User talk:FT2/Article supervision

From Wikipedia, the free encyclopedia

Thanks for reviewing this page!

Contents 1 Background to proposal 1.1 Effects of supervision 1.2 Typical cases 1.3 Existing policy and dispute measures that should work, but don't always 1.4 Common q's 2 Discussion 3 Supervision acceptance process 4 Supervisors' guide 5 Example of where this might be useful 6 Supervising the supervisors (grievance procedure) 7 What this needs... 8 What supervision is not 9 I rather like this 10 I appreciate the thought 11 How about 12 suggestion 13 A discussion 14 Suggesting a step 15 actual objectivity

[edit] Background to proposal

Most of the time, when an article gets mired down, clearing it up is fairly easy. In a few cases, the present dispute resolution handling isn't really up to the job. When that happens, it causes disproportionate problems, damage, and burnout to editors. In some cases (I can think of two immediately) the damage has gone on over a year before it was even stemmed. Recovering from that damage will take longer. One of the two hasn't recovered even a year beyond that. Many editors burned out over it.

The existing dispute resolution processes do not always address this well. Mediation requires good faith, arbitration is a point verdict rather than an ongoing guidance and is usually only reached at the end of a long trail. Also when bad faith is at play sometimes the genuine editors are wikistressed out and many leave long before that point; the damage is done. ANI, RFC and the like rely on individual editors at a point in time, and aren't really designed to address ongoing 'pushing'. This proposal would look at situations where an ongoing form of supervision is better suited, rather than mediation or arbitration.

Unlike most interventions, the intervention here would be to enforce the establishment of good editorship, in an environment where AGF cannot be assumed, and create a more level playing field for a neutral and objective approach to become established by editors who are prepared to respect Wikipedia editing policy. If the article could be better dealt with by some other avenue, or the dispute was not serious and established enough, it would be turned down as premature.

In this sense supervision parallels mediation; both take place over a medium period of time, but where RfM requires and assumes good faith, supervision is explicitly designed for disputes where good faith is dubious or not to be assumed. Where mediation looks to editors to understand policy and work together, supervision warns and removes editors who persistently refuse to edit in a policy-compliant manner.

These situations don't happen often, but they do happen, and WP:DR doesn't handle it well right now. When it does happen, a disproportionate amount of damage is done to the project and to its editors. In some cases, I think we need a dispute resolution avenue that doesn't need to go to arbcom, that's supervisory rather than judicial (somewhat like Mentorship), and which is accessible earlier on as one of many dispute options.

FT2 ^{(Talk | email)} 10:54, 11 July 2007 (UTC)

[edit] Effects of supervision

Effects of supervision.

At the least, bad conduct is identified and promptly removed, policy breaches promptly addressed, deadlocked debates freed up. Well meaning editors will still have to write the article, but will be freed from every edit being a 3 page essay, personal attack, and an uphill struggle. Problematic conduct will be much easier to deal with. Seriously deficient editing environments should not have to be borne by well-meaning editors in this manner. Supervision provides notice to all editors, on all sides, that policy violations will proactively meet rapid warning and short term removal as they arise, and for an extended period of time sufficient to inhibit bad actors, and re-establish good editorship and 'feeling' on the article. It thus takes the burden off editors who don't have resources, or means to effectively deal with it. It may address the concerns of many who tire of Wikipedia or specific articles due to a perception that Wikipedia doesn't presently address such issues effectively or rapidly enough. (This is a common theme in expert retention.) It is hard to abuse since supervision merely enforces that work on the article is policy-compliant, material presented is not misrepresented, and over a period of time identifies and cleans the article of issues caused by those who cannot or will not act appropriately (on both sides). As such, supervision is also a deterrent and educator. In extreme cases, I'm thinking of two serious sock-group POV wars I witnessed. Both lasted around a year before reaching arbcom, with huge damage to the articles and editorial environment concerned that's proven hard to fix (several editors left the project as well). Encouragingly, supervision would have addressed both of those much earlier, whereas no other dispute process tried had any serious effect.

[edit] Typical cases

Examples of dispute features where supervision might be useful.

The two cases I'm thinking of (one I was in there from the start, the other I wasn't) have these features in common: A group of editors appear to be acting towards a similar approach. But WP:RFCU doesn't show anything, and they are aggressive and WP:OWN the article by asserting strongly what the article talk page will consider a mainstream view. The control of the article is gradually acquired. There's usually a degree of subtlety about it, and the subtle use of sliiiiightly pushing it over weeks or months makes it hard to tell this is happening, and harder to present a case that it's POV. Claims that it's POV or not following policy meet the same response as any "civilized person meets the sharks": tactful suggestion that the NPOV editor needs to read up more, or doesnt know the real research, ultimately gliding into a strong "that person is known POV and biased" accusation. Almost inevitably not trapped before it's been dire for months, tried everything several times, and heading for arbcom. Even then, the full scale of it is suspected but unproven in most cases. Notably, supervision also for the first time addresses two areas which are serious problems when they arise at present, involving disputes where group-owned editing is creating a problem or dominating article editorship, but puppetry cannot be demonstrated. The focus of supervision is not on sock usage, but on policy usage. Supervision is thus capable of handling editors who act alone or in a grouplike manner, regardless of whether they are in fact puppets of some kind or not. This aspect also makes supervision one of the few approaches that (looking back) would have be both applicable and effective in handling various editors who were retrospectively identified as virulent and sneaky sock-puppeteers engaged in subtle but damaging POV warfare (as well as a while later, their suspected - but unproven - reincarnations).

[edit] Existing policy and dispute measures that should work, but don't always

Existing policy and dispute measures that should work, but don't always.

RFCU should help, but if it's partially behavior based, or suspected, often RFCU fails when later it was found there was undeniable evidence. That was the case in both the cases I'm thinking of, both of them did finally go to RFCU with not much gained by it. All that happened in each case was the sock-gang uses it as evidence of bias/POV by others, and clean heels for themselves. DR in general should help, but many steps are open to possible tendendiousness and a common handling is to "round robin" the issue, which with multiple editors not yet being confirmed as the same, is hard to address. SOCK (or meatpuppetry) would help, but seems to fail in practice (at least up prior to Arbcom). Often puppetry only becomes evidenced for certain a long time on. The near-inevitable answer by sock-actors before that point is 1/ their accusers are POV warring, and 2/ the fact they and the other editors all agree is not because they are puppets, but because they know the subject and others are biased. It's often hard under existing DR processes for editors to show puppetry in that scenario, despite the similar editorship. Content policies such as CITE / V / RS... this can be part of the game. Often cites may not be cited in a balanced manner (ie selective), and in many subjects Wikipedians have to assess from cited evidence what "the majority" or "notable" views are. In many subjects that can be contested - by ignoring evidence, by claiming the cites are biased, by claiming the cites are not reputable (theres always a critic for any view), etc. NPOV ... should solve that, but fails too, because of the fundamental aggressive stance that the view being taken is "the neutral view". ADMIN action often is sporadic at best, and any admin walking into this will probably be met by seemingly solid and good reasons why its "the other person's fault for bias". RFC - visitors may give advice, but often arent around enough to help with the "meat" of it, the actual editorial debate handling. Those that are often quickly tire of the hostile environment. RFM (including enforceable mediation) relies on co-operation and collaboration; mediation terminates without conclusion in the face of repeated stonewalling or non-co-operation. It is not very well suited as a dispute resolution tool to resolve situations where good faith is covertly dubious. ARBCOM half sorts it out, but it's a "point action" and not ongoing... and in such situations the "bad guys" seem to be very good at setting it up so they look on the side of the angels, unless examined closely. It's also the wrong kind of tool here; what's more useful here (and easier) is the inhibiting of improper activity for a while. It can take a long, long time to get to Arbcom, and because of muddied waters, hard to know (for editors) how to prove what they know is going on. Burnout long before this is quite common, leaving long term damage. In part this is because whilst arbitration is often effective, getting to that point (and with good evidence) is not a quick process. MENTORship and PROBATION are the closest things to supervision at present, but are only introduced sporadically, subsequent to arbitration, and even then subject to an arbcom remit determined case-by-case. It's an option within certain arbcom rulings, rather than a standard dispute resolution process that anyone can request at an early stage, and (for mentorship) typically applies to individual specified editor/s[1]. The usage is different from supervision, which seeks to create at an early stage and at editor request, a standardized and level policy-based field on an article, without focussing on any given editor/s, to allow good editorship to become re-established before problems go further.

