Forum:The current state of RW's community

< Index | Site-related discussion

PROOF!!! that RationalWiki is doing just fine

Moved here from RationalWiki:Saloon Bar--ZooGuard (talk) 17:19, 9 February 2015 (UTC)

Note: Nutty Roux thinks the following post is an irrelevant distraction from his very important opinion that people he doesn't like are ruinous garbage, regardless of how many of them are editing. His opinion is very important, and definitely merits putting a passive-aggressive disclaimer in front of someone else's post.

Here to settle the perennial debate of whether RW is dying:

Click so you can actually read it.

Notes:

• That massive spike in February 2013 is from when Brasov ran a massive vandalism attack on us. He singlehandedly added 150+ "editors" to the counter, but they were all IP accounts that got blocked after three or so vandal edits.
• Editors were counted using Special:Editcount, checking each month since our founding and tallying up the number of editors with 3 or more edits for that month.
• A threshold of 3 edits to qualify as "active" was used to filter out spam accounts and new users who fix two typos then never come back again.
• These numbers are off by 1 or 2 per month. I'm using data I originally gathered in mid-2014, and due to deleted pages and revisions, a small number of barely-qualifying users get bumped off over time. Recalculate these same numbers a year from now, and you'll have lost another 1-2 per month. This decay rate is actually pretty reasonable and there's no reason to panic about it.

In brief:

• Our number of editors is stable.
• Last I knew, our visitor count was about 2 million unique visitors per month. If visitor count was falling significantly, Trent or David would've said something.
• We aren't freaking dying. The old guard just needs to get over itself.

Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 01:20, 1 February 2015 (UTC)

Um, Stabby, you ARE part of the old guard. Noisemobile (talk) 01:23, 1 February 2015 (UTC)

In my heart I'm the same fresh-faced young rebel who joined in 2007. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 01:26, 1 February 2015 (UTC)

Jesus Christ RA. This shit again. Aceⁱ⁹ 02:33, 9 February 2015 (UTC)

*summons Christopher Monckton to cherry-pick the data* 141.134.75.236 (talk) 01:29, 1 February 2015 (UTC)

Get over itself? They're all but gone. If you're not just popping offer because there's been some mention of older editors leaving, I congratulate you on completely missing the point. Nutty Roux (talk) 02:44, 1 February 2015 (UTC)

I suppose your point is that the current community is garbage, and it's driving away the old guard.

But of course, the original old guard was dead by the end of 2009, when PalMD, AmesG, AKjeldsen, Interpreter, Icewedge, Hojimachong, GodlessLiberal, and others I remember fondly were gone. This is the second set of "old guard" we're losing. These things work in cycles; we're losing 2009-2010 veterans today. We already lost the 2007-2008 ones. A couple more years and we'll start losing 2011-2012-era editors, then the 2013-14 ones, and so on. Each set gets replaced by new users, who then get tired and fade away after a few years.

And I'm getting real tired of passing references to "the old days" by people who weren't even here for them. Hell, you joined in March 2009, by which point you'd missed my heyday, as well as the heyday of everyone else who was big in 2007 and 2008. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 03:14, 1 February 2015 (UTC)

It's fucking crazy how "kids these days" can be 2 years. I mean, at some point it becomes a Theseus' ship question. If none of the articles are the same, none of the editors are the same, and none of the style is the same, is it the same rationalwiki? (Yes, probably). Ikanreed (talk) 04:01, 1 February 2015 (UTC)

That spike in February '13 -- is that when there was a brief but intense influx of gender/identity politics/social justice-oriented editors combined with the Facebook group blowing up? I remember that as being later in the year, but can't remember what else would have caused such a sudden spurt.... Peace. AgingHippie (talk) 05:32, 1 February 2015 (UTC)

