiki) is like that.
[2011-02-25 19:30:18] <Philippe> funny definition :)
[2011-02-25 19:30:36] <Jalexander> yeah I'm not totally sure why it sounds like the start of the cabal when we are basically letting any of them join.
[2011-02-25 19:30:40] <Jalexander> this is not a hand selected group
[2011-02-25 19:30:53] <Jalexander> in fact the VAST majority of people who signed up I have never worked with
[2011-02-25 19:31:00] <Jalexander> and that's very good in my opinion
[2011-02-25 19:31:17] <Jalexander> they are a group of users who I haven't seen active elsewhere
[2011-02-25 19:31:20] <killiondude> I'm not following the conversation enough to give an opinion, Chzz (if that's what you were looking for :-P )
[2011-02-25 19:31:29] <Chzz> Jalexander don't get me wrong here: If it needs saying, then sure: I think James rocks� - but meh; I do too. Doesn't help here
[2011-02-25 19:32:18] <jorm> Since Jan 28 there have been 210,123 users with 1 edit created. Of those, 152,066 gave an email, which is 72%.
[2011-02-25 19:32:33] <Jalexander> hmm that's far higher then I thought. Thank you jorm
[2011-02-25 19:32:38] <Chzz> killiondude no hay problema; any time dude
[2011-02-25 19:32:43] <killiondude> :)
[2011-02-25 19:33:19] <jorm> i'm totally confused why everyone asks for editor stats and then completely ignores everything in the editor trends studies.
[2011-02-25 19:33:32] <jorm> like, for some reason, they think that there are other, more secret numbers out there.
[2011-02-25 19:34:05] |<-- Chzz has left freenode (Disconnected by services)
[2011-02-25 19:34:12] -->| killiondude_ (~Killer@wikimedia/Killiondude) has joined #wikimedia-office
[2011-02-25 19:34:33] -->| Chzz_ (~Chzz@wikipedia/Chzz) has joined #wikimedia-office
[2011-02-25 19:34:41] <Philippe> jorm, please dont' mention the secret numbers. You know they're secret. :)
[2011-02-25 19:35:04] <jorm> The ones published on stats.wikimedia.org? or the ones in the editor trends study? or what?
[2011-02-25 19:35:16] <Philippe> <grin> yes.
[2011-02-25 19:35:17] <killiondude_> Philippe needs to use sarcasm tags.
[2011-02-25 19:35:25] <killiondude_> #sarcasm
[2011-02-25 19:35:35] * Jalexander waves to Orionist . I saw you pop on earlier while we were talking and didn't say hi
[2011-02-25 19:36:06] <Orionist> Hi Jamesw. hi everyone!
[2011-02-25 19:36:18] === Orionist <5c60fd52@gateway/web/freenode/ip.92.96.253.82> ``92.96.253.82 - http://webchat.freenode.net
[2011-02-25 19:36:18] === Orionist: member of #wikimedia-office
[2011-02-25 19:36:19] === Orionist: attached to bear.freenode.net ``London, England
[2011-02-25 19:36:19] --- End of WHOIS information for Orionist.
[2011-02-25 19:36:29] <jorm> personally, i think that measuring what post-account creation guide works best is just like asking "which book on dealing with assholes works for you?"
[2011-02-25 19:36:44] |<-- killiondude has left freenode (Disconnected by services)
[2011-02-25 19:36:45] <jorm> it only measures the people willing to deal with assholes.
[2011-02-25 19:37:12] <Philippe> see, the beauty of this is, it's not a zero sum game :) We can put resources both places.
[2011-02-25 19:37:22] =-= killiondude_ is now known as killiondude
[2011-02-25 19:37:30] <Jalexander> jorm: That's possible. Though I feel that with someone nice you know when you are dealing with the assholes and greatly increase those odds. I know it's helped me on multiple occasions
[2011-02-25 19:37:48] <jorm> i have a half-finished thing about adding certain technology to the user account creation page to track people for, like, ever, for these kinds of studies.
[2011-02-25 19:37:50] <Jalexander> that's clearly an anecdotal and biased response, but I'd like to see if it's right or not
[2011-02-25 19:38:02] * Jalexander approves
[2011-02-25 19:38:19] <Philippe> Creeper.
[2011-02-25 19:38:20] <Chzz_> I am (pleasantly) astonished to learn that 70%+ of new users give an email addy
[2011-02-25 19:38:49] <Philippe> 70% of new users who make an edit*
[2011-02-25 19:38:53] <Chzz_> I thought it'd be way smaller
[2011-02-25 19:39:00] <Chzz_> yes, Philippe, indeed
[2011-02-25 19:39:11] <Chzz_> OK
[2011-02-25 19:39:14] <jorm> i don't really find that surprising.
[2011-02-25 19:39:29] <jorm> it's a single field, and our account creation screen sucks.
[2011-02-25 19:39:31] <Chzz_> I do, from NPP experience
[2011-02-25 19:39:50] <Philippe> OK, we're 40 minutes into this thing: James, what do we need to accomplish?
[2011-02-25 19:39:51] <jorm> people see all the warnings, don't read them, and assume that everything is required.
[2011-02-25 19:39:56] <Chzz_> enwiki NPP I'd guess that...what , killiondude? 75% is "utter shite"?
