As part of its strategic priorities, the Wikimedia Foundation has been increasingly concerned with ensuring a healthy level of editor retention, including both the recruitment and acclimation of newer editors and the continued activity of veteran ones.
As recently reported in The Signpost, many editors—and particularly many highly experienced editors—indicate that the complexity of rules and processes and the inadequacy of mechanisms to deal with problem editors are factors leading to decreased editor activity. Because the arbitration process impacts both of these areas of concern, improving it to reduce negative impact on the retention of participating editors is an important step towards meeting the strategic goals of the editor retention effort.
Editors are invited to suggest improvements to the arbitration process, with the goal of eliminating or reducing elements that negatively affect editor retention. Kirill [talk] 20:30, 27 May 2012 (UTC)
Editors who wish to suggest an improvement should create a new sub-section for their suggestion at the bottom of the page, using the following boilerplate:
===Suggestion from <User X>=== ====Editors who agree with <User X>'s suggestion==== # ====Discussion about <User X>'s suggestion====
If you wish to suggest multiple unrelated improvements, please create a separate sub-section for each to more easily allow independent discussion of each suggestion.
(broken into three points, for ease of commenting)
1. Arbitration isn't the problem, not really. This is because by the time anything reaches ArbCom, it is already helplessly fucked up. ArbCom only gets things once everything else has failed, and that means it is going to have the most contentious issues, the most disruptive users, and the most heated flame wars on the project.
2. The best thing that ArbCom can do, in my opinion, is to champion effective dispute resolution earlier in the process. I'm not sure how that would work; individual arbs could spend time at the DR forums, but if things break down there, that might force recusals later on.
3. As for changes to the Arbitration process, the only one I really have is that Arbs and clerks need to keep the case talk pages under much tighter control. Too often the parties and their advocates are allowed to clash unchecked on the evidence, workshop, and proposed decision talk pages. I'm not saying that they shouldn't be allowed to talk to one another, but I am saying that if ever there was a place for an absolute zero tolerance policy on personal attacks, it would be those pages. Shut down the fights early, using blocks and interaction bans if needed, because those flame threads cause needless escalation, and the anger carries over far after the cases are closed.
Sven Manguard Wha? 22:46, 27 May 2012 (UTC)
I feel that we had a much better and lively editing atmosphere a few years ago. Why? Actually, the answer is clear. This is not because of revert wars and editorial disputes, however unpleasant they were (when people do it, one of them eventually get tired and switch to doing something else). This is because of administrative sanctions: at least one contributor with whom I positively collaborated was indefinitely banned; a couple of others stopped participating after encounters with administrators, and a lot more, including myself were sanctioned, which did not help anyone of us (or others!) to contribute. I can agree that sanctions were fair, but they did not help anyone. That's the moment of truth... So, at the very least, I think there are two important questions for discussion.
My very best wishes (
talk)
23:26, 27 May 2012 (UTC)
The reduced participation was caused to significant degree by excessive use of administrative sanctions.
There are good and bad sanctions. Good sanctions are those applied to users who do more harm than good to the project. This could be for different reasons: vandalism, repeated copyright violations, lack of competence, nationalist/racist trolling (e.g. User:M.V.E.i.), blatant propaganda ( User:Jacob Peters), etc. Those are usually accounts of SPA type. In all such cases one should issue indefinite site bans, and the sooner the better.
Bad sanctions are excessive sanctions applied to users who do significantly more good than harm to the project. Those are usually users or vested type who contribute positively to content with many thousand edits. They also can exhibit problematic behavior, for example difficulty with communication/some incivility (e.g. User:HanzoHattori), sockpuppetry (e.g. User:Altenmann), improper canvassing (EEML case), civil POV-pushing, edit warring, etc. In all such cases, one should issue only relatively mild first sanctions in the interests of the project: topic ban or 1RR restriction for a few months (no longer than 6 months) or desysopping of administrators. If they did not get it, only then the initial official sanctions by AE, ANI or Arbcom should be gradually made longer up to indefinite (blocks by individual administrators and discussions on AE or ANI which resulted in no sanctions should not count). In addition, the sanctions with regard to productive contributors should never be made by individual administrators, even in the areas of discretionary sanctions. Unfortunately, this is frequently not the case: a productive contributor can be issued a very long editing restriction or indefinitely banned after their first ANI/Arbcom case which resulted in sanctions, as had happened in the cases of HanzoHattori, Altenmann, EEML and many others.
Our practices and polices should be changed to reflect this. My very best wishes ( talk) 11:42, 28 May 2012 (UTC)
Higly experienced editors are the very people who Ought To Know Better. If anything, we should be quicker to stop them, as we know that their flouting of the policies (e.g., in socking) is willful. It's probable that they'd be much easier to stop (a public warning ought to be good enough in most cases), but they have even less excuse for misbehavior than the average person. Also, being prolific means their potential for harm is much greater. If I were to cuss at someone in every 100th edit, I'd make eight or ten people per month think that Wikipedia was a rude, unfriendly place. My very best wishes ( talk · contribs), at the same rate of misbehavior, would have irritated half as many users last month. You should therefore be twice as quick to respond to profanity from me as you would to MVBW, and far quicker to me than to someone who does this once a year. WhatamIdoing ( talk) 21:54, 29 May 2012 (UTC)
Two main items for this:
1) Keep Arbitration processes largely unchanged.