(EC) I will never understand, especially within the last year, this community's obsession with alienating the so-called "old guard" (a label used so casually that it's really meaningless). A major reason that I am sick and tired of this place is because, especially in the last year or so, I've had more and more instances of my opinions being ignored or dismissed simply because "OMG!! OLD GUARD!!!" Like it or not, I'm a member of this community, too. And like it or not, I'm still a contributor here and still care about how this place looks. How/why does my having a different perspective on how I want things, and its implied correlation to the time I've been here, qualify its instant dismissal? It would seem to me that villainizing the "old guard" is little more than a code word; an othering of people we disagree with so that we can have an easier time ignoring their contributions here, and can more easily ignore any questions and skepticism that is raised. This is the same level of bullshit that causes normally level-headed people like myself to fly off the handle, act with contempt towards the current crop of regulars, and, yes, leave. In my case, it's the casual vilifying of us older editors that is my prime reason for no longer caring about the direction this place, and by extension the RMF, takes, and that includes taking my money elsewhere come fundraising time. Reckless Noise Symphony (talk) 05:44, 1 February 2015 (UTC)

Not really buying it. "Old guard" as a dismissal doesn't come up much. It comes up a lot as an excuse for why newer editors are destroying this wiki or whatever. I don't think anyone has said that long time contributors are a problem. Ikanreed (talk) 06:49, 1 February 2015 (UTC)

Oh? This whole shitstorm last autumn wound up boiling down to certain editors playing upon the "mean old guard" meme, even though it started with Brxbrx doing something uncalled for. And here's another example, of which I was NOT a party. Reckless Noise Symphony (talk) 07:23, 1 February 2015 (UTC)

I just finished reading your first example. I detect zero resentment of you for being an old-timer and a lot of Brxbrx's whingery and Tielec01 arguing against you — based on the grounds that you weren't upholding the Good Old Days' ethos enough. I pray your second example actually demonstrates what you claim. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 09:03, 1 February 2015 (UTC)

Your second example features three people, only one of whom (Brxbrx) is still here, directly accusing the old guard. Seven other people weighed in defending them.

I'm extraordinarily disappointed with your fall from grace, RNS. At some point, you've become a delicate flower who interprets a majority of people supporting your side as evidence that you're unfairly persecuted. You'd have been better off linking persecution complex in your earlier post. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 09:23, 1 February 2015 (UTC)

I find that surprising, since from eyeballing the recent changes it seems we are getting fewer and fewer edits. We did get the usual winter uptick this year too, but I could swear some years back 500+ edits per day wasn't that rare. --Someon (talk) 11:59, 1 February 2015 (UTC)

I have noticed that there are fewer editors with no life beyond RationalWiki. 2008 through 2010, most months had between 2 and 4 editors breaking 900 edits per month. Nowadays, it's uncommon to see even one person hitting 600. However, it's worth noting this isn't necessarily a bad thing: in the old days, the wiki was dominated by a handful of heavyweight editors. They drove policy and a lot of conflict (I say that know I was one of them), and their torrent of edits meant they could set the tone for articles across the wiki with little input from others.

Those days are over. Yes, robust communities tend to have a core of main users, but it's unhealthy for such a narrow group to have that much influence. We're actually more democratic than ever before — hence the sniping by Nutty Roux and RNS, who are frustrated with how their opinions aren't given the deference they once had. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 12:22, 1 February 2015 (UTC)

Not to mention wikiepdia isn't dead because their new article explosion is over. Rationalwiki has definetly hit the 80/20 threshold of covering pseudoscience. And there's not really any activity on conservapedia to mock anymore either. And every major theological idea has been hit to some degree or another. And even social justice stuff is pretty filled out, in spite of a lot of editors hating it. Ikanreed (talk) 22:40, 1 February 2015 (UTC)

As a noob, I've read many RW articles where I've just enjoyed the laugh and moved on. Most of my edits have been minor grammar, typo and +snark stuff. Why is that a problem? Lack of recent edits suggests that those articles were already near enough right (though I did once add a good-sized section into an article where I had local knowledge). Lots of WP articles are stable for the same reason. ProblemChimp (talk) 00:23, 2 February 2015 (UTC)

Looking at the demographic analysis recently (don't have the link at hand) most of RW's readers have high school education or below. Seems you have two opportunity/choices here: (1) take the ball and run with it using contributors who can pander/entertain/educate that market, or (2) recruit from that group of readers with high school education and below. nobs^{ISIS is a thorn in the world's ass and will be removed.} 01:14, 2 February 2015 (UTC)