[2011-02-25 19:40:04] <Chzz_> maybe 60%
[2011-02-25 19:40:06] <Jalexander> I know we were talking about the idea in the last meeting of using links that would allow us to see if someone clicked edit during that next session. Which sounds like a prelimiary idea to what you're talking about jorm.
[2011-02-25 19:40:10] <Philippe> (and if the answer is "talk through these questions, i'm fine with that)
[2011-02-25 19:40:29] <killiondude> I think closer to 50%
[2011-02-25 19:40:29] <jorm> i have technology designed that will do what you want but in a better way.
[2011-02-25 19:40:38] <Philippe> Jalexander: you're talking about click tracking?
[2011-02-25 19:40:43] <killiondude> But that's still a (frighteningly) large number.
[2011-02-25 19:40:59] <Jalexander> Philippe: That may well be it because we have a fairly small group. The question would first be if Orionist or someone else has any questions and has been kindly waiting while we rant
[2011-02-25 19:41:09] <Jalexander> Philippe: I am yes
[2011-02-25 19:41:15] === Lcawte <lcawte@Wikimedia/Lcawte> ``Lewis Cawte
[2011-02-25 19:41:15] === Lcawte: member of #wikimedia-office, #wikipedia-en, and #wikimedia
[2011-02-25 19:41:15] === Lcawte: attached to jordan.freenode.net ``Evry, FR
[2011-02-25 19:41:15] === Lcawte is logged in as Lcawte
[2011-02-25 19:41:15] --- End of WHOIS information for Lcawte.
[2011-02-25 19:41:18] <Philippe> k
[2011-02-25 19:41:19] <Chzz_> killiondude that's the stats we so desperately need - that's what I'm asking from WMG - take a day, look at all the new pages, and tell us- the % CSD'd, %valid, etc
[2011-02-25 19:41:36] <Jalexander> I used that word this morning, not sure why I didn't today :)
[2011-02-25 19:41:49] <killiondude> I don't know if they'd like what they see, chzz. :)
[2011-02-25 19:41:52] <killiondude> not good PR etc.
[2011-02-25 19:42:21] <Philippe> I dont think we really worry about PR when collecting data ;)
[2011-02-25 19:42:30] <jorm> our data is brutally bad.
[2011-02-25 19:42:38] <Chzz_> killiondude sure; it'd be pretty grim; but hey, from that, it'd be easy to move forwards
[2011-02-25 19:42:40] <Jalexander> to be honest I think if you asked many users and staff members they would give numbers that were less flattering then the real numbers ended up being
[2011-02-25 19:42:51] <Jalexander> for those %s
[2011-02-25 19:42:54] <Jalexander> and yes jorm, it is
[2011-02-25 19:43:40] <killiondude> I would ask "Then what is being used to combat all the crappy content added to Wikipedia?" but the answer would probably be "It's the community's decision what to do." or something.
[2011-02-25 19:44:00] <Chzz_> heh, yeah
[2011-02-25 19:44:10] <Philippe> killiondude: read the log of yesterday's office hours :)
[2011-02-25 19:44:12] <Chzz_> which is why I'm pressing for a baseline
[2011-02-25 19:44:19] <Chzz_> ZQ2: STATS - c'mon. We need to know. How many new users (1 contrib) join each day? how many stay >1 week, month, year?
[2011-02-25 19:44:19] <Chzz_> how many edit >1 article? No idea? This is easy to find; why don't we know? (I have asked S Walling about this, previously)
[2011-02-25 19:44:38] <Chzz_> to gather that would take...what? 10-20 hours man-time?
[2011-02-25 19:44:45] <Jalexander> I can tell you yesterday it was about 2200 who made more then an edit
[2011-02-25 19:45:07] <killiondude> I think that's Shirley's shtick (re: the wmf wanting participation no matter what kind of content is added), iirc
[2011-02-25 19:45:09] <apergos> we have that stuff inthe editor trends study right?
[2011-02-25 19:45:14] <Philippe> Chzz_: we are hiring a data analyst who can pull exactly those types of numbers ::)
[2011-02-25 19:45:20] <Philippe> on an ongoing basis
[2011-02-25 19:45:51] <killiondude> Nice.
[2011-02-25 19:46:08] <Chzz_> Philippe great; any chance of it happening before the next ice-age <forgive sarcasm; English, etc>
[2011-02-25 19:46:11] <jorm> we have that stuff in the editor trends study, correct, apergos.
[2011-02-25 19:46:19] <Philippe> pretty good chance.
[2011-02-25 19:46:25] <jorm> which is where i'm all confused when people keep asking for it.
[2011-02-25 19:46:32] <killiondude> Chzz_: we have a bugmeister!
[2011-02-25 19:46:48] <Chzz_> hm
[2011-02-25 19:46:55] <killiondude> (aka bugshyster)
[2011-02-25 19:47:30] <Jalexander> jorm: Where's the study living atm? (where can Chzz_ get it)
[2011-02-25 19:47:31] <Philippe> http://wikimediafoundation.org/wiki/Job_openings/Data_Analytics_Engineer
[2011-02-25 19:47:34] <Philippe> (currently posted)
[2011-02-25 19:47:44] <jorm> strategy
[2011-02-25 19:48:13] <killiondude> I think the "Major Gifts Associate" would be a fun job.