It's one of the few dispute-handling processes in Wikipedia that actually seems to work well for its intended function. While the evidence pages get dramariffic, the drama seems to be well-contained there (an ArbCom case's conclusion generally reduces the amount of drama related to the issue elsewhere on the wiki).
2) Increase the number of arbitrators (and clerks) to handle the load, and start bringing disputes to ArbCom sooner.
Quite a lot of problem-issues and problem-users spend years in DR, hit AN/I numerous times, cause large amounts of grief to large numbers of editors, and otherwise sap the productivity and will-to-live of tangentially-involved editors who would otherwise be productive. For most of these situations, it's pretty clear that the dispute is likely to end up at ArbCom eventually. Encourage bringing it there earlier, instead of jumping through as many hoops first. The wiki will be better for it.
-- Christopher Thomas ( talk) 02:09, 28 May 2012 (UTC)
Agree. The long term abusive people know how to misuse rather than violate policies. Currently the only place with a scope to review this is arbcom.North8000 ( talk) 11:47, 29 May 2012 (UTC)
Disagree. Yes, I can see the logic. There is no way for administrators to resolve any dispute unless people can resolve their disputes themselves. What remains? Shooting the horses by Arbcom and other administrators, and the sooner the better? No, simply do not shoot the horses. My very best wishes ( talk) 05:03, 28 May 2012 (UTC)
Disagree - shooting the horses is indeed resulting in further damage (or simply ongoing damage), while it does not resolve the actual issue that caused it. -- Dirk Beetstra T C 06:26, 28 May 2012 (UTC)
Comment. It could work, but then we need to have a different ArbCom system. What we need is a system in which the actual content disputes (that are usually at the heart of a dispute) can be looked into in more detail, and not just behavioral issues that some editors have. I'll write up a proposal along these lines below. Count Iblis ( talk) 17:09, 28 May 2012 (UTC)
Disagree. We don't need more people in control with bad procedures, we need better procedures that require less people and work quicker and make everybody happier because they aren't stuck in things which are irrelevant to building the encyclopaedia. Dmcq ( talk) 13:53, 29 May 2012 (UTC)
Suggest reading The Mythical Man-Month and reconsidering that suggestion at 2). It simply doesn't work that way. WhatamIdoing ( talk) 21:58, 29 May 2012 (UTC)
isaacl ( talk) 03:02, 28 May 2012 (UTC)
Arb process works pretty well. What's really needed for both editor retention and article quality is to make about 6 policy tune-ups so that they aren't so often used contrary to their purposes. The most frustrating/depressing thing for anybody anywhere who cares is to work where the bad guys win/rule via their cleverness at misusing policies. This is also why all Wikipedia articles on contentious topics are utter failures. But this isn't the place to fix that.
A suggestion is to expand the successful arb process to where it could straighten out destructive/nasty behaviors that escape ANI etc. due to wikilawyering cleverness. A good way to to this but still deal with only vetted situations is to take over RFC/U's (the toothless process) where the complaint has been substantiated. North8000 ( talk) 11:39, 28 May 2012 (UTC)
It's a leap to go from complexity of Wikipedia as a whole affecting retention to infer that complexity of ArbCom has significant impact on editor retention. Very few editors end up at ArbCom and they end up there because the rest of Wikipedia has inconsistent incoherent policies. Clear repeated policy violators will be blocked or banned before they get to ArbCom; likewise editors clearly following policies will never be brought here. The trouble is sometimes no one is quite sure whether they're following policy or not.
It would appear we lose the majority of good editors without ArbCom being a factor at all. There is a great weakness in our current dispute resolution system, in particular, manpower. Too often, disputes aren't addressed because of this shortage of available help from experienced editors to assist, and yes, the guidelines and rules are so plentiful that the two sides of the issue can point to guidelines that support their position, yet both sides be wrong. Or right. As to the manpower shortage, too many content disputes get pushed into places like AN/I, and only after the dispute has escalated to a point that the participants are completely frustrated by the lack of clarity. This is a terrible way to deal with disputes, and is very discouraging to editors as it gives them the impression that the only recourse in a dispute is to get someone blocked. Or to just give up and leave Wikipedia. If we could encourage more experienced editors to take an early role in dispute resolution, ArbCom would have a lighter load, and we would reduce the risk of losing more talent. This may require taking a look at dispute resolution as a whole. We have plenty of places to go for resolution (maybe too many), yet too few willing to help. Dennis Brown - 2¢ © 15:18, 28 May 2012 (UTC)
Factual disagreement. The most destructive editors are the ones who have learned how to do it via mis-using of policies rather than violating them. Those who violate them are already weeded out by the other processes (ANI etc.) North8000 ( talk) 11:52, 29 May 2012 (UTC)
Improve focus: the party bringing the dispute has to clearly state the dispute. While other matters may or may not be introduced to show motivation, for exanple, the scope should not be broadened. So I would see a checklist.