I know the Alexa rankings have demographic info about us Rob, but I don't know how accurate it is. Tielec01 (talk) 01:26, 2 February 2015 (UTC)

As I recall, Trent hates Alexa because it's so unreliable, and prefers not to cite it for any sort of insight into RW. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 11:00, 2 February 2015 (UTC)

Communities change with time, it would be flat-out amazing if they didn't. Typical participant lifecycle in almost any online community is 12-18 months. Clay Shirky: A group is its own worst enemy. Me, based on previous discussion here: Internet fundamentalism doesn't actually work. Nostalgia is another word for brain rot. You can't have the good old days back, not only because now isn't then but you are a different person now - David Gerard (talk) 10:33, 1 February 2015 (UTC)

I'm about as "old" as "old guard" can get, having originally been a member of RW1. I think that some sort of conflict began between the "the old guard" and the "hot young bloods" began pretty early in the life of RW2. For instance one of the earliest fights was over the perceived excessive interest in CP. Anyway, with time the "hot young bloods" inevitably transom into "the old guard" and the cycle repeats.--Bob^{"I think you'll find it's more complicated than that."} 19:00, 1 February 2015 (UTC)

Just saw this thread, I'm a little incredulous that I have been accused of playing the 'mean old guard' meme. Having said that; I (an obscure nobody) and Brx (the wiki's most admired user) did bully the old guard including the board, the moderators and the techs. In the future we need to ensure that the power balance I have over these users is redressed. RNS allow me to offer an unqualified apology to you for daring to question your decisions - what was I thinking?

Anyway, love the analysis it has shades of Larron whose posts are sorely missed - I think the bombastic title is great; although as others have noted, the argument centres around the quality of new editors rather than quantity. Not taking a side on that issue, DGs link is good too, for anyone that hasn't read it. Tielec01 (talk) 00:20, 2 February 2015 (UTC)

I find myself wondering what was special about February 2013. - Smerdis of Tlön, ^{A ⇒ ¬A.} 05:52, 2 February 2015 (UTC)

A massive vandal attack that added 150 or so "editors" to the list. It's in the notes section under the chart. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 07:26, 2 February 2015 (UTC)

Those were the good old days & we shall never see their like again. Weaseloid^{Methinks it is a Weasel} 07:37, 2 February 2015 (UTC)

At least not while open proxies are blocked - David Gerard (talk) 09:12, 2 February 2015 (UTC)

Looks to me like the site stopped growing in May 2011. Just looking at this particular metric, anyway. And if editors making fewer edits is a "good thing"... ħuman 01:01, 9 February 2015 (UTC)

Did someone say something about circlejerks? Nebuchadnezzar (talk) 01:40, 9 February 2015 (UTC)

FWIW

Monthly edits at RW by name-space	Pages created per month by name-space
166 busiest non-anonymous editors in 2014	106 most edited pages in 2014 The earlier version of this picture only included data from Jan - Sep 2014. My apologies

--larron (talk) 11:47, 3 February 2015 (UTC)

Hey, long time no see, dude. Hope everything's going ok.

And huh. Thought Gamergate would show up in the most edited list given its relative popularity and being a target for vandalism early on (then again, it started very late in the year and the write-protect came pretty quickly, so not as much time for such edits). Noir LeSable (talk) 08:53, 5 February 2015 (UTC)

Looking at the fossil record, the Gamergate article had nearly 500 edits by the end of the year, and its talk page had over 1,200, so I can't see why it's been missed off the chart when there are pages with a tenth of that on there. Wéáśéĺóíď^{Methinks it is a Weasel} 21:54, 5 February 2015 (UTC)

recent edits at different wikis
RationalWiki	Citizendium	Wikipedia (DE)	Conservapedia

I'm quite pleased with these snapshots of the activity at various wikis: Gamergate is still going strong here at RW. The pictures are less than 100KByte big, and you have to look at the original file to profit from the embedded links.--larron (talk) 14:00, 7 February 2015 (UTC)

So

Let's say that RationalWiki has less activity. What has caused this? What can we do to increase activity? 32℉uzzy; 0℃atPotato (talk/stalk) 01:56, 9 February 2015 (UTC)