[2011-02-25 19:48:43] <Chzz_> I'm still...sorry, I just doubt this; jorm is saying that, at the "create account" thingy, 70%+ give an email address? Sorry to keep asking; I just find it stunning, and need to ask one more time
[2011-02-25 19:48:58] <Chzz_> (ok, of people who make >1 contrib)
[2011-02-25 19:49:00] <Jalexander> killiondude: Perhaps, but also remember the pressure when you're dealing with large amounts of money if you make one wrong move with someone
[2011-02-25 19:49:01] * jorm gives up.
[2011-02-25 19:49:14] <Jalexander> yes Chzz_
[2011-02-25 19:49:18] <Chzz_> ok
[2011-02-25 19:49:29] <Jalexander> So Orionist asked me in PM about when we plan on starting
[2011-02-25 19:49:38] * Philippe is digging and does not find the data in the editor trends study page on strategy
[2011-02-25 19:49:41] <killiondude> Right. You just have to have good people skills (among other qualities).
[2011-02-25 19:49:41] <Philippe> but i'm also pretty dumb.
[2011-02-25 19:49:42] <Chzz_> well, that's great, Jalexander. I am suprised. In a good way.
[2011-02-25 19:49:43] <Jalexander> I wanted to answer that here as well so that everyone else could see that
[2011-02-25 19:50:53] <Jalexander> I think we're on target to be sending out a first (small) round of assignments late tonight (US time ). There is a loose end or two that I want to make sure is tied up so it's possible that it will be pushed until this weekend but "very soon"
[2011-02-25 19:51:06] <Jalexander> (trademarked)
[2011-02-25 19:51:14] <Chzz_> *from* who? Jalexander
[2011-02-25 19:51:26] <Jalexander> from would be me
[2011-02-25 19:51:34] <killiondude> ™
[2011-02-25 19:51:38] <Chzz_> ☝this
[2011-02-25 19:51:38] <Jalexander> yes that
[2011-02-25 19:51:55] <Chzz_> meh; ok; do your thing
[2011-02-25 19:52:10] |<-- stwalkerster has left freenode (Remote host closed the connection)
[2011-02-25 19:52:10] <Jalexander> As Philippe said this is a trial, things are likely going to change and I think that's ok
[2011-02-25 19:52:12] <Chzz_> but, if you wanna do your thing, why are you asking us? go for it
[2011-02-25 19:52:13] <Philippe> Does everyone know what to DO with the assignments?
[2011-02-25 19:52:15] -->| stwalkerster (stwalkerst@pdpc/supporter/student/stwalkerster) has joined #wikimedia-office
[2011-02-25 19:52:17] |<-- Beria has left freenode (Ping timeout: 264 seconds)
[2011-02-25 19:52:26] <killiondude> be bold
[2011-02-25 19:52:42] <Chzz_> be bald
[2011-02-25 19:52:56] -->| Sumsum2010 (6294a241@gateway/web/freenode/ip.98.148.162.65) has joined #wikimedia-office
[2011-02-25 19:53:05] * killiondude shudders.
[2011-02-25 19:53:10] <Jalexander> Chzz_: Because it doesn't mean things can't change (and usually do when people discuss it) and it allows us to know what to look at so that even if we go ahead with something "wrong" we're ready to change it.
[2011-02-25 19:53:17] <Jalexander> hey Sumsum2010, I'm glad you could join us!
[2011-02-25 19:53:18] <Chzz_> but srsly - if you wanna email n new wikipedians, and ask 'em "why" or whatever - ok, cool. why not? but, why ask us?
[2011-02-25 19:53:37] <Sumsum2010> Sorry i'm late, did I miss anything
[2011-02-25 19:53:39] <Philippe> Chzz_: wait.
[2011-02-25 19:53:46] <Philippe> Let's talk about the purpose of this study again.
[2011-02-25 19:53:53] <Philippe> We're emailing Wikipedians, yes, but not to ask why they leave.
[2011-02-25 19:54:03] <Philippe> It's to provide an intervention at critical inflection points.
[2011-02-25 19:54:10] <Philippe> For instance, when their first edit is reverted.
[2011-02-25 19:54:45] <Jalexander> Philippe: I know we've been discussing it some. We have some great email examples to start off with and I want to encourage everyone to add their own personal twists (and I want to see them! There are bound to be some great ideas)
[2011-02-25 19:54:56] <Philippe> OK, i just watned to be sure:)
[2011-02-25 19:55:02] <Orionist> Jalexander: Do you have some kind of "infrastucture" ready, like some sample messages etc. Maybe some templates or userboxes for the editors wee're going to guide, make them feel a part of a community and such? or is this going to give each guide the freedom to choose whatever they feel right?
[2011-02-25 19:55:25] <Chzz_> Philippe I *know* why their first edit was reverted. Don't need to ask 'em. I can tell, from...their *first edit*
[2011-02-25 19:55:31] <Philippe> no no no....
[2011-02-25 19:55:34] <Philippe> THEY don't know why
[2011-02-25 19:55:38] <Chzz_> mhm
[2011-02-25 19:55:49] <Chzz_> so tell 'em on their talk page
[2011-02-25 19:55:53] <Orionist> Ah! I really should work on my typing speed!