If any of these should be answered "No" the case should be rejected.
Agree. In addition, "the party bringing the dispute" should never be Arbcom itself or someone who is not involved in the dispute but tells: "look, these two guys can not resolve their dispute, so let's "help" by sanctioning one of them or the both". Why? Because these two guys would probably prefer to resolve their differences, rather than go to arbitration. Starting an unnecessary arbitration is a very bad idea, even if Arbcom finally decides to open the case. My very best wishes ( talk) 18:42, 28 May 2012 (UTC)
Yes. Mediation is skipped all the time in accepted cases. 71.212.251.217 ( talk) 04:36, 31 May 2012 (UTC)
I don't know how relevant 3O is, given that 3O is focused on content and ArbCom is focused on non-content. WhatamIdoing ( talk) 22:02, 29 May 2012 (UTC)
We need to have an ArbCom system which is able to study those issues that the editors in the dispute find the most relevant. In many cases these are content issues that neither Admins nor Arbitrators can directly address as they obviously need to be uninvolved. Also, the system has to be open minded about arguments by editors that sticking strictly to certain rules does not lead to a good outcome in some particular case.
What matters is how the editing proceeds, not per se if some rule X is violated by editor Y. If violating rule X is really a problem, that would become apparant also if you forget about rule X, study the problems and then come to the conclusion that someting along the lines of rule X should be necessary in precisely this case. Then because rule X already exist, it confirms the necessity of sticking to that rule. But if invoking rule X (perhaps in some narrow interpretation) is unreasonable and it is actually fuelling the problem, that will be mentioned in the final ruling.
The best way to implement such a system, would be for ArbCom to delegate accepted cases to a jury of experienced editors who remain anonymous. They then have the freedom to look at the problems in all its relevant aspects without any restrictions to keep away from specific content issues. The jury can interview the involved editors, they can read the sources on which edits that are under dispute are based, they can consult with external experts to get a better understanding of the topic area.
This will then give then a feeling of what the relevant issues in the topic area according to the sources are, and if these issues are fairly reflected in the articles with proper weight. Then that implies if some complaints of editors have merit or not. If the articles ar ok. that may be due to an uphill struggle by some editors, and their complaints about that will then be found to have merit. The jury will then compile a report and send that to ArbCom, who will make binding rulings.
Count Iblis ( talk) 18:20, 28 May 2012 (UTC)
In response to Mvbw's comment above, my concern is that any appointed or elected board would have difficulty evaluating disputed content. The people who are very familiar with a given subject are likely already taking part in the dispute; people who aren't, would have difficulty making a reliable, informed decision about that subject. This is also why I'm glad ArbCom doesn't handle content (I trust them to decide behaviour cases fairly, but I certainly wouldn't want them issuing decrees about what the tachyon article should cover or what the proper name of the tau particle is).
We already have a process that makes content decisions in contested areas: RFCs. For controversial topics, these tend to get wide participation, and be closed by a trio of uninvolved administrators rather than any single one. The most recent example I can name is the Muhammad images RFC, which produced about as sane a result as IMO was possible given the degree of controversy over the issue.
I'd be overjoyed if a board existed that I could trust to produce rational decisions about content issues, but given the involvement of expert editors in the debates, I suspect that article RFCs are about as good a process as we're likely to get. -- Christopher Thomas ( talk) 04:54, 29 May 2012 (UTC)
This (the suggestion) is an interesting concept, but over the years the community has been at odds with the notion of a "content committee". I think something is needed, but I'm not quite sure if this is what we're looking for. Steven Zhang Get involved in DR! 11:15, 29 May 2012 (UTC)
I'll try to be brief here. It is a lower level problem, with some limited arbitration concerns.
Stigma and neglect hurt existing editors, our arbitration concerns are for the more established editors (1000+ edits), but we lose most of our editors at the policy and procedure 'warnings' handed out in bulk by regular editors. Sure ArbCom sits atop the pile of actions, but if something goes to ArbCom we generally realize editor retention is not the concern, because most individuals would not stick through the case if they didn't want to contribute to Wikipedia. Every page on Wikipedia matters to someone, they work on it for free, stepping on anyone's toes (no matter who) could push them away long before an Admin or ArbCom ever get to it. Wikipedia's complex system of rules, procedures and arbitration exist for the greater good, but they are not as compatible with our ideals as one would expect. It is the role of the individuals who hold more power then ArbCom when it comes to editor retention and through the processes which dispute resolutions follow. They exhaust the parties and retention by trial of fire would be exceedingly low, the Wikipedia learning curve is steep and unfriendly and often enforced by the unforgiving. It is the community as a whole, rather then ArbCom, which bears the problem with editor retention.