We could delete all our articles and write new ones in their stead. Uncyclopedia's tried that a couple times, but people just kept bringing back the old articles. 141.134.75.236 (talk) 02:15, 9 February 2015 (UTC)

Yeah, but one-fourty-one, we want good ideas. 32℉uzzy; 0℃atPotato (talk/stalk) 02:35, 9 February 2015 (UTC)

Well, we could advertise I guess? (Which might attract more bad editors than good ones.) 141.134.75.236 (talk) 02:50, 9 February 2015 (UTC)

I would be more interested in talking about improving activity, as opposed to growth for growth's sake. Peace. AgingHippie (talk) 02:57, 9 February 2015 (UTC)

What do you mean by "improving activity"?

I don't think that growing RationalWiki's activity counts as "growth for growth's sake" at this time, because RationalWiki is still too small to make a large impact in the skeptical community, and gaining more editors (and hopefully subsequently a larger community, more articles, and better articles) will help to increase RationalWiki's importance. 32℉uzzy; 0℃atPotato (talk/stalk) 03:22, 9 February 2015 (UTC)

I think he means improving the quality of activity as opposed to increasing the quantity of activity. 141.134.75.236 (talk) 03:43, 9 February 2015 (UTC)

Pretty much. We have a lot of stubs that, I'm assuming, are still stubs because the users we have don't have the expertise to make them into better articles. So while I, of course, hope we can keep a critical mass of active and enthusiastic editors contributing on a regular basis, if we really want to "make an impact in the skeptical community," or hey, let's dream big, instead of preaching to the choir in the sceptical community, actually making a broad impact in the general community and driving debates and changing minds, we need people who are smarter than, say, me, more educated than, say, me, with more expertise and authority than, say, me. Peace. AgingHippie (talk) 03:51, 9 February 2015 (UTC)

Sure, we need better articles for a larger impact. To get better articles, both more editors and "better" editors will help, because everyone can add a reference or write a paragraph, even though not everyone can understand what that reference says. If we assume that our editors aren't significantly increasing their education, then the only way to get "better" editors is to get new editors who happen to be "better". So it seems to me that getting more editors should definitely be a goal for RationalWiki, especially if the "critical core" is in danger. 32℉uzzy; 0℃atPotato (talk/stalk) 04:10, 9 February 2015 (UTC)

Our articles are of a better standard - and a better expected standard - these days than they were before 2010, IMO. Editors are not at each other's throats as much as they were then. (You think the arguments these days are chicken coopery? Hah!) We have similar edit numbers if less people spending their entire goddamn lives here (apart from me, but I am an actively dull suburban dad who delights in not having a life).

So basically, we need to get more n00bs interested in contributing at all - David Gerard (talk) 10:57, 9 February 2015 (UTC)

Thing is that older (i.e. long term) editors have used up all their knowledge/enthusiasm on subjects germane to RW's mission, so there are, as I see it, three ways to increase the editorship:

1: acquire editors with more knowledge on existing subjects - there are several instances within the last twelve or so months where this has happened.

2: find more subjects within the mission for editors to have a go at - Gamergate, Bitcoin etc, come to mind.

3: expand the mission -?

Scream!! (talk) 13:05, 9 February 2015 (UTC)

Not sure that's it. I frequently write here about things I don't know about beforehand, so I do research to write them up. (You know you're onto a good one when you feel stupider after writing it up.) - David Gerard (talk) 13:44, 9 February 2015 (UTC)

1) and 2) are laudable. 3) is less so, as I think a reasonably tight focus is what makes a non-wp wiki productive (notwithstanding some recent attempts to define some on-mission topics as off-mission). And we mustn't forget 4: Keeping existing articles up-to-date with new material. It's not as if antivax is likely to shut up shop any time soon, for example. Queex_{chthonic murmurings} 14:09, 9 February 2015 (UTC)

(1) How does one "acquire" editors? (2) RationalWiki seems to do that automatically, often. Should we try to increase our coverage of current events over the level that we do now? (3) I'd be cool with a mission change. Our current mission doesn't directly mention the topics of (a) social justice, (b) non-authoritarian political extremism, (c) conspiracy theories, and (d) logic, language, and argumentation, all of which take up significant parts of RationalWikispace. But changing the mission doesn't really affect how editors act, in the same way that the new blocking policy isn't really any more influential than the old. 32℉uzzy; 0℃atPotato (talk/stalk) 16:37, 9 February 2015 (UTC)