[2011-02-25 19:55:53] <Chzz_> fix templates
[2011-02-25 19:55:55] <Jalexander> we're going to start off with a very small group of 10 so that you can take a bit of time to look at each one and use some personal touches
[2011-02-25 19:56:05] <Jalexander> Orionist: It's both a blessing and a curse :)
[2011-02-25 19:56:07] <Chzz_> :-S
[2011-02-25 19:56:13] <Philippe> Chzz, it's push versus pull messaging.
[2011-02-25 19:56:21] <Sumsum2010> Jalexander: do we have lists of uses compiled yet of are they still being made?
[2011-02-25 19:57:13] <Jalexander> We do have some infrastructure, some good same messages on the talk page and a couple more I'm going to put out. I think we want to provide some freedom though so that people can share what they're doing, what works and is different etc.
[2011-02-25 19:57:16] <Chzz_> can't see a point here; I can do all this shit myself. And I have. And, consequently, I *know* why we don't get new editors. If you wanna try work it out, sure, go for it. Have fun. When you reach same conclusion as me, in a year or ten, let me know.
[2011-02-25 19:57:26] |<-- killiondude has left freenode (Quit: 42)
[2011-02-25 19:57:47] <Jalexander> Chzz_: What do you think we are going to do. I want to make sure we're on the same page
[2011-02-25 19:58:04] <Jalexander> Sumsum2010: Not yet, I hope to have those out late tonight, and if not that this weekend at the latest.
[2011-02-25 19:58:23] <Chzz_> Jalexander emailing new users with a polite, friendly "hi" is a good idea. The rest is not
[2011-02-25 19:58:29] <Sumsum2010> ok, so check my email
[2011-02-25 19:58:54] <Jalexander> Sumsum2010: Aye, though I wouldn't wait up all night for it if you're in the US :)
[2011-02-25 19:59:06] <Philippe> Chzz_: What are you referring to when you say "the rest?"
[2011-02-25 19:59:07] <Jalexander> Chzz_: What do you think the rest entails at the moment.
[2011-02-25 19:59:23] <Philippe> I'm trying to make sure we're clearly communicating what we're trying to do.
[2011-02-25 20:00:34] <Sumsum2010> J: ok tat'll work
[2011-02-25 20:01:04] <Chzz_> Philippe the concept of a 'project' to look at things holistically, is what I mean by "the rest"- whilst it is a laudable goal, the goalposts are not defined - and that problem, of not defining goalposts, seems endemic to WMF.
[2011-02-25 20:01:37] <Chzz_> also Jalexander ^
[2011-02-25 20:02:04] <Philippe> OK, but let's narrow this down. In this study, what is it that you think we're doing? I'm not trying to badger, I really want to be sure that we're not working at cross purposes:)
[2011-02-25 20:03:09] <Chzz_> C'mon folks, this is easy: #1 how many new editors stay more than a month? #2 (ideas; solutions. email them in person? sure, ok, try that) measure it. #3 how many new editors *now* stay more than a month?
[2011-02-25 20:03:33] <Chzz_> did it work? should we do it more? is it worth the $$$ and timetimetime?
[2011-02-25 20:03:42] <Philippe> That's so not what we're doing here.
[2011-02-25 20:04:07] <Philippe> We're working from a hypothesis of something that experientially we believe will work. We're trying to provie that it does or does not.
[2011-02-25 20:04:10] <jorm> C'mon, philippe! it's easy.
[2011-02-25 20:04:20] <Chzz_> ok, Philippe; you have the floor; tell us what you wanna do, and why it is worth investment
[2011-02-25 20:04:28] <Philippe> OK, thanks :)
[2011-02-25 20:04:45] <Philippe> First of all: investment... James is cheap. literally, we pay him in peanuts. So this is a pretty low cost study. :)
[2011-02-25 20:05:00] <apergos> poor guy
[2011-02-25 20:05:05] <apergos> I hope you like peanuts
[2011-02-25 20:05:26] <Philippe> But what we're doing here is starting from the hypothesis that measured caluculated interventions at times when we're likely to "lose" people will result in long term survivability of those people.
[2011-02-25 20:05:33] <Philippe> So we're testing that hypothesis with this study.
[2011-02-25 20:05:54] <Philippe> We're doing an intervention when we think we might be at an inflection point, and then also when we determine there's a need.
[2011-02-25 20:06:00] <Chzz_> <this might be lulz BUT, we have other channels for lulz. Can we stay on-topic please. And we can call James a thingy elsewhere>
[2011-02-25 20:06:25] <Philippe> If, at the end, a greater percentage of these users stick around than the norm, we repeat the study to be sure it's not an anomaly, try the various control groups, and see if we've found something that works.
[2011-02-25 20:06:36] <Philippe> If we have, then we see if there's a way to scale it with software tools.
[2011-02-25 20:06:56] <Chzz_> Philippe so, can I put a proposal to you? will take me 4 mins...
[2011-02-25 20:07:19] <Orionist> If you allow me to chime in here...
[2011-02-25 20:07:25] <Philippe> Orionist: of course.