To surmise: Invoke low-level binding dispute resolution (For two editors to avoid each other for a week), dispel the negative appearance of ArbCom in its role and refresh our responses so that the newest and most unfamiliar Wikipedia editors can be assisted rather then warned or punished for their infractions. Let's not make a walled garden. ChrisGualtieri ( talk) 05:13, 29 May 2012 (UTC)
I think a major improvement in the atmosphere could be brought around by having a better way of dealing with content disputes. It is pretty clear there are problems with arbitrating on content disputes so it would be better if the requirement was mostly removed. The problems dealing with these problems lead to ratty behaviour with people provoking civility disputes so as to get some sort of decision. My preferred solution would be to to have RfCs which had a freezing effect for about three months which could be used when an edit felt things were going nowhere and was just generating heat rather than light. When such an RfC is passed the decision would be fixed till the end of the term. Saying it should be changed can be counteracted by pointed to the RfC rather than any arguments of substance. Persistence in asking for or trying to change the decision could be treated as disruptive behaviour.
This would allow for a breather in intractable disputes and allow people to get on with their lives elsewhere in Wikipedia until the term ends. Some editors would be frustrated that they can't get their edit in but things don't have to happen immediately. Overall I think except for the warrior caste we'd be much happier and more productive. Dmcq ( talk) 08:04, 29 May 2012 (UTC)
Delaying the problem doesn't fix it. North8000 ( talk) 11:58, 29 May 2012 (UTC)
RfC/U has pretty much been reduced to a paper trail creating exercise in preparation for some higher tier of the DR process.
And I might point out that creating an rfC takes almost as much work as creating an RfArb, and is almost always typically drama filled by its very nature.
This much work (and stress) should at least possibly have more of a resolution than: "the user says they will try to do better in the future" (especially when that's an easy lie to tell).
No, I don't want lynch mobs to have the ominpotent power over individual editors, but we need to do something to fix this process. - jc37 11:44, 29 May 2012 (UTC)
RFC/U needs teeth. Period. It is farcical how ineffective it is. You reduce ArbCom cases by bolstering RFC/U by providing oversight by either one Arb or one admin from a pool of preselected, rotating admins. It needs to be a "lower court" in some ways, with consequences, simple formality, and the ability to be appealed in AN, similar to how we do DRV. What I see is that it often emboldens editors who are problems, instead of fixing problems. They can smirk, play nice for a few weeks, knowing nothing will come of the process, yet the process is needed to provide this paper trail. RFC/U should be retooled and used where AN/x is improper, the case is more complicated, and the problem isn't a single event but a pattern of overall concern. There should be a limit to the sanctions passed, clearly defined, but it needs "clerks" with teeth, who can choose to get involved (pass sanctions to anyone involved) or choose to not get involved (community handled event, no action required). As it currently stands for dealing with problem users, we have AN/I and we have ArbCom, and nothing effective in between. Dennis Brown - 2¢ © 11:17, 31 May 2012 (UTC)
We need a way for semi-automated editing tool usage to be revokable. If bot owners were doing what I constantly see these tool users doing, they and their bots would likely be blocked. And as these tools provide a way for not only great disruption, but also essentially fait accompli, this is clearly a major problem. - jc37 11:44, 29 May 2012 (UTC)
The need for this is demonstrated at least weekly at AN/I. Having the ability to disable automation of another editor would be an effective tool to deal with problem users before a situation escalated to the point that a block becomes the only remedy. Editor A:-)Brunuś is but one recent example. Dennis Brown - 2¢ © 13:21, 29 May 2012 (UTC)
The Committee should be more proactive, being more liberal about accepting cases which involve minor but egregious aggressiveness and incivility.
The Arbitration Committee should work toward taking effective action when the problem is small. My experience both as an editor and as an arbitrator is that it is very difficult to get anything done with respect to particularly nasty situations when they first start. Eventually, of course, action is taken because the behavior begins to affect many people and many articles.
The downside is that doing heavy things regarding problems when they are small is hard to explain both to other arbitrators and to the community. And, of course, sometimes the problems are small or can be mitigated without getting heavy, and making a issue of something is inappropriate.
As to leaving Wikipedia, I created a fork because of aggressive and incivil behavior by other editors, came back to Wikipedia and helped form the arbitration committee as a way of dealing with aggressive and incivil behavior by other editors and now feel that while a great deal was accomplished the system is ineffective. "Laws, like cobwebs, catch small flies, great ones break through before your eyes," Benjamin Franklin. User:Fred Bauder Talk 13:42, 29 May 2012 (UTC)
I don't know if Arbitration has that much of an impact on editor retention as suggested by ArbCom itself. In my perspective, Arbitration affects a comparatively small number of articles and editors, whereas the rest of the encyclopedia are free to continue along its way. On the other hand, the cases that ArbCom deal with highly controversial topics and highly controversial editors.
That being said, the editing community needs to retain the prerogative – as well as needs to remain proactive – in handling complex situations collectively on their own, without Arbitration whenever they can. I support the status quo in that ArbCom remain as the last resort in dispute resolution, to be employed when all other attempts at community-level resolution have failed. Whenever possible, collaboration and consensus-building, not leadership or its accompanying bureaucracy, need to resolve the many issues that they face.