New blocking policy? Where's that? Peace. AgingHippie (talk) 16:38, 9 February 2015 (UTC)

RW:B. (Before, the policy page was basically nothing.) 32℉uzzy; 0℃atPotato (talk/stalk) 16:47, 9 February 2015 (UTC)

Equally important is why it happened, as that informs any changes made to improve. I think the only serious charge for why has been that new people are driving the old people out, but if the old editors were really that much more active, they'd swamp out new editors. I have no alternative explanations myself. I have no idea what happened. Ikanreed (talk) 15:13, 9 February 2015 (UTC)

Or it could just be that old editors left, for any of a host of other reasons (bored, moved to another site, felt that their field of editing was complete enough, etc.). One thing that I noticed in LArron's graphs is that the amount of Main edits has stayed mostly the same, while Conservapedia edits (which used to make up 1/3 of edits) have declined to maybe 1/20th of edits; it might be that Conservapedia-watchers have largely fallen away from RationalWiki. 32℉uzzy; 0℃atPotato (talk/stalk) 16:37, 9 February 2015 (UTC)

The rationalwiki namespace also fell off hard, if we're doing that analysis. Was it more popular to chat on the Saloon bar in the past? Ikanreed (talk) 16:41, 9 February 2015 (UTC)

I think there's a lot less discussion about policy, for one. 32℉uzzy; 0℃atPotato (talk/stalk) 16:47, 9 February 2015 (UTC)

As [ZooGuard] requested

Moved here from User talk:ZooGuard.--ZooGuard (talk) 17:22, 9 February 2015 (UTC)

That awkward moment when I claimed the community isn't declining, but it turns out we peaked in 2011 :-/

Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 03:03, 2 February 2015 (UTC)

Let's assume that this graph accurately represents "the community". "The community" went from ~80 to ~60. Not really that large of a change, but pretty sustained.

How could RW reverse this decline? 32℉uzzy; 0℃atPotato (talk/stalk) 03:36, 2 February 2015 (UTC)

Also: Can we get a best-fit equation for this? 32℉uzzy; 0℃atPotato (talk/stalk) 03:37, 2 February 2015 (UTC)

I sent you an updated Excel file that has that chart in it. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 04:00, 2 February 2015 (UTC)

The best fit for this graph certainly does show that 2007 - April 2014 peaked in 2011. But, if the best fit line is to be believed, RW has been doing well since April 2014 and is due for an upswing. 32℉uzzy; 0℃atPotato (talk/stalk) 04:12, 2 February 2015 (UTC)

BTW, @ZooGuard and @Weaseloid: Should we be having these discussions at the Bar, so everyone else can see 'em? 32℉uzzy; 0℃atPotato (talk/stalk) 04:19, 2 February 2015 (UTC)

And here it is extended 1 year using the best fit line. Probably a wee bit optimistic, and if the line is extended out 10 years it quadruples and then shoots into the ground. 32℉uzzy; 0℃atPotato (talk/stalk) 04:27, 2 February 2015 (UTC)

That's because it's not a great best fit equation for making general predictions. Excel doesn't know how to add constraints on future behaviour and just tries to match whatever best fit works best for the given data set. Also, that uptick in the final few months is skewing the data, making the best fit equation even less useful. - Grant (talk) 04:50, 2 February 2015 (UTC)

Yep. And it wasn't even a good fit in the 1st place. 32℉uzzy; 0℃atPotato (talk/stalk) 05:16, 2 February 2015 (UTC)

Exactly. Unfortunately the "spikey" nature of that data-set makes a best fit dubious at best. I don't believe it will be possible to make further concrete predictions until we can see whether the current spike is another temporary one (as has happened several times in that data set) or the start of a longer-term trend. - Grant (talk) 05:23, 2 February 2015 (UTC)

I'm going to write a Python script that queries RW and exports the result as a CSV file so you won't have to do this by hand in the future. As for the right place of the discussion, I suggest migrating this and the overlong Saloon Bar thread to a Forum page.--ZooGuard (talk) 12:54, 2 February 2015 (UTC)