[2011-02-25 20:07:26] <Jalexander> Please do Orionist
[2011-02-25 20:07:40] <Philippe> Chzz, you're welcome to make a proposal, but it should be to James, and not to me. I just swoop in and use big words.
[2011-02-25 20:07:42] <Philippe> He runs the study.
[2011-02-25 20:07:56] <Orionist> What group of editors are we going to target, edit count wise
[2011-02-25 20:08:13] <Philippe> Orionist: This is editors with at least one edit.
[2011-02-25 20:08:22] <Philippe> So it's the very smallest registered user group
[2011-02-25 20:08:29] <Philippe> (smallest by number of edits)
[2011-02-25 20:08:57] * Chzz_ <aside> Jalexander can't use big words; spells 'em wrong /me goes back to notepad
[2011-02-25 20:09:16] <Jalexander> Not true, I use them still I just spell them wrong
[2011-02-25 20:09:20] <Orionist> Isn't it better to target, say, editors with 50+ edits?
[2011-02-25 20:09:42] <Jalexander> Orionist: well one of the big things we want to try and target is if we can pull those people across the barrier
[2011-02-25 20:09:47] <Philippe> Orionist: the idea of this is to see if we can increase stickiness in this most vulnerable population.
[2011-02-25 20:09:52] <Jalexander> we have a LOT of people who only make like 1-5 edits
[2011-02-25 20:09:54] <Philippe> This is where our biggest drop off is
[2011-02-25 20:10:00] <Orionist> Some editors maybe just signing up to make a couple of edits or upload a photo
[2011-02-25 20:10:17] <Philippe> Perhaps so. But we know from past studies that many are run off for other reasons.
[2011-02-25 20:10:23] <Orionist> editors with about 50 edits are probably prepared to commit
[2011-02-25 20:10:42] <Philippe> If they're prepared to commit, they don't really need us to help them commit. :)
[2011-02-25 20:10:49] <Orionist> But they may not be aware of every aspect of the pedia
[2011-02-25 20:11:01] <Philippe> Absolutely. But tht's a different problem from the one we're trying to solve.
[2011-02-25 20:11:03] <Chzz_> Philippe A) Look at "a day on English Wikipieda" - all new articles. Pick a day from 1 month ago. Show us how many are CSD'd and under what cats; what % are shite? what % are 'fixable'?
[2011-02-25 20:11:04] <Chzz_> B) for one month, choose 'n' of the fixable. For n/2 (let's say, odd numbers) email them. Try helping.
[2011-02-25 20:11:04] <Chzz_> C) evaluate effectiveness Compare to control
[2011-02-25 20:11:05] <Chzz_> ???
[2011-02-25 20:11:05] <Chzz_> profit
[2011-02-25 20:11:25] <Philippe> Chzz_: What's the problem that you're trying to solve there?
[2011-02-25 20:11:34] <Chzz_> editor retention
[2011-02-25 20:11:51] <Chzz_> *~new editor retention
[2011-02-25 20:12:04] <Philippe> OK, so your hypothesis is that the reason we lose new editors is because new articles are deleted?
[2011-02-25 20:12:59] <Philippe> We know that that's true for some percentage of them. But we know that for a greater percentage (which I do not have in front of me) that they leave for reasons unrelated to new articles... in fact, they never tried to start one.
[2011-02-25 20:12:59] <Jalexander> aye, I think what Philippe said is exactly the question. I would really like to make sure those 50+ editors stay too (and you're right they probably are a bit more ready to commit) but I think it's a different dynamic and likely a different study that could definitely take some of the same ideas.
[2011-02-25 20:13:02] <jayvdb> that, coupled with the way we treat them as they try to help
[2011-02-25 20:13:09] <Jalexander> (responding to Orionist )
[2011-02-25 20:13:09] <apergos> suppose you choose new editors who didn't contribute something fixable. maybe they will be just fine as editors if they survive past the deletion (I of course know some cases where this happened)
[2011-02-25 20:13:34] <Philippe> Remember, not all editors create articles :)
[2011-02-25 20:13:37] <Philippe> Most of them don't.
[2011-02-25 20:13:48] <apergos> ( that was to Chzz_ )
[2011-02-25 20:13:50] <Philippe> So, Chzz, I think there's value to what you suggest.
[2011-02-25 20:13:56] <Jalexander> I know of at least 2 editors who's first edit was spam... (they are admins now)
[2011-02-25 20:13:58] <Philippe> But I think that it doesn't get to the heart of the matter,
[2011-02-25 20:14:14] <Philippe> which is that we know from various measurements that people leave because they are confused, frustrated, and don't know how to get help.
[2011-02-25 20:14:20] <Chzz_> sec...sorry...finding. I'm slow
[2011-02-25 20:15:09] <Orionist> Jalexander: My point is, editors with 50 edits may make better quality editors, maybe more effective than trying to convert infrequent editors
[2011-02-25 20:15:38] |<-- Sumsum2010 has left freenode (Ping timeout: 272 seconds)
[2011-02-25 20:15:41] <jayvdb> here is yesterdays example: https://secure.wikimedia.org/wikipedia/en/w/index.php?title=Wikipedia:Administrators%27_noticeboard/Incidents&oldid=415818634#Stone_and_Stone
[2011-02-25 20:16:10] <jayvdb> the new contributor is still with us, so maybe this treatment doesnt hinder them
[2011-02-25 20:16:18] <Philippe> Orionist, that's a good hypothesis, but it doesn't particularly solve the question at hand, which is "how do we prevent the people in THIS set from leaving?"