Lastly, what many people (even me) forget is that dispute resolution is no quick fix. Highly controversial topics are going to require lots of time and legwork to make progress. Simply passing the buck to ArbCom is not going to reduce the amount of time and legwork needed. -- MuZemike 18:42, 29 May 2012 (UTC)
ArbCom isn't the problem. The problem is that all too often, ArbCom is the only solution. The closest other mechanism for fulfilling the most common need need (RFC/U) is toothless. The first sentence was said by Dennis Brown in a discussion above, the most brilliant summary I've seen on this page.
The processes and delays at ArbCom do, indeed, make it seem like it is functioning as a "court" using "evidence" and "findings."
Therefore I suggest:
The aim in all of this is to totally reform ArbCom, and thus to improve editor retention. It will reduce the elements that affect editor retention, while allowing ArbCom to focus on greater matters than whether editor Gnarph has improperly edited on the topic of "NASCAR championship incidents" or the like. Collect ( talk) 13:02, 30 May 2012 (UTC)
Much good work/ideas in there on structure & procedure, but I tend to disagree with the underlying premise that such is the problem. North8000 ( talk) 11:03, 31 May 2012 (UTC)
My personal opinion is that Arbcom is not the disease but a symptom of it. When cases come to Arbcom its usually because its a contentious site wide issue that has gone unresolved in other venues. When dealing with editors, once the case gets here Arbcom historically has the view that the accused must be guilty and tend to favor banning the user or sanctins against them and they are presumed guilty on arrival with the attitude that if they hadn't done anything wrong they wouldn't be here. Additional problems include a general lack of consideration of discussions, the results are inconsistent and frequently poorly and vaguely written, they make unprecedented decisions about how to interpret different policies and perform a variety of other actions that make the community generally uneasy. I would say there is a general feeling of distrust amongst the community towards arbcom for many of the aforementioned reasons. Many of the cases (both editor and non editor related) that come here are very high profile and attract well known and experienced users so the conduct and decisions by the Arbcom are viewed far outside the temple of Arbcom. A poor decision or comment ripples out across the pedia and causes discussion in other venues and editor talk pages. For these reasons I feel that yes Arbcom decisions do affect editor retention because people see the decisions they make and it causes them to leave. Not in droves necessarily but it does affect it. Kumioko ( talk) 14:47, 30 May 2012 (UTC)
Requiring the panel of arbitrators, who are also very active day-to-day admins, to vote on findings and penalties leads to very frequent questions of prior involvement and conflicts of interest, actual and apparent. It results in a star chamber with all of the inherent injustice that entails. Frequently diffs shown in support of findings do not support them.
After the proposed decision is drafted, Randomly select a jury of ordinary editors, without regard to admin status, to vote on findings and penalties, and toss out any penalties which don't match the approved findings. This has been a bedrock principle of Western jurisprudence for centuries, and there is no reason that dispute resolution in Wikipedia should continue to discourage disputants and onlooking editors by flying in the face of it.
71.212.251.217 (
talk)
04:46, 31 May 2012 (UTC)
Yes, but what does it mean to be constructive in collaborative projects? I think it means mostly two things: (a) helping each other; and (b) openly discussing any problems, whatever they might be. Therefore, speaking about the retention of editors and Arbcom, I would suggest the following policy changes to encourage collaboration and communication in the project.
I have little to add. But wanted to lend my weight to some points other people have raised.
That's it. Shooterwalker ( talk) 07:21, 11 June 2012 (UTC)
I think the most common situation that really harms retention is this. Editor A (who could be a newbie or not) makes a change that is not unreasonable [4]. Very soon an experienced editor X comes to blindly revert his edit without explanation [5] instead of making a compromise edit, as he easily can [6] (I am giving the simplest example; note the presence of nationalistic POV, "Jewish colonel in the Soviet army" in last version). Things like that really harm desire of people to contribute. In the more complicated cases, the "reverter" frequently explains his revert like this: "I am not sure that your changes represent an improvement compare to the previous version" (then editor A must provide a long explanation, but he might not have enough time and patience). Why revert if you are not sure or not really familiar with the subject, and why revert if you can easily make a compromise version? Unfortunately, I do not know how to fix it (except not writing wrong essays like WP:BRD that encourage exactly such behavior). I know that Arbcom started punishing people for edit wars more seriously, but this is not a good solution. It actually forces reasonable people (who do not want to be sanctioned) to leave political/history subjects to " tigers" who are willing to take risks and edit war. It only helps that we do not have a lot of articles currently owned by "tigers"... My very best wishes ( talk) 13:14, 12 June 2012 (UTC)
As part of its strategic priorities, the Wikimedia Foundation has been increasingly concerned with ensuring a healthy level of editor retention, including both the recruitment and acclimation of newer editors and the continued activity of veteran ones.
As recently reported in The Signpost, many editors—and particularly many highly experienced editors—indicate that the complexity of rules and processes and the inadequacy of mechanisms to deal with problem editors are factors leading to decreased editor activity. Because the arbitration process impacts both of these areas of concern, improving it to reduce negative impact on the retention of participating editors is an important step towards meeting the strategic goals of the editor retention effort.