Any chance you could furnish me with the raw numbers for now? I feel the urge to bust out some DLM magic on this bad boy. Queex_{chthonic murmurings} 2015-02-09T17:30:45

Here's a Dropbox link to the Excel file containing all the raw numbers I gathered. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 17:52, 9 February 2015 (UTC)

Looks like two superimposed sine waves to me - one with a period on the order of 8-12 months and a much longer one that we are about half way through. Which would mean we are due for a short term drop over the next few months, but a glorious gentle recovery over the next 3 or so years. I think there is also a very small subtle linear increase underlying those two functions, perhaps on the order of 2-3 more users per year being on this list. ħuman 18:11, 6 March 2015 (UTC)

How about that spike on search traffic tho ikanreed ^{You probably didn't deserve that} 18:18, 6 March 2015 (UTC)

Best fit

This discussion was moved here from User talk:Stabby the Misanthrope.

I checked all of Excel's available best-fit lines on your data, and the best best-fit line for the number of active users is a graph with equation: Number of active users = 109.37 * (time-in-months^0.2585) with an R² of 0.781. So, according to the current graph, RationalWiki will continue to grow at an ever-decreasing rate. Whaddya think? 32℉uzzy; 0℃atPotato (talk/stalk) 01:34, 2 February 2015 (UTC)

That makes sense. We're not exploding, but we're definitely not shrinking. Slow growth feels like a good fit. In a couple hours I can send you an updated version with separate columns focusing on editors making 30+ edits per month and 300+ edits. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 01:38, 2 February 2015 (UTC)

Kewl. I gotta do some real work, but MOAR DATA is always nice. 32℉uzzy; 0℃atPotato (talk/stalk) 01:41, 2 February 2015 (UTC)

The fits are meaningless. You probably have to use generalized least squares instead of ordinary. I suspect the bars are all within error of each other. Sterile (talk) 00:38, 10 February 2015 (UTC)

No Sterile. Nutty is wrong, RA is right. That's all he gives a fuck about. Aceⁱ⁹ 00:43, 10 February 2015 (UTC)

@Sterile: I'm new to anything like this. Does Excel have that capability? If not, what program should be used? 32℉uzzy; 0℃atPotato (talk/stalk) 00:51, 10 February 2015 (UTC)

Beware forward projecting on arbitrary data. Unless you have a causative model that aligns with your correlation, you know very little. Ikanreed (talk) 14:33, 10 February 2015 (UTC)

There is no particular reason to fit the data. There is real no basis for a mathematical model to fit to, and hence and mathematical function is arbitrary. Basically, the data are and any fit is bullshit unless you have a reason otherwise. Sterile (talk) 16:18, 10 February 2015 (UTC)

Right. That's what I was saying. It can be impossible to tell a quadratic equation from a sine wave from a hyperbola from an (inverted) logarithm from a exponential curve, over some data ranges. And that's when your data is perfect and has no errors or noise. Ikanreed (talk) 16:24, 10 February 2015 (UTC)

It's even worse than that. If you expect pressure to be inversely proportional to 1/volume because little atoms exert small amount of force and decreasing the volume increases that force per area, and hence the smaller volume, the higher the pressure. Here, there is no real basis for a particular model, no connection between number of editors and months. (Maybe a steady state over time if there is an editor "half-life", hence y = a*x, after a particular time, but even that has to particular reason to be valid.) Just because Excel has different regressions doesn't give you a valid reason for using them. You can plot the data with fits all you want, but they have no meaning and certainly no predictive value. Sterile (talk) 16:41, 10 February 2015 (UTC)

Another point to ponder: Though our overall rate of deletions is reasonable (according to my sampling ca. 4.6%), it isn't constant, but decreasing over time, especially as many of the older articles in the former Conservapedia-namespace were deleted. Therefore one can assume that editors of the early years are underrepresented. --larron (talk) 17:03, 10 February 2015 (UTC)

