User:Statistics~enwiki/VP

Archive of conversations started on Village pump (miscellaneous)

Comparative contributions to the Encyclopedia and to supernumerary namespaces
In light of the discussion here, it may be useful to review Snowspinner's contributions to Wikipedia.

Snowspinner's edits to article namespace, 1 January 2005 - 15 April 2005

Snowspinner's edits to all namespaces, 1 January 2005 - 15 April 2005

For comparison purposes, the contributions of Everyking, netoholic, John Gohde, and Snowspinner:

--Statistics compiled by rrcaballo AT yahoo.com


 * For further comparison, User:Jimbo Wales has made 94 edits since January 1st, exactly three of which were to the article space, and only one of which was not about the edit war on Autofellatio, making him the least valuable of all. But welcome back, |orthogonal. Snowspinner 06:28, May 1, 2005 (UTC)
 * Oh, and he basically runs the Foundation, doesn't he. Oh yeah. &#8212;Charles P. (Mirv) 17:06, 1 May 2005 (UTC)

If there is a point in any of this, it escapes me. One can make good contributions either in a talk page or an article; one can make bad contributions either in a talk page or an article. -- Jmabel | Talk 06:38, May 1, 2005 (UTC)


 * It is provides data about stalking, I think. Pcb21| Pete 08:57, 1 May 2005 (UTC)


 * The point, I suppose, is either that Snowspinner does far more harm than good, since he contributes so little to the encyclopedia, or that he persecutes people who do a lot more good than he does. Everyking 14:21, 1 May 2005 (UTC)


 * That, and he's part of a whole class of non-editors who are not very interested in contributing to the encyclopedia. but are very interested in controlling and ordering around those who do. &#8212;Charles P. (Mirv) 17:02, 1 May 2005 (UTC)


 * I have no experience with John Gohde, so I have no idea whether reverting large numbers of his edits is a service or a disservice to Wikipedia. I can certainly say, though, that I have seen "contributors" where reverting 90% of their edits would enhance Wikipedia. -- Jmabel | Talk 23:46, May 1, 2005 (UTC)


 * John Gohde seems to have been banned for a year, starting yesterday. I'll let everyone draw their own conclusions on what this means about his contributions, 'cos I don't know the details of his case. -- Eugene van der Pijll 00:45, 2 May 2005 (UTC)

Snowspinner's edits: pages accounting for 1% or more of his total 424 edits to Wikipedia namespace, 1 January 2005 - 15 April 2005

Of Snowspinner's 424 edits to the Wikipedia namespace, 156, or 37%, were to an Arbitration page.
 * At least 31 of his edits to the Wikipedia namespace, or 7%, involved altering or removing edits made by John Gohde, or listing evidence against John Gohde.

Of his 230 edits to the Wikipedia_talk namespace 132, or 57% were to an Arbitration Talk page.
 * At least 19 of his edits to the Wikipedia_talk namespace, or 8%, were either direct responses to or disputation with John Gohde, or edits to a John Gohde Arbitration sub-page.

Of his 69 edits to User namespace, 3 were to John Godhe's own user page,
 * another 35 were to Snowspinner's sub-page of evidence against John Gohde, "User:Snowspinner/MNH_Evidence",
 * for a total of 38, or 55%, of his edits to the User namespace.

Of Snowspinner's 200 edits the User_talk namespace 19, or 10%, were to John Gohde's User_talk page.


 * what --Golbez 11:00, May 1, 2005 (UTC)

If a bunch of you have a problem with Snowspinner, start an RfC, so that he has a decent chance to reply. Frankly, this effort at prosecution by press release disgusts me, and probably should be considered a violation of the policy against personal attacks. -- Jmabel | Talk 23:54, May 1, 2005 (UTC)
 * These are merely summaries of what's available under "user contributions" for any user. Aggregating numbers is prosecution? Besides, as I understand the structure here, isn't Snowspinner Wikipedia's Public Prosecutor? If you'd like to discuss your concerns offline, please email me -- rrcaballo AT yahoo.com

What's your point? If you think something should be done, write an RfC, but the Village Pump is no place for your bellyaching. WP:ANI is at least slightly more on-topic for this discussion which is nothing but a large personal attack. RickK 66.60.159.190 17:49, 2 May 2005 (UTC)


 * I'm not sure why this is seen as a personal attack. It's just an attempt to study Wikipedia by analyzing patterns of edits. In all cases, I've endeavored to impartially show all edits, or to extract the most obvious aggregations ("top ten", "over 1%", etc.) of edits. Of course, being an outsider, I may have missed other patterns in the data. If so, I hope you'll point them out, either here, or, in confidence via email.


 * Nor am I singling out any one editor; Snowspinner is far from the only editor whose statistics I've analyzed. But when he was accused of stalking here, it seemed most fair to him and to all concerned to make available a portion of my work-in-progress, so that the discussion could reference real facts and figures rather than unsupported and overheated rhetoric. I'm not sure why this is being taken a "personal attack".