[2011-02-25 20:16:22] <Jalexander> Orionist: It may be, I honestly don't know. The angle at this stage is a bit different because of the quantity
[2011-02-25 20:16:26] <jayvdb> rough em up and see who stays
[2011-02-25 20:16:47] <Jalexander> the amount of users that make 1-5 (and 1-10) edits and drop of is ENORMOUS the amount who make 50 is tiny
[2011-02-25 20:16:53] <Chzz_> .me oh fucksocks; my logs are too huge; sorry still looking...
[2011-02-25 20:17:28] <Chzz_> Jalexander 0-1 is *way* bigger
[2011-02-25 20:17:56] <Jalexander> and so one of the ideas for this particular project is "ok, well if we can pull 50 of those 100's of thousands to the 50+ category we've made an enormous % change on that 50+ category
[2011-02-25 20:18:31] <Orionist> I see
[2011-02-25 20:18:33] <Philippe> So, for instance, we know that rouhgly 40% of survey respondents said they left beause editors were difficult to work with.
[2011-02-25 20:18:37] <Philippe> That's an intervention that we can help with.
[2011-02-25 20:18:54] <Philippe> Only 13% left because writing an article and getting it to stick was too hard.
[2011-02-25 20:19:02] <Chzz_> Jalexander remember it is about people. One new user who is 'good' is worth 1000 mono's
[2011-02-25 20:19:09] <Philippe> Chzz.... c'mon
[2011-02-25 20:19:10] <Philippe> let's not do that
[2011-02-25 20:19:29] <Jalexander> 1. let's not target individuals 2. I disagree with that statement as a whole
[2011-02-25 20:20:06] * Chzz_ doffs hat; accepts that is was not appropriate to individualize
[2011-02-25 20:20:09] <Philippe> Only 12% of the folks who answered the survey left because the rules around article creation were too complex.
[2011-02-25 20:20:34] <Philippe> So the idea here is that rather than focusing on teh 12 and 13% wins... we'll focus on the 40% reason... by providing targeted intervention.
[2011-02-25 20:20:45] <apergos> see this is the sort of logic that lets people say "one experienced editor's time is much more valuable than trying to make him convert a bunch of new editors who may none of them turn out to be better than mediocre"
[2011-02-25 20:21:23] <apergos> 40% sounds pretty good to me
[2011-02-25 20:21:24] <jayvdb> Philippe: how was the survey conducted
[2011-02-25 20:21:25] <Jalexander> it's certainly an argument I've heard many times and one that I don't agree with
[2011-02-25 20:21:39] <Philippe> jayvdb: this was done by email, with follow up telephone calls.
[2011-02-25 20:22:01] <jayvdb> selection method was kosher ?
[2011-02-25 20:22:06] <Philippe> It was.
[2011-02-25 20:22:12] <Chzz_> Philippe Jalexander it isn't just about numbers though; it's about quality. Increatingly so. I personally have 'helped' over 9000 new users; I think about 10 of 'em are truly active. However, *fuck* they're a good ten.
[2011-02-25 20:22:14] <Philippe> Randomized selection, yada yada
[2011-02-25 20:22:24] <jayvdb> Philippe, ta
[2011-02-25 20:22:38] <Philippe> jayvdb: would you like the link to the survey?
[2011-02-25 20:23:25] <Philippe> But, Chzz.... in the 30,000 that leave in the time it takes you to get the 10... there might be 100 good ones.
[2011-02-25 20:23:26] <jayvdb> nah. i was checking that the data you were giving wasnt from some 'wiki survey' done on outreach or similar
[2011-02-25 20:23:28] <Philippe> That's what we don't know.
[2011-02-25 20:23:32] <Philippe> jayvdb: nope.
[2011-02-25 20:23:42] <Chzz_> those 10 people - and I can paste their names, if you want - do more than 10,000 other new users
[2011-02-25 20:24:29] <Jalexander> Chzz_: It isn't all about numbers, obviously not. but numbers certainly can come into play. I know I've talked about my 80/10/10 philosophy (which I know you think the %s need adjusting, which may be true)
[2011-02-25 20:24:40] <Philippe> But,Chzz_ ...you're starting by being forced to prove a negative. We don't know what the others could have done, because we ran them offf.
[2011-02-25 20:25:01] <Chzz_> Philippe yes, you are quite right; which is why targetting that tiny %age is so critical; we're wasting money on useless fuckwits
[2011-02-25 20:25:13] <Philippe> No.
[2011-02-25 20:25:14] <Philippe> We're not.
[2011-02-25 20:25:20] <Chzz_> yes
[2011-02-25 20:25:21] <Jalexander> but for the sake of argument. Let's use you're numbers 9000 new users, 10 of them have become "Highly active users" who have in turn probably helped more new users.
[2011-02-25 20:25:22] <Chzz_> we are
[2011-02-25 20:25:24] <Chzz_> :p
[2011-02-25 20:25:26] <jayvdb> I think you're all wasting money on Wikipedia
[2011-02-25 20:25:27] <Philippe> We're running off highly educated people, who have trouble with our community.