[edit] Common q's

Q&A.

Would it create a cabal or such? Probably not, no more than recognized structures like RfM do. The aim is deliberately low key, and simply to address persistent and visible problem conduct in order to allow a level playing field and a degree of fresh air to be re-established for editors under stress. Is this a big job to do, is it WP:CREEP? Since it's not about writing the article, it's a quick job. Scan the page now and then, check assertations in disputes seem fairly founded, has anyone been acting up, or is everyone playing nice? Is this necessary? I think on balance, something is, and this might do the job. Its an extra option on dispute handling. There's an actual visible need for it, because there are actual visible disputes which hit this problem - and stay there a year or more, and it does a lot of damage to goodwill and the encyclopedia. We don't have our house in order for this type of vandal - and I think we need to do so more. Would it remove power from editors to some clique No. It creates a supervision and policy enforcement regime option for articles whose editors are screaming for it, at a stage before they hit the inevitable burnout or arbcom far down the line, and before most of the damage is done and well-meaning editors have left. It's also deliberately quite tightly focussed and limited in duration to around 2 months or under (long enough to get good editing practices back after bad actors have become obvious or mended their ways, short enough to be recognized as a temporary measure). Who joins it, how, and what stops bad actors joining that project Given the situation involved, experience is obvious. Because of the need to block, it's probably mostly sysops, and perhaps a nomination system, or something like the system that works for the medcab. A lot of thought has gone into avoiding both bureaucracy and problem editors, to check consensus without being "heavyweight", and the current proposals (especially the current "hybrid" proposal) seem to look good. There's two proposals on the table so far, others are welcome. Nb - criteria and processes are listed more fully on the proposal subpage. Don't admins do that anyhow They should, but in most cases a strict attitude on conduct is at the whim of individual admins, happens much later on in the dispute (if at all), and is sporadic rather than ongoing. It doesnt provide a steady sense to contributors that this article is being managed back to stability; there isn't the context that the article is under extended supervision by a group of neutral experienced editors, which this approach would allow. Place in dispute resolution Parallels mediation, so RFC may well have been used already, as might mediation, but probably not yet arbitration. Used when a serious entrenched deadlock is becoming apparent, usually due to warfare, ownership, or non-AGF, and where simply enforcing an appropriate playing field and policy compliance would probably solve most of the problem in the view of participants. Also used if mediation fails due to being 'blocked' and enforcement of policy-based editing seems a more helpful or simpler remedy to participants, than arbitration. Appeal? Yes - RFC, ANI and Arbcom most likely. Mediation less so given the kind of situation this will most be applicable to.

FT2 ^{(Talk | email)} 02:23, 10 July 2007 (UTC)

[edit] Discussion

I've taken the liberty of some grammar corrections, as well as a couple of tweaks. You're welcome to revert it (since it's your writeup). Peace.Lsi john 12:44, 10 July 2007 (UTC)

Thanks, yes, escalation to arbcom would be the norm, as I was thinking it too.

The "warning and/or block" was deliberate though, as was "may/will" - I can see cases where a short term block for a 1st act might be appropriate, and not every fact will be personally checked by supervisors. Bearing in mind this is a context where bad faith editing may be common, I'm not much minded to tie hands that way, although I don't have hugely strong feelings on it. Are you okay if I put those two word choices back as it was? If consensus feels it should be "warn then block" rather than "warn and/or block", I won't disagree. FT2 ^{(Talk | email)} 12:07, 11 July 2007 (UTC)

Also, should it be "supervision lasts for a maximum of 2 months" or "supervision will only exceptionally last more than 2 months"? FT2 ^{(Talk | email)} 12:07, 11 July 2007 (UTC)

Should it state explicitly, "Simple revert warring or addition/removal of content, unless accompanied by these, are not usually sufficient to request supervision." ? FT2 ^{(Talk | email)} 12:07, 11 July 2007 (UTC)

It's an interesting idea, and I think that something like it would be very welcome. I hope that it would also help out with ownership issues; I've seen in more than one case where a group of editors will latch onto an article, summarily revert changes they dislike, and dismiss or ignore any editor that brings up any issues on the talk page. A few points to consider:

• Currently, the proposal calls for supervisors to self-select what articles they will supervise. I'm afraid that might lead to supervisors with a bias toward an issue choosing to supervise an article that comes up for supervision. This is likely to worsen, rather than ameliorate, problems.
• Supervisors should probably be instructed to give additional leeway toward newer editors, at least at first. There's a big difference between someone who's still stumbling through learning how things are done and needs guidance, vs. someone who's been around long enough to know better and is being deliberately obstructionist. (Of course, once the newer editor has been told "Hey, you shouldn't do that" and keeps it up, they can get the hammer just like anyone.)
• How would supervisors actually be chosen in the first place? If that's self-selection too, it could lead to the same problems as above, people who wouldn't be very good at it or will (intentionally or unintentionally) use the position in a biased manner. If there's a committee or something that approves them, a lot of caution would have to be used in order to ensure that they're selected well.
• Supervisors should be given the option of banning an editor from editing the supervised article (and optionally its talk page) for a period of time rather than blocking. Of course, if the editor disregards such a ban it should be enforced by block.
• Sometimes there's a topic or group of articles, rather than a single one, which tends to have the same problems involving the same editors. In cases like this, it probably would help to have one supervisor in charge of those problem articles, since they'll be familiar with the parties involved and the situation in general. (If there are a lot of articles involved, it may also be useful to have a group of supervisors involved, if they'd be too much for one person to reasonably do.)

This stuff aside, I think something like this is needed. There are articles out there which are in pretty desperate need of help, and for which RfC, mediation, and the like aren't going to work on, because some parties just are not going to listen. At the same time, sometimes the level of disruption doesn't get quite high enough to result in it going to ArbCom (and if anyone does get in there and start fighting the entrenched owners, they'll find themselves sanctioned at ArbCom too!). I think this would be a good way to fill in that gap. Seraphimblade ^{Talk to me} 22:51, 13 July 2007 (UTC)

Seraphimblade, speaking for myself, thank you. As a new editor, I would have welcomed something like this. I jumped right into a situation where an editor 'looked good', 'smelled good', was 'established', and knew all the right things to say an do.. and used virtually all the tactics that this seeks to mitigate. I, too, think it provides a much needed way to fill an existing gap, where most admins 'assume good faith' and being forced to repeatedly prove otherwise is a very arduous task to hang on the necks of new editors (who are often simply accused of complaining about an established editor with 15,000 edits). Peace.Lsi john 22:59, 13 July 2007 (UTC)

Quick responses to the above:

• Self-selection/supervisor matching to articles/supervisor neutrality is an important question. It doesn't affect the core issue "is this a useful process", but it's a part of that process that needs discussion. I've opened a new section for this, below, with some ideas.
• Supervisor guidance would be provided, similar to deletion debate guidance. This would cover issues such as "don't WP:BITE (which should be obvious anyhow)
• Supervisor nominations, I've opened a subpage: /Article supervisor nominations with some ideas. It's in need of collaboration and editing, but the principles seem workable.
• Bans, blocks, leeway, and so on - most supervisors are experienced sysops, with access to reasonable discretion on what's appropriate, who the community feels are trusted to exercise that role well. A quick and clear means of review needs to be provided to disputants (similar to Wikipedia:Deletion review) who feel that this has been improperly used. There'd also be some sensible limit on escalation of bans and blocks, with recourse to Arbcom if more is felt appropriate. Again, this is part of supervisor guidance to be discussed further.
• The proposal already allows for multiple articles to be covered by one supervision. It also allows further articles to be requested to be added, and for editors within supervision to have their conduct elsewhere taken into account or held to the same standard (thus blocking the loophole whereby a bad actor could otherwise just start up games elsewhere).