A possible model (one that mimics many websites) is an exponential to steady state. (Steady state as some editors arrive and some leave at corresponding half lives.). That would give a function like y=a-b*exp(-k*t). If you believe that (I don't), the editorship would top-off at a steady state of around 353 editors. But a sudden influx or outflux due to an event or slow decline in interest over time could easily change that. Sterile (talk) 17:12, 10 February 2015 (UTC)

So, what we know from serious analysis is that we don't actually know much? Ikanreed (talk) 17:19, 10 February 2015 (UTC)

*Shrugs.* What were you expecting? Sterile (talk) 18:01, 10 February 2015 (UTC)

DLMs with R

Some quick and dirty DLM modelling. A DLM model with a trend term, MLE estimates used for V and W (because the crappy R package I found doesn't handle discounting methods), on the 'users with x+ edits' data. In all these, green is the data, brown the one step ahead forecasts (with 95% DI), orange the estimate of the trend parameter (with 95% CI), red=zero.

The trend estimate is somewhat conservative. Better models would incorporate the spike in the 3+ edit data, and try to introduce different trend phases in order to capture the behaviour. Over-fitting might be a problem in that case, as would assumptions regarding the underlying situation. In any case, I'm not being paid for this, so that's enough work. Queex_{chthonic murmurings} 14:02, 10 February 2015 (UTC)

RationalWiki:Content dumps

Since people are interested in doing statistics, could you post a full (incl. all revisions) site dump? Maybe you could even host it on https://archive.org/. Right now the current-revs-only http://rationalwiki.org/dumps/rwcurrent.gz is almost a year old. --Someon (talk) 02:03, 10 February 2015 (UTC)

archive.org!! of course!!!!

I actually have the anonymised versions of the Squid logs going back a few years too - that raw data can go on archive.org as well. (Then someone can build a http://stats.grok.se equivalent if they like.) I'll try to remember to get in touch with them ... - David Gerard (talk) 16:01, 10 February 2015 (UTC)

Feb 2015

In Feb 2015 there were 942 users with 1+ edits (versus 941 in January), 358 users with 3+ edits (versus 365), 76 users with 30+ edits (versus 78), 11 users with 300+ edits (versus 11), and 1 user with 1000+ edits (versus 1). Since February is 10% shorter than January, seems like a good sign. FuzzyCatTomato (talk/stalk) 18:52, 1 March 2015 (UTC)

Probably need to click the image for a good picture. 32℉uzzy, 0℃atPotato (talk/stalk) 22:13, 1 March 2015 (UTC)

September 2014 still seems inexplicable. Ikanreed (talk) 14:35, 2 March 2015 (UTC)

It was just after Gamergate started, perhaps it was Gater trolls? Bicycle^wheel 20:55, 3 March 2015 (UTC)

In 14-Sep the number of editors went down, maybe to go harass Ms. Quinn instead of RW? Cømrade FυzzчCαтPøтαтø (talk/stalk) 21:47, 3 March 2015 (UTC)

I'd love to see that chart with the vertical axis changed to number of edits - so there would be a total edits top line, and areas below it showing the "influence" of editors with >x edits. ħuman 18:15, 6 March 2015 (UTC)

Slow weekend

Usually RW hits 400 edits per day, but we've had maybe 600 over the last 3 days. What gives? FᴜᴢᴢʏCᴀᴛPᴏᴛᴀᴛᴏ﹐ Esϙᴜɪʀᴇ (talk/stalk) 01:00, 10 March 2015 (UTC)

I don't think that this is the right place for this conversation.--ZooGuard (talk) 16:19, 10 March 2015 (UTC)

Not precisely a wrong one. Dips happen. G*m*rg*t* was quiet. I was busy - David Gerard (talk) 19:31, 10 March 2015 (UTC)

@ZG, where else would this go? Saloon?