[2011-02-25 20:25:34] <Philippe> This is a chance to fix it.
[2011-02-25 20:25:46] <Philippe> We're running off people with expertise in every area under the sun.
[2011-02-25 20:25:47] <jayvdb> attract Wikisource contributors and you will have very good Wikipedia editors
[2011-02-25 20:25:56] <Chzz_> Is this the right room for an argument?  :)
[2011-02-25 20:25:58] <Philippe> jayvdb: there's a certain logic to that.
[2011-02-25 20:25:58] <Jalexander> those 8990 users who are not "highly active users" are in no way not valuable
[2011-02-25 20:26:31] <Chzz_> Jalexander ok, math: what is that as a %age of all users?
[2011-02-25 20:26:35] <jayvdb> is there a wrong room for a vigourous discussion about this? ;-)
[2011-02-25 20:26:49] <Jalexander> even if only 1% of the encyclopedia users do 90% of the work. If we cut off about 80-90% of them we would fail.
[2011-02-25 20:26:53] <Philippe> 8990 as a percentage of the active user base? About 10%.
[2011-02-25 20:27:04] <Chzz_> wat?
[2011-02-25 20:27:19] <Philippe> Our active editor base is around 80,000.
[2011-02-25 20:27:41] <Jalexander> there is a much smaller % who do a huge amount of the work, that's clear. But there is also a huge % of people who do valuable work and just like we could not succeed without that 1% we could not succeed without the 80%
[2011-02-25 20:27:43] <apergos> more than 10% then
[2011-02-25 20:27:52] <Jalexander> we could probably cut 10-20% dead weight, I'll give you that
[2011-02-25 20:27:56] <Jalexander> (and always would)
[2011-02-25 20:28:42] <Jalexander> it is a red herring to say "well we can ignore the 10-100 edit group because all we need is the 50k edit group" or similar
[2011-02-25 20:28:43] <Orionist> Did you try to look for similar numbers from other sites that have user-generated content? Editor retention/activity? They might prove useful for comparison
[2011-02-25 20:29:00] <Orionist> what do they have and we don't?
[2011-02-25 20:29:22] <jayvdb> wikiHow has some very good analysis on their community
[2011-02-25 20:29:24] <apergos> the assumption that people who have difficult encounters with editors are part of that dead wood by virtue of their having had those encounters... is problematic to me. some folks don't like the existing culture, it doesn't mean they would be bad contributors
[2011-02-25 20:29:26] <Chzz_> heh; go set up one, Orionist; sheesh. there *is* no comparison to enwiki. Thank FSM.
[2011-02-25 20:29:33] <Philippe> Orionist: I wish we had better numbers for that, and it's on my list of things to look at. But in almost all cases, the answer starts with "they identified the points of vulnerability to losing people and shielded from it."
[2011-02-25 20:29:37] <Jalexander> Philippe may have a better answer to that question then I do Orionist . I haven't seen great numbers but they may have done some for the Strategy sessions.
[2011-02-25 20:30:02] <Jalexander> I did notice an interesting stat in the fundraising survey we did earlier this year
[2011-02-25 20:30:29] <Jalexander> where almost exactly the same % of donors (obviously different pool) who said they edited (about 1/3) said they did other user driven activities
[2011-02-25 20:30:33] <Chzz_> Philippe Jalexander u can contact me. this ain't constructive. I'd rather go edit an article. I'm sure you understand. kthxbai
[2011-02-25 20:30:35] <Jalexander> such as reviews
[2011-02-25 20:30:38] <--| Chzz_ has left #wikimedia-office
[2011-02-25 20:31:00] <Philippe> eyeroll
[2011-02-25 20:31:13] <Philippe> Jalexander: I had forgotten that. You're right
[2011-02-25 20:32:04] <apergos> I wonder why people *want* to exclude folks that don't deal well with a hostile editing atmosphere :-(
[2011-02-25 20:32:19] <apergos> (sorry. just that's how it looks at the moment)
[2011-02-25 20:32:49] <Philippe> Well, this is something I've seen a lot: people don't know or believe the statistics.
[2011-02-25 20:32:57] <Philippe> That's why I want to do a much more aggresive job of putting them in front of people.
[2011-02-25 20:32:59] <apergos> hmm right
[2011-02-25 20:33:13] <Jalexander> even with a hostile editing atmosphere I think a community can help it. When you know someone who does appreciate your work you are far less likely to storm off
[2011-02-25 20:33:25] <jayvdb> that is another reason for making wikisource the first wiki experience for newcomers - it isnt a hostile environment
[2011-02-25 20:33:28] <apergos> it's certainly worth trying
[2011-02-25 20:33:58] <jayvdb> its hard to fight about fixing spelling corrections ;-)
[2011-02-25 20:34:06] <apergos> you think? and yet
[2011-02-25 20:34:08] <Jalexander> when you come in, make yoru first edit, get slapped and don't hear anything but "bad!" then they are far more likely (in my opinion) to leave then if they also heard a "hey, sorry this can get tough but I'm glad you're here"
[2011-02-25 20:34:18] <apergos> on wikt we can duel to the death over spelling corrections! :-D
[2011-02-25 20:34:29] -->| mdale (~mdale@unaffiliated/mdale) has joined #wikimedia-office
[2011-02-25 20:34:30] <Jalexander> jayvdb, unless you're changing it to American/British English
[2011-02-25 20:34:31] <jayvdb> apergos ;-)
[2011-02-25 20:34:47] * Philippe likes wikisource a lot
[2011-02-25 20:34:56] <jayvdb> Jalexander, apergos: wikisource has a correct version for every edit: the one that was printed.