FT2 ^{(Talk | email)} 11:46, 19 July 2007 (UTC)

Good idea, but I don't think it's worth creating a separate "election/RFA" process to create supervisors. Just have admins who believe they can handle it join a Supervision WikiProject, slap a template on their userpage, or something. That's the way Mediation works, let them volunteer. If you are too worried about certain controversial admins calling themselves Supervisors and screwing up, you can try to pick one or a few highly respected person to hand out the Supervisor badge; that's the way Clerks work. But I wouldn't worry about it, most admins know whether or not they would have the temperament to work as Supervisors. --AnonEMouse ^(squeak) 05:01, 20 July 2007 (UTC)

You reckon? I'd be glad of an easy version. My main concern is (like you can probably tell) that these will in a way be handling the really bad situations. True, they don't have to adjudicate or mediate or arbitrate, they just have to watch for policy violations, but it still needs good judgement and fine balance, and many admins may not have the temperament for it. Do you really think that a system of self nomination and "supervisor approval" will work? It won't end up as some clique of self-selected elite minded aggrandizers or anything childish? Standards'll stay high and visibly neutral and utmost integrity? Self-selection woouldn't be seen as nepotism? That's the sort of thing that was bothering me. If consensus did feel "don't worry" it'd be a relief :) Further thoughts & reasoning? FT2 ^{(Talk | email)} 05:57, 20 July 2007 (UTC)

So long as there's an eye kept on it somehow, I don't think it would be problematic. I think AnonEMouse is generally right in that admins generally don't do that, but even one unfortunate exception could create a lot of trouble. I don't think we need anything as dramatic or formal as RfA or even the Medcom selection process, but supervisors would have some more responsibility than a Medcab mediator—those don't block anyone in the course of what they're doing. Maybe have the supervisors occasionally "look in" on one another, to make sure nothing's being done wrong and offer each other suggestions? Seraphimblade ^{Talk to me} 06:13, 20 July 2007 (UTC)

My thoughts, which I've removed and snipped, in light of the following compromise idea that actually might retain by far the best of both worlds - keep substandard supervisors out (my concern); and also keep it simpler and lighter (AnonEMouse's idea).... FT2 ^{(Talk | email)} 11:48, 20 July 2007 (UTC)

Try this hybrid idea that just occurred to me: supervisors are appointed more readily and easily (Mouse's suggestion), but the initial appointment is always provisional, for a 3 month period (min 3 supervisions), after which there is a review (a quick and simple one, similar to deletion review) if anyone has objections to making the role permanent. As a safety net they can be demoted or warned by supervisors with a sustained (unremedied) concern in that period.

Hybrid proposal

Any experienced editor meeting the standard for supervisorship may nominate themselves (or be nominated). Provisional supervisorship (usually for 3 months/3 full supervisions) is agreed by supervisor consensus. Provisional supervisors actions will be overseen (without intervention) by another supervisor who is there to help, 'double check' and coach them.   Provisional supervisors may be subject to recall, warning or demotion (by consensus of supervisors and without blemish of character) for unaddressed concerns or if it is deemed unlikely they will meet appropriate standards of skill, judgement, community confidence and effectiveness. Supervision requires a considerable balance of skill, experience and attitude, and no adverse inference should be drawn if this happens. Coaching or advice will often be available if requested, and reapplication is encouraged at any future time when concerns have been addressed. At the conclusion of provisional supervision, any editor in good standing may offer good-faith comments or objections. Objections will carry more weight if supported by evidence. Full supervisorship is granted (after consideration of any objections) by consensus of supervisors, or (if no substantive good-faith objections) after a period of 5 days, in either case confirmed by a bureaucrat.

I think the advantages are obvious - it keeps the formalities right down by removing the intense fact finding and debating up front; it keeps people on their toes for several supervision cases which is actually more useful, and by then they're either doing well (and everyone knows it) or causing concerns (and people will raise specific blocks or actions rather than general discussion). The presence of such a review as a possibility, and the knowledge their initial supervisorship is provisional only, will act as a powerful incentive to do it well, a disincentive to gamers, and by 3 months/3 cases it'll probably be habitual if they're genuinely up to it. Maybe even combine SeraphimBlade's idea and use that time interval to have another supervisor look in occasionally too. Thoughts? FT2 ^{(Talk | email)} 11:48, 20 July 2007 (UTC)

I like that! Not totally unrestricted, but not overly bureaucratic either. (The only thing is, if you want crats to confirm the promotion, well, you'll kind of need the crats to agree to do that!) Seraphimblade ^{Talk to me} 03:14, 24 July 2007 (UTC)

In general, I want to say, that I think that this proposal would make a tremendous addition to WP:DR. I have been in situations where a clique of editors have stonewalled all of my other DR efforts including ignoring the advice given by WP:3PO, refusing to Mediate, and resorting to gross incivility and gaming tactics. At the time, I was inquiring if something like "Article supervision" even existed. I am glad to know that it soon very well may exist. I think this would be most helpful! -- Levine2112 ^discuss 22:04, 30 July 2007 (UTC)

It would also help for the supervisor of moderated areas to be named at the start of an article, with the suggestion that proposed changes be discussed with them first.

There may also be a case for a panel of experts to act as supervisor, particularly in areas where no one person has an exhaustive knowledge of the subject. It should be remembered that some subjects are by definition dogmatic, and often extremely subjective: we should ensure that provisions for NPOV are particularly respected in such areas. There may be a case for such a panel to act as peer adjudicators in case of a dispute: if so, easy access should also be provided.

Another aspect is that Wiki supervisors are rarely the world's leading experts on their subject: a supervisor may need to back down before an editor backed by an external expert. Similarly, there may be a case for a supervisor to be superceded were someone with greater knowledge to appear: the quoted instance of someone with 15000 edits simply proves they are generalists, not specialists, and the very number of edits probably disqualifies them from supervision of detailed areas: they may be better suited for higher levels of Admin.

I also think there should be an added onus on any editor, to act as a mentor in answering questions on their contribution, particularly where further in-depth knowledge is sought. The supervisor should, of course, be able to cover his subject fairly exhaustively!

Might I also express my thinking that some subjects may need a tyro version and a specialist version: it might also be useful to have the resident Admin guru named as Mentor on the page, to answer questions which go beyond the scope of the article.

My personal qualification: I am an advisor to a Belgian Supreme Court judge, and employed by a top European international diplomat for my particular research skills. I am in particular the originator of the Franco-German European history project, whereby the chauvinistic aspects of history are balanced by stating both sides of the story. Jel 07:46, 15 September 2007 (UTC)

[edit] Supervision acceptance process

This section is for debate and discussion on:

1. How do people propose articles they think need supervision?
2. Who discusses these, and how is a decision made?
3. How do potential article supervisors become matched/allocated to articles?
4. How are genuine issues with a supervisor (if any) identified, and what happens if past issues exist?
5. How many supervisor(s) and do they all have the same role?

My tentative thoughts on these:

• 1. If an article is suitable for supervision then this should be evident even from a brief individual examination of the article, the discussion page, and the requestors (and others') short statements.
• 2. There are role models in other processes. Actually this one is is much simpler than some dispute processes since all that's needed is statements by any editor(s) why supervision is/isn't appropriate, and consensus by supervisors whether it is a genuinely appropriate case. Anyone who has a view that the dispute is or isn't suitable for supervision can make such a statement. Supervision agreement doesn't require agreement of "blame", just supervisor consensus whether (given the state of editorial discussion and statements submitted) supervision would seem appropriate.
• 3. Seraphimblade raises a concern people will choose articles they have a preconception or bias. I would hope we can keep supervision to people who generally don't act with bias (and remove those who do on review). I am open minded on this one; similar to mediation, a supervisor who is deemed to have acted with bias, wouldn't survive as a supervisor; that's a pretty powerful disincentive. Maybe any supervisor without prior COI can show an interest in any case, but within those it's "longest waiting is allocated" (to remove the element of free choice)? Maybe there's a section "editors offering to supervize" including space for "objections"?

(Note: Specific, reasonable objections on the basis of previous involvement or bias (with the article, subject area or editors) should always be taken seriously, as with mediation and arbitration. Objections which are primarily obstructive need not be.)