@DG, sounds right. Looks like you and Ryulong were both busy, since GG's had only a few dozen edits. oʇɐʇoԀʇɐϽʎzznℲ (talk/stalk) 21:00, 10 March 2015 (UTC)

Edits, edits by amount of edits

Here's the edits. Excel: https://www.dropbox.com/s/d7l0xfvr52ytnod/RationalWiki%20users.xlsx?dl=0
RW's doesn't seem to be doing so hot. We've got maybe 33% more users than in '11-'13, but less or comparable edits. αδελφός ΓυζζγςατΡοτατο (talk/stalk) 01:27, 24 March 2015 (UTC)

IS there a motive to trying to prove we're "not doing so hot"?--Miekal 01:37, 24 March 2015 (UTC)

It's useful information. We do need more editors - we have way more content than curators. I believe Conservapedia space has gone quiet, and there were people for whom that was their primary interest - David Gerard (talk) 11:38, 24 March 2015 (UTC)

LArron's graph has CP as maybe 20% of edits (now maybe 1%), while mainspace edits have continued to rise regardless. FU22YC47P07470 (talk/stalk) 14:27, 24 March 2015 (UTC)

@DG: If RW wanted to get more editors, how would it? 32℉uzzy, 0℃atPotato (talk/stalk) 01:04, 25 March 2015 (UTC)

Recruit your friends!! JOIN THE CULT - David Gerard (talk) 08:55, 25 March 2015 (UTC)

Have you/the RMF considered reinvesting any of the funds raised into any form of advertising or recruitment? For example, you can run Facebook advert campaigns for only for a few bucks a day and generate a lot of engagement. You have to be a bit careful with the targeting and keep an eye on what's happening, but it may be worth trying for a few days to see if it's worthwhile. Wéáśéĺóíď^{Methinks it is a Weasel} 01:12, 18 April 2015 (UTC)

I wouldn't consider it very worthwhile. The Official FB Page has 3000 likes and a few regular readers, the Twitter about the same; they occasionally net an editor or two. I've barely posted to either in a few weeks, been stupidly busy. - David Gerard (talk) 22:39, 18 April 2015 (UTC)

Note about the FB, though: I'm fairly sure that under Facebook's new rules for organizations etc. most of those 3000 don't see your posts. Could be misunderstanding, though. PacWalker 01:53, 2 May 2015 (UTC)

If you're a page editor or above, it shows you the count of views. Typically 500-1000, could be lower or higher - David Gerard (talk) 17:43, 2 May 2015 (UTC)

Well that's nifty. PacWalker 17:44, 2 May 2015 (UTC)

Mar 2015

Excel: https://www.dropbox.com/s/d7l0xfvr52ytnod/RationalWiki%20users.xlsx?dl=0 Cømrade FυzzчCαтPøтαтø (talk/stalk) 22:21, 1 April 2015 (UTC)

April 2015

Excel: https://www.dropbox.com/s/d7l0xfvr52ytnod/RationalWiki%20users.xlsx?dl=0 FU22YC47P07470 (talk/stalk) 01:46, 2 May 2015 (UTC)

I know Alexa is unreliable, but it says we've fallen from 15,000th most visited to 20,000th since Oct. [1] FuzzyCatTomato (talk/stalk) 01:59, 2 May 2015 (UTC)

Don't know what the hell that graph is doing; both the number quoted next to it and on the toolbar are 17611ish. PacWalker 02:03, 2 May 2015 (UTC)

Or is the 17611 a moving average of what's in the figure? PacWalker 02:03, 2 May 2015 (UTC)

It's a moving average, I think. Whatever it is, the current amount is down. FüzzyCätPötätö (talk/stalk) 17:59, 2 May 2015 (UTC)

So we had a (comparatively) ridiculous amount of traffic on 1 April? That's amusing. PacWalker 18:00, 2 May 2015 (UTC)

Heh. RW's traffic appears to be somewhat event-based; someone posts a link on Reddit and we get a few trolls, etc. Probably someone linked to RW somewhere popular. FrizzyCatPotato (talk/stalk) 18:04, 2 May 2015 (UTC)

Mid-May

By the current numbers, May would have about 17,000 (+2000) edits and 1,400 (+450) users. Cømrade FυzzчCαтPøтαтø (talk/stalk) 17:33, 13 May 2015 (UTC)

Yeah, nope, (as expected) the numbers were waaay off. About 14,000 (-1000) edits and 1,000 (+50) users. More importantly, methinks, is the drop from about 16,000th most popular website to 18,500th. Graphs update tomorrow. ʇυzzγɔɒтqoтɒтo (talk/stalk) 01:35, 31 May 2015 (UTC)