Editors are invited to suggest improvements to the arbitration process, with the goal of eliminating or reducing elements that negatively affect editor retention. Kirill [talk] 20:30, 27 May 2012 (UTC)
Editors who wish to suggest an improvement should create a new sub-section for their suggestion at the bottom of the page, using the following boilerplate:
===Suggestion from <User X>=== ====Editors who agree with <User X>'s suggestion==== # ====Discussion about <User X>'s suggestion====
If you wish to suggest multiple unrelated improvements, please create a separate sub-section for each to more easily allow independent discussion of each suggestion.
(broken into three points, for ease of commenting)
1. Arbitration isn't the problem, not really. This is because by the time anything reaches ArbCom, it is already helplessly fucked up. ArbCom only gets things once everything else has failed, and that means it is going to have the most contentious issues, the most disruptive users, and the most heated flame wars on the project.
2. The best thing that ArbCom can do, in my opinion, is to champion effective dispute resolution earlier in the process. I'm not sure how that would work; individual arbs could spend time at the DR forums, but if things break down there, that might force recusals later on.
3. As for changes to the Arbitration process, the only one I really have is that Arbs and clerks need to keep the case talk pages under much tighter control. Too often the parties and their advocates are allowed to clash unchecked on the evidence, workshop, and proposed decision talk pages. I'm not saying that they shouldn't be allowed to talk to one another, but I am saying that if ever there was a place for an absolute zero tolerance policy on personal attacks, it would be those pages. Shut down the fights early, using blocks and interaction bans if needed, because those flame threads cause needless escalation, and the anger carries over far after the cases are closed.
Sven Manguard Wha? 22:46, 27 May 2012 (UTC)
I feel that we had a much better and lively editing atmosphere a few years ago. Why? Actually, the answer is clear. This is not because of revert wars and editorial disputes, however unpleasant they were (when people do it, one of them eventually get tired and switch to doing something else). This is because of administrative sanctions: at least one contributor with whom I positively collaborated was indefinitely banned; a couple of others stopped participating after encounters with administrators, and a lot more, including myself were sanctioned, which did not help anyone of us (or others!) to contribute. I can agree that sanctions were fair, but they did not help anyone. That's the moment of truth... So, at the very least, I think there are two important questions for discussion.
My very best wishes (
talk)
23:26, 27 May 2012 (UTC)
The reduced participation was caused to significant degree by excessive use of administrative sanctions.
There are good and bad sanctions. Good sanctions are those applied to users who do more harm than good to the project. This could be for different reasons: vandalism, repeated copyright violations, lack of competence, nationalist/racist trolling (e.g. User:M.V.E.i.), blatant propaganda ( User:Jacob Peters), etc. Those are usually accounts of SPA type. In all such cases one should issue indefinite site bans, and the sooner the better.
Bad sanctions are excessive sanctions applied to users who do significantly more good than harm to the project. Those are usually users or vested type who contribute positively to content with many thousand edits. They also can exhibit problematic behavior, for example difficulty with communication/some incivility (e.g. User:HanzoHattori), sockpuppetry (e.g. User:Altenmann), improper canvassing (EEML case), civil POV-pushing, edit warring, etc. In all such cases, one should issue only relatively mild first sanctions in the interests of the project: topic ban or 1RR restriction for a few months (no longer than 6 months) or desysopping of administrators. If they did not get it, only then the initial official sanctions by AE, ANI or Arbcom should be gradually made longer up to indefinite (blocks by individual administrators and discussions on AE or ANI which resulted in no sanctions should not count). In addition, the sanctions with regard to productive contributors should never be made by individual administrators, even in the areas of discretionary sanctions. Unfortunately, this is frequently not the case: a productive contributor can be issued a very long editing restriction or indefinitely banned after their first ANI/Arbcom case which resulted in sanctions, as had happened in the cases of HanzoHattori, Altenmann, EEML and many others.
Our practices and polices should be changed to reflect this. My very best wishes ( talk) 11:42, 28 May 2012 (UTC)
Higly experienced editors are the very people who Ought To Know Better. If anything, we should be quicker to stop them, as we know that their flouting of the policies (e.g., in socking) is willful. It's probable that they'd be much easier to stop (a public warning ought to be good enough in most cases), but they have even less excuse for misbehavior than the average person. Also, being prolific means their potential for harm is much greater. If I were to cuss at someone in every 100th edit, I'd make eight or ten people per month think that Wikipedia was a rude, unfriendly place. My very best wishes ( talk · contribs), at the same rate of misbehavior, would have irritated half as many users last month. You should therefore be twice as quick to respond to profanity from me as you would to MVBW, and far quicker to me than to someone who does this once a year. WhatamIdoing ( talk) 21:54, 29 May 2012 (UTC)
Two main items for this:
1) Keep Arbitration processes largely unchanged.