• 4. Since we're expecting bad faith and smearing in some of these disputes, one must expect the occasional unfounded or bad faith objections to supervisor appointments and actions, too. That's partly why a high standard is necessary. I think the attitude is, that supervisors are trusted to be neutral, until evidenced otherwise. Well founded objections can be heard, but it's down to the supervisor concerned to comment and/or offer to recuse if the case isn't frivolous. Again, a supervisor who should recuse but doesn't probably won't last long.
• 5. Many thoughts, leaving this one for others to consider. Heres a couple of ideas that seem to count, for me: 1/ One supervisor alone is more than able to supervise an article or dispute. 2/ Multiple supervisors provide cover for bad editing and a chance to bounce off each other. 3/ Multiple supervisors, one of whom is actively involved, the other watches and supervises their supervision, would be an extra way to ensure neutrality if neutrality was ever a problem (a double check on everything).

So really my view is, make supervision accessible to users, but set the standard for consistent supervisor conduct quite high. If the standard is high, there'll still be many who make it, but it will pay off well in antagonistic disputes. FT2 ^{(Talk | email)} 11:20, 19 July 2007 (UTC)

I certainly can see your point. I would hate to see people able to go "Hey, X reverted vandalism one time six months ago to the article (s)he's supervising, so (s)he's an involved party! Not fair!". I think, so long as those who are eligible to supervise in the first place are carefully screened and closely monitored, most of the rest will become mainly a nonissue. Seraphimblade ^{Talk to me} 17:02, 19 July 2007 (UTC)

I think there should also be some forum/noticeboard for editors working in a supervised article to complain about any possible COI or unfairness in terms of the assigned Supervisor. This way, any grievances have a proper place and are keep off of the volatile article talk page or the Supervisor's talk page. Otherwise, I fear that this may lead to incivility directed toward the Supervisor and thus adding even more incivility/hostility in the already charged environment. -- Levine2112 ^discuss 22:01, 30 July 2007 (UTC)

There's likely already to be a subpage per supervision, containing stuff like requests, and a note of agreement and actions related to that supervision. The main options for review of actions by supervisors are, all review requests on one page, or, all matters related to a supervision subpaged under that supervision. Thoughts? Review/grievances is probably an aspect that needs discussing -- if only to ensure it's designed to help rapidly distinguishes between baseless complaints by thwarted POV warriors vs. otherwise and reach rapid fair conclusions. Maybe set up a new section on this page to look at it? FT2 ^{(Talk | email)} 22:39, 30 July 2007 (UTC)

I think a section for grievances is a very good idea. If nothing else, if some POV warrior files a baseless grievance over a perfectly valid action and several uninvolved parties quickly tell them so, it will probably get them to get over it and move on rather than screaming on the talk page or ANI for days on end. (And of course, if a supervisor really is acting inappropriately, that's something that certainly needs taken care of too.) Seraphimblade ^{Talk to me} 23:05, 30 July 2007 (UTC)

Definitely, both ways need cutting. New talk page section for discussing it? FT2 ^{(Talk | email)} 23:25, 30 July 2007 (UTC)

[edit] Supervisors' guide

Although brief, there probably should be a one-page guide, a bit like the administrators' guide to the deletion process. This section is for discussion and points that might be worth putting in it.

Attitude

• Supervision is as much about supporting and guiding enjoyable collaborative good quality editing, as about rectifying and removing editors of problematic editing.
• Utmost neutrality and good conduct should be the norm, and seen to be the norm.

Handling dubious conduct

• Excessive patience (especially of gamey type activity) is not needed, but if there is doubt whether good faith is being shown or not, impatience is usually inappropriate. Some people struggle to "get" policy for a while. But ultimately, refusal to get the point can be sufficiently unhelpful even if the intent is good. Plus, it's a known "game" some bad faith actors play well.

COI

• Prior involvement as a collaborator on the article, or a 'history' with editors who feature in the dispute, can give rise to COI concerns. This should not be taken to extremes - a few conflict-free edits 3 months ago or having blocked an editor for 3RR last year, is not likely to signify any substantial "involvement".
• If there is a COI issue, then handling could be anything from recusal, through to asking other eyeballs to review matters and proposed sanctions for actioning. Supervisors are expected to ensure they meet and exceed communal expectations of neutrality and COI avoidance, and to visibly check where necessary, and ensure their documenting of actions is transparent to uninvolved others.
• That said, obviously reasonable actions and neutral handling of the supervision should not be disrupted or inhibited due to questionable or doubtful claims of COI. For example, an editor who repeats a personal attack that is an obvious and repeated breach of policy, should not be able to wikilawyer that they are "in dispute" with the supervisor who therefore is in breach of COI for addressing it. Genuinely dubious actions can and should be reviewed (see next) but obviously evidenced GAMEy, POINTy and spurious ones need not be taken as seriously unless independent voices also concur.
• Contributions should aim to further the dispute resolution goals of supervision and encourage policy-based editing, explain the views of policy, address problematic conduct or dubious stances, and help stalled debate move forward. Supervisors should not make controversial edits or engage overly in editing, style and content issues; they should suggest improvements rather than dominate. Their role is to empower good-faith editors and as part of that, at times to suggest improved approaches, but never to replace them.

Grievances, accusations, requests for review,

• How to complaint, and what is/is not going to be considered.
• For supervisors, how to handle accusations of abuse, bias, wrong/unfair actions, misunderstandings, etc. What to do, what's expected.
• The process and criteria for rapidly obtaining a view whether there is a basis or not, should it be taken further or just summarily ruled, etc.

Sanctions

• Blocks placed under supervision should be marked with "ASU" at the start of the narrative, to make clear the block was given under the article supervision. This is to help others WP:AGF in future if it's an isolated instance -- blocks may be given under supervision for borderline conduct issues that might have escaped notice in other circumstances, and that's worth noting on the block narrative for fairness' sake.
• Possible block/ban length guide TO DISCUSS: - It shouldn't be necessary to make the escalating short term article bans or temp blocks longer than a few days. 24 hours for the first few offences (rising to 72 if needed) should be the norm, or perhaps even 12 hours for minor infractions where bad faith is dubious but unhelpfulness needs "heading off". Only if it's clear there is a seriously problematic editor, where the motive is to give the article and other editors a break from them as well as deter their future conduct, should longer be considered. (Serious disruption is probably handled in line with normal approaches -- blocks of a week to a month are not uncommon for serious serial offenders in the usual course of adminship, even without supervision.)

Notices and keeping informed

• Supervision is meant to be strict on borderline and questionable editing activity - it's a semi-final measure for the tough disputes to avoid Arbcom. So when an article enters supervision users should probably be asked to read the policy and guide, and be told what's looked for in that article, so they know what is in order and what isn't, and check if unsure.
• Users who believe they may fall foul of supervision, and would like to be sure they do not, should talk with the supervisor to discuss their concerns. However for the sake of neutrality it is important to note that ultimately, the same (minimal above-borderline) editing standard is needed from all, indifferent as to reasons for non-compliance -- this being the basis of article supervision. Non-confrontation, collaboration, referring thoughts via the supervisor, are easy options if needed.

others? FT2 ^{(Talk | email)} 13:19, 24 July 2007 (UTC)

[edit] Example of where this might be useful

I kind of stumbled across a pretty heated debate at Template talk:Infobox NFLactive. I've suggested mediation to the parties, but one editor is not willing to participate [1], and neither seems willing to back down even a bit. The situation, unfortunately, seems to be headed rather needlessly for RfC and arbitration, as several attempts at intervention have met with no success whatsoever. I think this is a perfect example of where a mechanism like this would be useful—something that needs non-voluntary intervention, isn't going to get solved by just blocking someone, but at the same time probably doesn't need a full ArbCom case either. Seraphimblade ^{Talk to me} 23:12, 30 July 2007 (UTC)

[edit] Supervising the supervisors (grievance procedure)

I think starting a new section for grievance procedures is probably a good idea. I don't think it probably needs to be anything too formal, as of course the actions of anyone are always open to community and ArbCom review. I think, however, it might be a good idea to establish some contacts which an editor with a grievance could contact by email, since many editors (rightly or wrongly) might fear retaliation if they make a public complaint. We also could establish a subpage for editors who wish to complain on-wiki, and ask for a peer review of sorts. If the grievance raises serious doubts about the ability of the supervisor to fairly adjudicate, (s)he could be replaced by someone else; if not, everyone can simply be told to move right along. Seraphimblade ^{Talk to me} 23:53, 30 July 2007 (UTC)

I agree. I think if we were to set up an Article Supervision Noticeboard (AS/N), that one of its possible uses could be for posting grievances concerning a supervisor. Another option could be to set up a wholly unique noticeboard just for grievances about Supervisor. -- Levine2112 ^discuss 00:44, 31 July 2007 (UTC)

Makes sense, and especially the point about email for fear of relatiation. I hadn't thought of that one but it's probably valid (it'll be out there). Some basic thoughts on what it has to achieve (feel free to add/edit/comment):

1. Basic assumption, some supervisors may genuinely act too rough, but often grievances or concerns may well be spurious or the complaint of exactly the bad actors that supervision is intended to thwart.
2. Supervision applies and requires policy-based editing. It's geared to removing some pretty tenacious problems from stonewalled articles, often by bad actors who are skilled at "borderlining" it. So there is a degree of "yes its supposed to be strict" too.