[2011-02-25 20:34:58] <Philippe> The Executive Orders section there makes me very happy.
[2011-02-25 20:35:48] <apergos> so what numbers need to be publicized?
[2011-02-25 20:36:06] <Philippe> apergos: I think what we need to do is find influencers, and get them in a room together, and go through the surveys.
[2011-02-25 20:36:17] <jayvdb> well, American/British could be fought over the transcription of spoken texts.
[2011-02-25 20:36:35] <apergos> did they get into the signpost?
[2011-02-25 20:36:49] <Philippe> Yes, but that still relies on people to read it
[2011-02-25 20:36:51] <Philippe> and understand.
[2011-02-25 20:36:54] <Philippe> I think we need to presentit.
[2011-02-25 20:36:59] <Jalexander> and be interested from the start
[2011-02-25 20:37:15] <Jalexander> they have to be interested enough to 1. click through and 2. read beyond the first paragraph or so
[2011-02-25 20:37:18] <Jalexander> 2 significant steps
[2011-02-25 20:37:42] <Jalexander> (and depending on where they started from perhaps a couple more clicks)
[2011-02-25 20:37:45] <apergos> you would want a couple of eye popping numbers to be right there in the lead
[2011-02-25 20:37:52] <Jalexander> yeah I think you're right
[2011-02-25 20:37:54] <apergos> so that people will have the desire to click through
[2011-02-25 20:38:09] <Philippe> It wouldn't hurt anything.
[2011-02-25 20:38:14] <Jalexander> Orionist: Did we get your question well enough (and get to them all?)
[2011-02-25 20:38:22] === mrmist <~mrmist@pdpc/supporter/active/mrmist> ``mrmist
[2011-02-25 20:38:22] === mrmist: member of #wikimedia-office, #wikimedia, #wikipedia-en, and #wikipedia
[2011-02-25 20:38:22] === mrmist: attached to hubbard.freenode.net ``Pittsburgh, PA, US
[2011-02-25 20:38:22] === mrmist: away with message ``I am away (screen detached) Please leave a /msg.
[2011-02-25 20:38:23] === mrmist is using a secure connection
[2011-02-25 20:38:23] === mrmist is logged in as mrmist
[2011-02-25 20:38:23] --- End of WHOIS information for mrmist.
[2011-02-25 20:38:34] <apergos> frankly...
[2011-02-25 20:39:03] <apergos> having a piece in some widely read press about this (if it could be done in an appropriate manner)
[2011-02-25 20:39:30] <apergos> I mean I bet the intersection of "influential editors" and "people who read the NYT" is pretty large
[2011-02-25 20:39:50] <Philippe> apergos: This ain't really sexy enough for the Times.
[2011-02-25 20:40:06] <Philippe> Besides, one of the ways to create change is to get buyin from influencers.
[2011-02-25 20:40:15] <Philippe> And you can get buyin pretty effectively in personl
[2011-02-25 20:40:17] <Philippe> person*
[2011-02-25 20:40:19] <apergos> sure
[2011-02-25 20:40:32] <apergos> but you'll reach only a teeny tiny amount of people in person
[2011-02-25 20:40:36] <Philippe> Well, some people can. I usually just start wars.
[2011-02-25 20:40:42] <apergos> heh
[2011-02-25 20:40:42] <Jalexander> it's the pin pointing that is the molehill atm I think
[2011-02-25 20:40:48] <Philippe> Yeah, that's why you get the influencers who will go out and do the same thing over and over agani.
[2011-02-25 20:40:50] <Orionist> Jalexander: yeah, kind of. I'm trying to throw ideas.
[2011-02-25 20:41:10] <Jalexander> the pool of the most influential is likely fairly small
[2011-02-25 20:41:42] <Jalexander> Orionist: I approve of that technique :) I have a tendency to do that (throw out ideas and get a debate going, figure out the real issues including the real reasons I want to do it)
[2011-02-25 20:42:12] <Philippe> (get stomped on by Philippe, cry in the corner...)
[2011-02-25 20:43:03] <apergos> no stomping.
[2011-02-25 20:43:09] <Jalexander> The smart influencers make the others think they came up with the idea and sit in the corner smiling
[2011-02-25 20:43:13] <Philippe> or crying.
[2011-02-25 20:43:30] <Jalexander> no I'm allowed to cry, just not once fundraising starts again
[2011-02-25 20:43:36] <Jalexander> or on test days
[2011-02-25 20:44:01] <Philippe> OK, kids... i'm going to go make dinner.
[2011-02-25 20:44:02] <Philippe> Ya'll have fun.
[2011-02-25 20:44:07] <apergos> later
[2011-02-25 20:44:28] <Philippe> ta
[2011-02-25 20:44:29] |<-- Philippe has left freenode (Quit: Philippe)