It's one of the few dispute-handling processes in Wikipedia that actually seems to work well for its intended function. While the evidence pages get dramariffic, the drama seems to be well-contained there (an ArbCom case's conclusion generally reduces the amount of drama related to the issue elsewhere on the wiki).
2) Increase the number of arbitrators (and clerks) to handle the load, and start bringing disputes to ArbCom sooner.
Quite a lot of problem-issues and problem-users spend years in DR, hit AN/I numerous times, cause large amounts of grief to large numbers of editors, and otherwise sap the productivity and will-to-live of tangentially-involved editors who would otherwise be productive. For most of these situations, it's pretty clear that the dispute is likely to end up at ArbCom eventually. Encourage bringing it there earlier, instead of jumping through as many hoops first. The wiki will be better for it.
-- Christopher Thomas ( talk) 02:09, 28 May 2012 (UTC)
Agree. The long term abusive people know how to misuse rather than violate policies. Currently the only place with a scope to review this is arbcom.North8000 ( talk) 11:47, 29 May 2012 (UTC)
Disagree. Yes, I can see the logic. There is no way for administrators to resolve any dispute unless people can resolve their disputes themselves. What remains? Shooting the horses by Arbcom and other administrators, and the sooner the better? No, simply do not shoot the horses. My very best wishes ( talk) 05:03, 28 May 2012 (UTC)
Disagree - shooting the horses is indeed resulting in further damage (or simply ongoing damage), while it does not resolve the actual issue that caused it. -- Dirk Beetstra T C 06:26, 28 May 2012 (UTC)
Comment. It could work, but then we need to have a different ArbCom system. What we need is a system in which the actual content disputes (that are usually at the heart of a dispute) can be looked into in more detail, and not just behavioral issues that some editors have. I'll write up a proposal along these lines below. Count Iblis ( talk) 17:09, 28 May 2012 (UTC)
Disagree. We don't need more people in control with bad procedures, we need better procedures that require less people and work quicker and make everybody happier because they aren't stuck in things which are irrelevant to building the encyclopaedia. Dmcq ( talk) 13:53, 29 May 2012 (UTC)
Suggest reading The Mythical Man-Month and reconsidering that suggestion at 2). It simply doesn't work that way. WhatamIdoing ( talk) 21:58, 29 May 2012 (UTC)
isaacl ( talk) 03:02, 28 May 2012 (UTC)
Arb process works pretty well. What's really needed for both editor retention and article quality is to make about 6 policy tune-ups so that they aren't so often used contrary to their purposes. The most frustrating/depressing thing for anybody anywhere who cares is to work where the bad guys win/rule via their cleverness at misusing policies. This is also why all Wikipedia articles on contentious topics are utter failures. But this isn't the place to fix that.
A suggestion is to expand the successful arb process to where it could straighten out destructive/nasty behaviors that escape ANI etc. due to wikilawyering cleverness. A good way to to this but still deal with only vetted situations is to take over RFC/U's (the toothless process) where the complaint has been substantiated. North8000 ( talk) 11:39, 28 May 2012 (UTC)
It's a leap to go from complexity of Wikipedia as a whole affecting retention to infer that complexity of ArbCom has significant impact on editor retention. Very few editors end up at ArbCom and they end up there because the rest of Wikipedia has inconsistent incoherent policies. Clear repeated policy violators will be blocked or banned before they get to ArbCom; likewise editors clearly following policies will never be brought here. The trouble is sometimes no one is quite sure whether they're following policy or not.
It would appear we lose the majority of good editors without ArbCom being a factor at all. There is a great weakness in our current dispute resolution system, in particular, manpower. Too often, disputes aren't addressed because of this shortage of available help from experienced editors to assist, and yes, the guidelines and rules are so plentiful that the two sides of the issue can point to guidelines that support their position, yet both sides be wrong. Or right. As to the manpower shortage, too many content disputes get pushed into places like AN/I, and only after the dispute has escalated to a point that the participants are completely frustrated by the lack of clarity. This is a terrible way to deal with disputes, and is very discouraging to editors as it gives them the impression that the only recourse in a dispute is to get someone blocked. Or to just give up and leave Wikipedia. If we could encourage more experienced editors to take an early role in dispute resolution, ArbCom would have a lighter load, and we would reduce the risk of losing more talent. This may require taking a look at dispute resolution as a whole. We have plenty of places to go for resolution (maybe too many), yet too few willing to help. Dennis Brown - 2¢ © 15:18, 28 May 2012 (UTC)
Factual disagreement. The most destructive editors are the ones who have learned how to do it via mis-using of policies rather than violating them. Those who violate them are already weeded out by the other processes (ANI etc.) North8000 ( talk) 11:52, 29 May 2012 (UTC)
Improve focus: the party bringing the dispute has to clearly state the dispute. While other matters may or may not be introduced to show motivation, for exanple, the scope should not be broadened. So I would see a checklist.
If any of these should be answered "No" the case should be rejected.