   • When an article enters supervision users should get asked to read the policy and guide, so they know what is in order and what isn't, and check if unsure.

3. Both good and bad actors may legitimately need the ability to ask/double-check privately, with someone they can reasonably believe is neutral and not part of supervision. The reassurance that the opinion is truly neutral is important.
4. Bad actors often try to play the game, so a rapid channeled and very specific "this is what do do if you have a concern, and this is what will be done about it" will help forestall using complaints to play the game or trying to dislodge supervision. We don't want every other spurious complaint to be able to sprawl onto every other admin page out there.
5. Ideally we want a system that has a "pre filter" - have they discussed with the admin concerned, and then, is this good or bad faith (ie has it got any real basis or is it just a bad-faith actor trying escape the decision to supervize). That's first, and fairly quick. If it looks like it could have some basis, then it needs a way to look closer, and maybe more eyeballs.
6. I think maybe, a simple process would be:

   1. Editor gives details and evidence, and asks any admins/supervisors of their choice (free choice or from some list, eg "admins willing to review supervisions") for an indication whether they have a good case that the supervision is improper. This can be done in private by email, on talk pages, and up to (say) three admin/supervisors. Or to the community via 3O/RFC if they so wish.
   2. Admins/supervisors who respond, log the fact an inquiry is made on the supervision subpage, but not the details. For example "Initial inquiry by USER on DATE, result = agree/disagree/maybe/recuse/decline to respond. {reason/comment}. ~~~~". This allows forum shopping to be detected; too many requests and admins will decline to respond.
   3. At least one "agree" is needed to start a review of a supervision or a supervisors actions, modelled on WP:DRV.

7. This process should be mandatory as a first step. If the user is unhappy afterwards, then in any event there's always appeal to ANI or 'arb. But making this a necessary first step means that other DR processes don't get junk complaints by disgruntled bad-faith actors out "forum shopping", and they can see what the results where and review quickly if complaint/rejection was justified.

Anyhow, that's my initial ideas. Comments, alternatives, and review process development? FT2 ^{(Talk | email)} 02:11, 31 July 2007 (UTC)

[edit] What this needs...

In my opinion, is a time limit - for instance, that an editor can be a supervisor of whatever for a maximum of three months. This is to prevent people from abusing it for article ownership. >Radiant< 09:41, 2 August 2007 (UTC)

I don't think that's a bad idea at all. If it lasts that long and has failed to fix the problem, it's probably time for arbitration anyway. Seraphimblade ^{Talk to me} 11:14, 2 August 2007 (UTC)

There's actually a 2 month time limit on supervision duration (not 3) built into the proposal - is that any better? ("Article supervision typically lasts no longer than two months.") If it's not resolved by then it's almost certainly an arbcom matter as SeraphimBlade says, and appropriate to prevent OWNership and ensure detachment, exactly as you reckon.

Also see Q&A comment (4): " .... It's also deliberately quite tightly focussed and limited in duration to around 2 months or under (long enough to get good editing practices back after bad actors have become obvious or mended their ways, short enough to be recognized as a temporary measure)."

The other benefit of a time limit is to add some sense of urgency; 2 months is quite a long time, and giving a time limit not only addresses OWNership but also makes clear it's not a "drag out and stonewall" process. It supports the idea that policy based approaches and decent standard editing are expected to be adopted on the article/s pretty much "starting now", not at some vague time in future. (I was actually considering going for 4 - 6 weeks if I could have been sure it was enough to stabilize and settle things post-edit-war, but I couldn't be sure and 4 weeks might not completely shake off POV warriors with time to settle afterwards, so I figured, these are known to be bad disputes, so "allow 2 months, but often less, pretty much absolute maximum" of tight policy and conduct application. If it isn't solved by then, it's almost certainly either insufficiently tight supervision/unaddressed bad conduct, a genuinely fundamental content/NPOV question being mediated or seeking WP:DR on content, or an arbcom matter.)

Next thoughts? :) FT2 ^{(Talk | email)} 12:32, 2 August 2007 (UTC)

[edit] What supervision is not

Added this small section due to a good point raised off-wiki that (given supervisors role in enforcing policy) there should be a more clear demarcation that they act as supervisors, and not editors.

I'm not 100% sure it's the utter best wording, more eyeballs please on this one to double-check? FT2 ^{(Talk | email)} 12:32, 2 August 2007 (UTC)

Update:

Actually this is a question of balance, eyeballs needed. I have (almost deliberately I guess) left both of these statements in:

On the one hand it states that a degree of involvement is reasonable, essentially since their experience of policy and good practice will be useful (original text):

"They may choose to edit or improve the article and take part in debates, but must take care that any edits are appropriate and reflect a policy based approach, their conduct is always of a high standard, and that their actions as supervisors are strictly based upon conduct and evidence, not upon a preference for a given viewpoint."

On the other hand I also seem to have - somewhat contradictorily - added in parallel, a very tight restriction (following introspection about editors who can both debate with, and block, others), and yet not deleted the above:

"Supervisors comment on the talk page only, and to further the goals of supervision, rather than to engage in content debate. As such, supervisors' primary role is supervisory and mentoring, not editorial." [...] "Contributions should aim to further the dispute resolution goals of supervision, encourage policy-based editing, explain the views of policy, address problematic conduct or dubious stances, and help stalled debate move forward."

There's a balance here, and some ideas would be good. Discussion? FT2 ^{(Talk | email)} 18:44, 2 August 2007 (UTC)

Maybe something that indicates that, while supervisors may edit, the edits should be mainly policy-based (removal or questioning of unsourced material, apparent original research, removal of BLP/nonfree image violations, and the like) or uncontroversial maintenance stuff (formatting/spelling/grammar fixes, vandalism reversion). Whatever they do, they should make sure that their editing does not show partiality or bias toward any given side. If they want to be a participant in the debate, someone else should be supervising. Seraphimblade ^{Talk to me} 21:34, 2 August 2007 (UTC)

I've moved some explanatory wordage to the list of "things to cover in the guide page", replacing it with a one-liner. It was unnecessary detail for the main page. And had a go at getting the above issue into better shape. Its a step, not yet completely dealt with by far, through. So "What supervision is not" has had a bit of a rewrite, with an explanatory addition to the guide notes above. Thoughts? Has this made it better? Can others improve it further? FT2 ^{(Talk | email)} 09:48, 15 August 2007 (UTC)

[edit] I rather like this

The idea is a very good one. We have to be careful it doesn't get too complicated, but this - or something very, very similar - is something Wikipedia needs. Cheers, Moreschi ^Talk 20:30, 12 August 2007 (UTC)

I do too. At first blush, it seems like a very good idea, and certainly one worth trying out. Maybe we should try a "pilot program" if enough people agree, on one topic or small set of troublesome articles, and see what issues arise. I imagine this could be particularly useful in the angry nationalist conflicts in out-of-the-way places, where many admins, or indeed editors, have little actual editorial knowledge, but can clearly see when policy is being trampled. Good work FT2. Antandrus (talk) 21:47, 12 August 2007 (UTC)

I think this is in pretty good shape, perhaps ready to be moved to project space and proposed? Seraphimblade ^{Talk to me} 22:38, 12 August 2007 (UTC)

[edit] I appreciate the thought

but this will just add more red tape without solving anything. The problem is that you will lack manpower of dedicated supervisors. WP:3O and WP:RFC are in place, but they do not work because for truly uninvolved editors, the threshold of acquainting themselves with a subject they know nothing about and have no interest in is too great. My experience with trolls or cranks (e.g. Talk:Rigveda) is that they indulge in slow revert warring, but at some point, they simply cease to present sources backing up their view. Once it becomes clear that their opinion simply isn't defensible in academia, they resort to endless hand-waving or forum-shopping, but the debate is really over. What we need are admins who recognize once the debate is over, and enforce the article version that is in fact backed up by academic sources. This isn't agnostically "neutral", it requires that "supervisors" look at sources, and recognize the status of each side of the "dispute". This is already firmly established in Wikipedia policy (WP:UNDUE, WP:RS), the problem is simply with enforcing it. dab (𒁳) 09:22, 13 August 2007 (UTC)

I don't agree that it adds red tape. An additional dispute process available for those intractable disputes which are messy and entrenched, but might respond to strict policy imposition, is a neat and probably very effective way to address many serious disputes and educate their participants, without escalating to flood Arbcom.