 * In any case, I'll think you'll find that my post of today (below) will alleviate your concerns about this being in any way an attack, but if it does not, please, please, let me know how you think I could better analyze these statistics, so that I can add your thoughts before publication of the final version.
 * -- rrcaballo AT yahoo.com

(history link)

Comparison of edits by Wikipedia Founder Jimbo Wales and editor Snowspinner
(I'm posting portions of my study here, on Wikipedia, in hopes of getting feedback from as many actual Wikipedia users as possible. Please help me to improve this analysis by adding your thoughts here, or by emailing me in confidence.)

Snowspinner has suggested that his pattern of contribution to Wikipedia is similar to that of Wikipedia founder Jimbo Wales's contributions.

In broad terms, Snowspinner is correct.

But when a detailed breakdown of contributions is examined, there are large dissimmilarites:

While both Snowspinner and Jimbo Wales make most of their edits to non-article namespaces, it's readily apparent that Wales spends most of his time (a full 65%) in User_talk, talking to other users; Snowspinner spends most of his time in Wikipedia, and most of that on Arbitration pages.

Also of note is the time spent discussing articles: while Wales makes few edits, he spends about over ten percent of his time in Article Talk discussing with other users what should be in articles; Snowspinner spends more time in the more contentious arenas of Deletion voting than in (possibly) non-adversarial Talk.

While Wales spends only a little more than one percent of his time in Wikipedia_talk, much of which involves discussion of policy, Snowspinner devotes over fifteen percent of his time to this.

(While Snowspinner spends, proportionally, twice the time in the User namespace as Wales, the difference isn't actually as great as it first appears, as it's primarily the result of Snowspinner compiling evidence, within his own user-subpages, for future Arbitration cases; this should be appreciated when looking at these numbers.)

Much of the proportional differences noted can no doubt be accounted for rather simply: Jimbo Wales and Snowspinner have very different roles on Wikipedia. Wales directs the Foundation, and as such plays a role of mediator, cajoler, conciliator and consensus-builder: he works primarily to build up Wikipedia, to iron out differences and to bring users together.

Snowspinner's role on Wikipedia is more adversarial, given that he works closely with the Arbitration Committee, at that Committee's request   in his role as "Public Prosecutor". (Snowspinner established what he calls the "District Attorney's Office", but as Wikipedia doesn't have districts, the title "Public Prosecutor" or "Attorney General" is perhaps more accurate.) To do that job, Snowspinner must necessarily spend his time seeking out evidence of wrong-doing, compiling evidence against wrong-doers, and bringing them to justice.

Snowspinner additionally contributes to Wikipedia as an enforcer of policy (e.g., 72 edits to Administrators' Noticeboard) and as to a lesser degree as a formulator of policy (44 edits in Wikipedia, 23 in Wikipeda_talk).

When the details are taken in account, then, Snowspinner's pattern of edits is only in a very superficial way similar to Jimbo Wales's. This reflects their very different roles on Wikipedia: Wales nurtures the Encyclopedia and the Foundation as a whole; while Snowspinner polices and prosecutes individual editors who get out of line, and formulates policy to keep them in line.

Of course, this is only a preliminary analysis of statistics by an outsider, and I'm sure it can be improved by experienced Wikipedia users. Please contribute your thoughts here, or in confidence to: -- rrcaballo AY yahoo.com


 * Why are you going through so much trouble to analyze Snowspinner's edit percentages? Yeah, I know that there's arbitration and all, but it seems rather irrelevant. How do his edit percentages even matter? Does his "role" in Wikipedia make a difference? Obviously no one has the same role as Jimbo.


 * - Pioneer-12 15:01, 12 May 2005 (UTC)


 * p.s. The analysis is impressively detailed, though. If you like doing this, perhaps you might be interested in doing horse racing analysis. You could win big money at the track if you could analyze horses as well as you do edits. It could be a profitable use of your skills.


 * As much as I'd like to take this opportunity to take a dig at Snowspinner, I can't. The first few times I read this, I let it pass. But I dislike to see bad precedent set.


 * Statistical analysis may demonstrate something of value -- when applied to groups. I don't see it is revealing to compare any two users. For that matter, I would hesitate to draw any conclusion from the edit ratios of the entire membership at a single point in time. What is the proper ratio of Userspace edits to Mainspace, of WPspace to either?


 * The proper domain of statistics is the aggregate and in the comparison of like to like. Often the nearest similar dataset is that which applied to the same project in the past -- thus, historical trend analysis. This automatically normalizes a lot of possibly misleading values.


 * I am nearly at the point of downloading the entire database and doing some real analysis on it, and you may be sure I will be slow to draw conclusions. For what it's worth, the *first* thing I intend is to throw out all user names in favor of internal system ID numbers. &mdash; Xiong &#29066; talk * 04:20, 2005 May 13 (UTC)

(history link)