A part of the reason it's helpful is that many admins do not seem comfortable varying between AGF when deserved, and drawing a line rapidly when patience (should) run out. In part that's because there is actually no (or no effective) communal sanction for a tough line on covert bad actors who play the game well and tread the borderline quite cleverly, not uncommonly played out for months at a time.

So it's not just that admins should deal with it; it's that the sanction for admins to deal with it tightly does not formally exist. An admin who did so would probably risk being perceived as over strict or over-intolerant, and moderately fair game for ANI complaints or the like.

The thought here is that a formal way to say "the borderlining that has gone on here must end, good collaborative editorship is now non-negotiable on this matter due to stress and damage" may help. A formal way to alert editors in a debate that the dispute must be brought to a close, that the game-like behaviors are being called now, and that to that end, there is going to be strict imposition 1/ of policy, 2/ of editorial standards, and 3/ of curtailing behavior which is normally fine but in some contexts means a game is being attempted, will help. That is not a normal part of the editing environment, and its deliberate inhibition of edit war behaviors would be perceived as excessively strict for many disputes where the traditional relaxed wiki approach is more beneficial.

However, this approach (or one like it) is probably the only tool that would effectively solve the few really tough cases though. Arbcom is over-used in this context, and shouldn't be used as it is, just because mediation etc has failed. But lacking communal sanction for individual admins to take a stricter line on borderline conduct, that's how it can go, and that's what actually happens (regardless of what "should ideally" happen) due to it. That's the background. Thoughts? FT2 ^{(Talk | email)} 14:33, 13 August 2007 (UTC)

[edit] How about

How about we mainspace this and take it for a test run? >Radiant< 13:18, 15 August 2007 (UTC)

Do you really think it's ready? (honest question)

I had ther thought that we run it by critical individuals, see what negatives are perceived, and discuss to be sure we address them best as possible. Maybe that was an over cautious approach. Maybe it should just be discussed and tried. But we don't know what concerns might be obvious (to some people), and there are bound to be some. Do you think it's in fact tight enough to post to main space? Do we need a list of "things that need deciding or fixing" before opening up the debate to wider input? Is it really as simple and non-CREEP as it can be and do the job?

I have no doubt the overall value of the approach, but whether the small print is up to standard for proposal in mainspace is something I'd like more views/reassurance on. Maybe a short straw poll - how ready is this, what if anything needs doing first, what if any are the likely big concerns and does it address them well? Beyond that, thasnks - the view of confidence is reassuring, Radiant!. :) FT2 ^{(Talk | email)} 14:36, 15 August 2007 (UTC)

Good question. I'll look into it, but give me a few days if you don't mind. >Radiant< 10:42, 16 August 2007 (UTC)

Works well for me; (see below) I'm away myself for a few days too, till Monday. I've dropped you some more thoughts via email if that helps - you may have already got them of course. have a good weekend if I don't get online while away! :) FT2 ^{(Talk | email)} 11:28, 16 August 2007 (UTC)

[edit] suggestion

I like a good deal of this, and think it might soon be ready. It meets a real need, one I've experienced in trying to respond to RfCs on articles. I would however suggest one major change:

I dont think using blocks as part of the process will be beneficial, good faith or bad. I've never know a block to increase compliance. Rather, in bad faith editing, the editors tend to try to push one another into a blockable offense, and settle the question that way. this should be discouraged. The way to get compliance and improvement is immediate and consistent reversion of out-of-poicy edits. DGG (talk) 22:13, 15 August 2007 (UTC)

This works in certain limited areas, but not others. In particular, it's suited to article edits and reversions, where different people add edits and revert edits. But in edit wars, a lot of the time, the real meat of the dispute is on the talk page; editor X takes an unhelpfully narrow view, that effectively stonewalls, or the like. Another editor comments that (some actually neutral) editor seems to be a little biased or unaware of the field... and so on. It's often only 6 or 8 months later that a solid enough case of pov warring can be made and proven to arbcom; even then it's not always easy. By then there are only one or two good-faith editors left.

Genuine out-of-policy article edits and obvious vandalism/attacks are already reverted and dealt with a lot of the time. The kinds of disputes this is intended to handle are often on the talk page discussion, or bad-faith borderlining with the intent of pushing stuff, and careful gaming to keep below the usual DR horizons as long as possible.

The approach you suggest doesn't touch that, and rarely helps such talk page disputes (which can be found in many places) because these aren't article edits and can't be helped by "revert out-of-policy edits" approaches. They are often deliberately tangled up to make it hard to resolve. It's worth looking at the kind of behaviors this policy clamps down on. Reversion approaches would not really help the types of disputes this proposal is intended to address. Most disputes of this kind last months, and finally head for arbcom. By then the damage is long done and long enduring, both to editors and the project. People usually have been driven off, long before. Sadly, there is no apt DR measure less than arbcom at present, which effectively curtails these. See above (Q&A) for comparison with present DR options.

Hope that helps clarify, response and discussion welcomed if you have more thoughts. FT2 ^{(Talk | email)} 23:18, 15 August 2007 (UTC)

have you any evidence that the approach you suggest here would work? It would work in one sense, it would get trouble makers banned, but I don;'t see how it would get articles improved. What I think you are suggesting comes down to, is that all POV-pushers should be removed from editing an article and more neutral editors take over. There will be some disputes about what counts as hopelessly pov-pushing editors, though. I think there might be cases that need this approach, but I dont see how to limit it. I am not prepared to say that any WP editor, myself included, will be truly impartial once started editing articles of this sort, and I fear the consequences will be imposing a point of view. I agree on the need to do something, but i think it should start by reinforcing the more customary methods. For example, how about simply compulsory mediation? DGG (talk) 04:18, 16 August 2007 (UTC)

Good questions, DGG. Long answer coming up:

Whether or not it counts as evidence, a lot of it stems from the several cases I've now seen of this sort of thing. Some background: I spent a year as one of the 'dupes' in this sort of thing, with sock-vandal HeadleyDown. I also last month helped present arbitration evidence on DPeterson, where I wasn't one of the dupes, but others were - again for about 13 months. Having been there once helped me recognize what was going on when I saw it 2nd time around happening to others. In each case, the editors concerned had not realized just how badly they were being played with, and it wasn't obvious (in the 'provable/visible to others' sense) until very late in the 'war' that they were in fact being warred upon. Mediation was tried in both cases, and eventually ended up in a mess of distraction, with the warriors agreeing to it, and then subtly making sure it "just didn't work" - no useful results came of it.

Other pov wars are shades of this - people who have found that stonewalling, sidelining, distracting, slight slanting initially, casting gradual doubt, wearing others down, and gradual growth of pov is effective where outright disruption isn't. The majority of moderate-severe-serious pov wars include a fair number of these, and other matters specifically addressed by the proposal. (In fact, without these sorts of behaviors, edit wars almost can't exist, or are almost all resolvable by good-faith DR). Editor X takes an unhelpfully narrow view or slight stance, that sounds reasonable to an unfamiliar outsider, but effectively stonewalls, or the like. They comment they aren't happy with the other editor's views, and feel the other editors need to be more compliant. Two other editors (often being unprovable and unsuspected socks or meats) present rationales in their own terms agreeing with them, and one of these comments that the neutral editor seems to be a little biased or unaware of the field. A cite is presented selected from some credible source, that seems to suggest this is the case (this is easy since almost every view in any debate has concerns raised by some credible source). The original editor X agrees, adds a second credible cite, and asks the neutral editor to politely to check facts before suggesting things. The pov war gradually develops over a few more weeks. The talk page becomes "heavy", RFC/3O is sought, but the RFC/3O visitors just see a few users commenting that some user is not respecting policy, and a messy discussion with no clear flagrant breach of policy beyond a sense of heavy dispute and accusations. No flagrant (unmissable) conduct issue is visible by incoming editors who haven't been closely involved with the article, so the visitors comment on what should happen and then relatively soon they depart the dispute. It's 6 or 8 months later that a solid enough case of pov warring can be made and proven to arbcom; even then it's not always easy. By then there are only one or two good-faith editors left.

This is more typical of how a bad pov war goes. Entrenched positions where policy is blocked from being established in the dispute, aren't resolved by any of the current methods. The test of this is that rarely have DR processes helped in any case I'm aware of, even when conduct is visibly problematic to participants. Admins who seek to enforce policy on this borderline behavior are seen as being over-quick to act heavily, so the sock-warriors usually are warned but not nearly as often actually sanctioned by any admin. It typically goes through the whole gamut of DR, to arbcom, and because of its nature usually incurs 4 - 12 months of damage before any genuine action is taken. The arbitration itself is a big job too, more work. Often a case is made that these are not pov warriors, but just need mentoring. And so it goes on.

So this proposal was borne of the thought, "many editors have seen such tactics; cases that only stop when there is enough of a mountain of good evidence, months later, to arbcomize. What makes them so hard to prevent and allows them to grow so long, and what would have helped other editors sooner, and reverted the editorial environment to normal without taking a year to do it?"

The answer was simple and easy. A stricter line on borderline questionable conduct, and close enforcement of policy, for just long enough that gamers couldn't sustain their games. A month or two of tight policy enforcement, with all games and game-like behaviors called and addressed, would have broken the pattern much more easily, even in the extreme cases noted above where it was hard to make a case "X is edit warring". Administrators almost never do that in practice (and for good healthy reasons); so there need to be some form of process that allowed the communally sanctioned statement to be made, "This dispute is unacceptable, strict policy and dropping of unhelpful approaches needs to now be followed by all".

So to return to your questions, the aim is not as you suggest, to ban people. Far from it - there's actually no long term (or even medium term) block or ban component suggested, and any short term ones are identified as worth ID'ing as such in the block log to ensure others can AGF and not overly read too much into them. I'd be fine putting a "length of block" limit into it, to clarify that if you feel it wise. Long blocks are neither needed nor intended as a part of the proposal. Short term escalating blocks of 24 hours, rising to 1-2 weeks (at absolute maximum for serious repeated disruption), are all that's needed to make it work. The aim of supervision is not to ban editors, it's to make it hard to effectively warrior the article - to acually disrupt the borderlining that the really problematic edit warriors seem to universally rely upon and use. That is why there is so much emphasis on short term blocks only in it. Someone like DPeterson who unhelpfully stonewalls debate, or uses "smoke" to distract, can be told with authority, "That didn't actually answer a legitimate policy-related concern" -- and if they continue not responding to policy related points, they are removed a short while for disruption of the discussion. They come back, and again, they can choose, work with policy or be short term removed. They can't tie the article up in red herring argument for ages (as such editors invariably do), because the issue isn't arguments, it's conduct - behave well (including answering questions to the point) or have a 24-48 hour break from it. It's amazing how fast going through that a couple of times will educate people who stonewall and refuse to get the point that policy is to be complied with, and how it interferes with the actual flow (and core techniques) of edit warring. It's also amazing how grateful bona fide editors are to have the borderline game player finally required to comply with policy and good conduct. I have tried it to a limited extent, but lacking communal sanction for a tough line I felt obligated to allow huge amounts of AGF even though obviously the gamer was a bad faith actor. None the less, it seemed to have visible potential to help, where mediation and other DR's did not.

The aim of this proposal is to provide a means for the community to say for the more problematic pov wars, "We agree that this dispute is messy and seems 'gamey', and without need to analyze blame or present evidence, we would therefore sanction admins taking a much stricter line than usual on policy and drawing a tight line on all unhelpful or unconstructive conduct for a few weeks, and even slightly game-like tactics, and clean it up." The alternatives, including communal sanctioned mediation, are all analyzed above (Q&A) and all have failed to help such disputes. For one thing, almost all readily accessible DR processes other than arbcom itself, require good faith desire by editors to engage in resolution, but in these cases there is a covert intent to eventually not let resolution go anywhere, but to derail it by "smoke" and mess. This proposed process has huge advantages by comparison. It doesn't need to allocate blame, doesn't need to ask if the case-bringers are good or bad faith editors, victims or perpetrators or an edit war, doesn't ask for evidence of blame and fault, is effective for bona fide editors and bad-faith actors, and works for honest users, warriors, and sock users indifferently and alike. All that is needed is an evaluation, is there an intractable situation, that looks like NPOV isn't having a chance to be established. Is there a problem bad enough that strict imposition of policy and good quality editing should be sanctioned? That's a decision that can be made quite easily in such cases, by the mere reading of the talk pages. Supervision itself is deliberately behavior based, as proposed. Support and assistance is given, edit in an appropriate manner and supervision will help you - whatever 'side' you are on. Try to fake it (ie appearance of good editorship but in fact tendentious or unhelpful), and supervision will temporarily remove you (and then let you back a day or two later to see if you got the idea policy matters) - no matter what 'side' you are on. Very clean in concept, very simple and pretty easy to keep neutral.

A long response, to good questions. I'm sure I haven't addressed them all. Maybe reading what I've added here, you'd comment with your follow-on thoughts? :) FT2 ^{(Talk | email)} 09:34, 16 August 2007 (UTC)

(Note: away the weekend, will read and respond when away or if unable, on return. Just a heads up for your info! - FT2)

[edit] A discussion

The following are some (anonymized and with permission) Q&A's of a dialog with a fairly respected admin, covering what he/she felt were the main thoughts on reading this proposal the first time:

• Q: So supervisors have to be admins? Follow-on question: How would they sanction people who are editing tendentiously if they can not block them?

A: Not necessarily, and thats deliberate - I dont like cliques. But they would need to handle blocks, so I've kind of left the door open - if someone has the skill and experience and isn't an admin, and can show it's not a problem (eg, maybe they'd run it by someone else) then fine. I dont know how one would do that, but the principle and the open door, not to create "yet another cabal" by accident, seems important, even if in practice most supervisors will be admins.

• Q: How is it ensured that the supervisor is neutral?

A: Two ways. The first is the same as all dispute resolution tests - lack of involvement, 3rd party review of grievances, being fleshed out on (this) talk page. Second, because these kinds of dispute are the nastier ones (by definition), the standard is or should be high, which means a record of general repute etc elsewhere. The /Nominations page is fairly specific about what sorts of qualities are useful and what ones arent. Neutrality has been got round on every other dispute resolution process though, so it's clear we have the ability to find neutral folks for a DR case in general.

• Q: What if one group of tendentious editors takes a strong liking to the supervisor and their opponents do the opposite?

A: It shouldnt matter. The process is purely behavior based. If one side are bad actors and hate it, or the other isde are good actors and love it, or both sides are bad actors and the supervision starts by examining one sides case (and the other side love it)... it shouldnt matter. It should be effectively, a clamp down on edit war and problematic conduct, whoever by. Over the short to medium time of supervision, any side playing up will be called on it and expected to improve, any side not playing up and acting reasonably will be fine.

• Q: When are you going to move it to the namespace?

A: That's been discussed on the talk page. There will be objections -- what are they, have we done the best we can. is it ready, will the community like it. So ... after a fair bit more review by those who have been involved, I hope. The wording is too much "one persons work" and that's not okay for me (WP:OWN needs to be visibly applied, even if I think my own work is good quality). Even if everyone else likes it, I'd still like to be collectively sure as a community its good, because many people out there will have concerns, perhaps, and those need anticipating and respecting.

• Q: There've been DR cases where the final decision is 100% one person's work. But seriously, I don't think who wrote what parts is a reason to keep it your userspace longer. If the proposal reads as you think it should, then it's ready to be moved. Doesn't matter that you're the main author.

A: The latest post (short) on the talk page covers that exact question. In short, a review for weaknesses and possible issues

Concerns over the potential for abuse of process were raised:

• Q: First of all, wikilawyering on the talk pages in question about the appropriateness of sanctions applied could become a distraction. Where the focus shifts from editing the article to debating behavioral issues. How would the supervisor keep the discussion on track?

A: On these kinds of disputes, the discussion is already off track as a starting point - badly so. Or else it's probbably not serious enough to merit this process. Smoke, accusations, puppets, and polarization. The kind of article this would apply to, a group of bad actors are keeping it tied up in circles. That's the starting point. I agree, if it was on track then discussion of conduct would take it off track. But this kind of dispute is best imagined as reasonably good faith editors being unable to progress the article due to gaming tactics. The first step is to say "sorrry, thats not okay, this is a valid concern, please answer it", then "Sorry, thats still not a helpful answer", and to do so with authority to clean up the discussion page so that article talk can continue. Supervision won't take a dispute debate off topic, if it's on topic and conduct isn't stalling it then it's not needing supervision.

• Q: Another issue would be the supervisors themselves - are you sure you'd be getting the right people nominating themselves for the post?

A: The right people... we get them on other DR processes. I'm not sure theres an inherent reason to assume we couldnt find admins (say) who could be neutral, experienced in disputes, and seen (by the community) as neutral. But I agree, it is important to do so, and not just pay lip service.

• Q: On the topic of supervisors, I could foresee this position attracting people more interested in adding a notch to their belt, aquiring another position, than actually helping out. You'd need a removals process able to deal not only with abusive supervisors but with incompetent and status-seeking ones.

A: Agreed. The nom process took a lot of thought. You get self-aggrandisment as a motive for RFA, Arbcom, mediation, RfB, most "positions". Its not yet clear if theres a best way to avoid it. But there are editors who genuinely can be neutral in a debate, however tough... those could do it. The removals issue was considered too. I dont know if we have the best answer possible, but the current proposal I like most is based on AnonEMouse's idea.

Questions over neutrality of role-fillers dogged arbcom, administratorship, checkuser policy and several other DR processes at one time or another. In each case it was resolved. If the role needs filling, a way to find people who can do it will happen, it seems.

• Q: Removal is "by consensus discussion and evidence of involved and uninvolved parties (usually at WP:ANI)". How would that work out?
• Q: Follow-up question: I'd think you'd want a specific process for removing supervisors, rather than a discussion on the incidents noticeboard

A: This is part of the "stuff I want more eyeballs on" before mainspacing it, because some bad actors will use anything to create a mess, or smoke, or confuse the issue, theres a definite 2 stage process. First, bad faith grievances need to be checked for, and then good faith ones discussed somewhere. What's ideal is if all grievances must get agreement from at least one admin that there is a basis for a complaint. Any admin, at all, of the complainants choice. if no admin is prepared to say "there is a genuine evidenced basis of bad conduct here", then theres a problem. Something like that maybe - not sure the detail. But something that makes it hard to present a case unless there is *something * to it.

• Q: But why admins?

A: Drama cut-down. We're first checking whether there a non-spurious case to discuss at all. Asking "any user" might include socks or others with bias, then you have "is there consensus" etc. At that stage all thats sought is, "is there a genuine basis to discuss", nothing more. After that then yes, open forum if there is a legitimate (non-spurious) issue. Again, its to cut down on the stuff bad actors do. A common tactic is to try and create a case from nothing, or stall, or cite only marginally relevant diffs, or unbacked allegations. If their grievance cannot find even one admin prepared to say "yes, I second this to be discussed as a genuine grievance and misconduct of the supervision, or bias, or whatever", then its probably not going anywhere. I use the "dedicated POV warrior test" -- would it have stopped [names removed; well known pov warriors]? Ask any user to second the proposed complaint - no. Ask an admin to second the proposed complaint - yes.

• Q: Would one issue that stretches over several articles be handled by one supervisor or several?

A: Yes - one (or 2) supervisors to cover a "case" (of connected articles). ACT and AT had the exact same actors on both sides, all thats needed is to watch conduct on both - the same person/people can do that

• Q: "Repeat cases, where it is established that one group is consistently thwarting the process will be referred directly to arbcom." What if both groups are doing this, and by the time they have all been sanctioned appropriately, there are no editors left?

A: I think thats a real possibility in some cases. If an article is being dictated by two groups of edit warriors, one on each side, there is a long term entrenched position such that the dialog which got articles like "abortion" more neutral is being long term stalled, and no productive editing is happening, what should happen? Historically it gets to arbcom, which sanctions one or both sides anyway. But a better result is, there are often non-POV warring editors watching (eg those who have left due to the stonewalling and attacks). Visibly remove the warriors, sooner, and those will perhaps come back. In any event, the article should not be written by warriors just because nobody else is active on it. By contrast, remove the warriors, someone will probably fix it.

(Clarification point: - Supervision does not inherently provide for long term sanctions. It uses short term sanctions that escalate from 24 hours to (say) around 1-2 weeks maximum only, and its aim is to actively prevent potential or actual edit warriors of any agenda from gripping the subject or dispute for a time. It has the expectation that this allows good editorship to be re-established, and as a useful side-effect it can help highlight and clarify for future, the participants and nature of any games, which is often covered by "smoke" as part of those games. It does not, itself, aim to do more. The sense is that this alone is likely to be able to pull most of its target discussions around.)

• Q: I think you should move it to the namespace and see how it does.

The person I asked stated the responses had been "persuasive" and that their concerns seemed to find answers. What I think this shows is, there is a good case, but the concerns which might be raised need articulating and addressing (and checking if there is a better way to handle any of them). However the core idea is not far off track, and answers to the key concerns raised to date are tending to be seen as reasonable -- or close to it.

If refined, and mainspaced, perhaps a trial might be allowed for a few months. But the need to clarify details and review with diligence, is also obvious, and that's awaiting being done. It's not yet been done at all. The above dialog is reassuring, though. FT2 ^{(Talk | email)} 01:54, 21 August 2007 (UTC)

[edit] Suggesting a step

Adding another step to a process seems on the face of it only to prolong the agony, but I wonder whether the following might actually help cut to the chase: as a topic comes under supervision, the editors involved are asked to begin their work by creating and agreeing upon an outline-- just a list of the sub-topics and points that they feel need to be made. This would include positioning contradictory statements in a planful way, either within a relevant sub-topic, or in two different sections. Editors who could not constructively manage this task would not be permitted to proceed with the article under supervision at that time, although they could come back later and attempt to create an outline if they were ready to do so.

The outline would be a short piece and therefore require less intense or lengthy supervision, and it would provide a criterion for deciding whether supervision of the actual article could be done effectively without absorbing all of someone's energies.

Having an outline available might also act as an invitation for editors who had not been in the fight to come in and do some of the work, thus diluting the effects of the contentious ones.

I don't mean, of course, that the outline as created would have to remain as it was forever-- it could be changed when that seemed desirable-- but it would act to provide some boundaries as well as to test the willingness of editors to act cooperatively.Jean Mercer 18:30, 3 September 2007 (UTC)

I actually tried that in a few cases; it can help. The question is, how directive does one want to be. I'd support an "As one option, supervisors may seek to clarify the article by..." in the guide. Possibly better to not tie hands, generally. But a useful tool. FT2 ^{(Talk | email)} 19:46, 3 September 2007 (UTC)

[edit] actual objectivity

In my experience here, in the outside world, and from what one can see in Citizendium, having an expert does not necessarily assure objectivity. Many will have a distinctive POV. One of the 19th c. EBs had an ed on Shakespeare who did not think he wrote the plays. There will be a great incentive for such people to come here. DGG (talk) 19:42, 15 September 2007 (UTC)