Agree. In addition, "the party bringing the dispute" should never be Arbcom itself or someone who is not involved in the dispute but tells: "look, these two guys can not resolve their dispute, so let's "help" by sanctioning one of them or the both". Why? Because these two guys would probably prefer to resolve their differences, rather than go to arbitration. Starting an unnecessary arbitration is a very bad idea, even if Arbcom finally decides to open the case. My very best wishes ( talk) 18:42, 28 May 2012 (UTC)
Yes. Mediation is skipped all the time in accepted cases. 71.212.251.217 ( talk) 04:36, 31 May 2012 (UTC)
I don't know how relevant 3O is, given that 3O is focused on content and ArbCom is focused on non-content. WhatamIdoing ( talk) 22:02, 29 May 2012 (UTC)
We need to have an ArbCom system which is able to study those issues that the editors in the dispute find the most relevant. In many cases these are content issues that neither Admins nor Arbitrators can directly address as they obviously need to be uninvolved. Also, the system has to be open minded about arguments by editors that sticking strictly to certain rules does not lead to a good outcome in some particular case.
What matters is how the editing proceeds, not per se if some rule X is violated by editor Y. If violating rule X is really a problem, that would become apparant also if you forget about rule X, study the problems and then come to the conclusion that someting along the lines of rule X should be necessary in precisely this case. Then because rule X already exist, it confirms the necessity of sticking to that rule. But if invoking rule X (perhaps in some narrow interpretation) is unreasonable and it is actually fuelling the problem, that will be mentioned in the final ruling.
The best way to implement such a system, would be for ArbCom to delegate accepted cases to a jury of experienced editors who remain anonymous. They then have the freedom to look at the problems in all its relevant aspects without any restrictions to keep away from specific content issues. The jury can interview the involved editors, they can read the sources on which edits that are under dispute are based, they can consult with external experts to get a better understanding of the topic area.
This will then give then a feeling of what the relevant issues in the topic area according to the sources are, and if these issues are fairly reflected in the articles with proper weight. Then that implies if some complaints of editors have merit or not. If the articles ar ok. that may be due to an uphill struggle by some editors, and their complaints about that will then be found to have merit. The jury will then compile a report and send that to ArbCom, who will make binding rulings.
Count Iblis ( talk) 18:20, 28 May 2012 (UTC)
In response to Mvbw's comment above, my concern is that any appointed or elected board would have difficulty evaluating disputed content. The people who are very familiar with a given subject are likely already taking part in the dispute; people who aren't, would have difficulty making a reliable, informed decision about that subject. This is also why I'm glad ArbCom doesn't handle content (I trust them to decide behaviour cases fairly, but I certainly wouldn't want them issuing decrees about what the tachyon article should cover or what the proper name of the tau particle is).
We already have a process that makes content decisions in contested areas: RFCs. For controversial topics, these tend to get wide participation, and be closed by a trio of uninvolved administrators rather than any single one. The most recent example I can name is the Muhammad images RFC, which produced about as sane a result as IMO was possible given the degree of controversy over the issue.
I'd be overjoyed if a board existed that I could trust to produce rational decisions about content issues, but given the involvement of expert editors in the debates, I suspect that article RFCs are about as good a process as we're likely to get. -- Christopher Thomas ( talk) 04:54, 29 May 2012 (UTC)
This (the suggestion) is an interesting concept, but over the years the community has been at odds with the notion of a "content committee". I think something is needed, but I'm not quite sure if this is what we're looking for. Steven Zhang Get involved in DR! 11:15, 29 May 2012 (UTC)
I'll try to be brief here. It is a lower level problem, with some limited arbitration concerns.
Stigma and neglect hurt existing editors, our arbitration concerns are for the more established editors (1000+ edits), but we lose most of our editors at the policy and procedure 'warnings' handed out in bulk by regular editors. Sure ArbCom sits atop the pile of actions, but if something goes to ArbCom we generally realize editor retention is not the concern, because most individuals would not stick through the case if they didn't want to contribute to Wikipedia. Every page on Wikipedia matters to someone, they work on it for free, stepping on anyone's toes (no matter who) could push them away long before an Admin or ArbCom ever get to it. Wikipedia's complex system of rules, procedures and arbitration exist for the greater good, but they are not as compatible with our ideals as one would expect. It is the role of the individuals who hold more power then ArbCom when it comes to editor retention and through the processes which dispute resolutions follow. They exhaust the parties and retention by trial of fire would be exceedingly low, the Wikipedia learning curve is steep and unfriendly and often enforced by the unforgiving. It is the community as a whole, rather then ArbCom, which bears the problem with editor retention.
To surmise: Invoke low-level binding dispute resolution (For two editors to avoid each other for a week), dispel the negative appearance of ArbCom in its role and refresh our responses so that the newest and most unfamiliar Wikipedia editors can be assisted rather then warned or punished for their infractions. Let's not make a walled garden. ChrisGualtieri ( talk) 05:13, 29 May 2012 (UTC)
I think a major improvement in the atmosphere could be brought around by having a better way of dealing with content disputes. It is pretty clear there are problems with arbitrating on content disputes so it would be better if the requirement was mostly removed. The problems dealing with these problems lead to ratty behaviour with people provoking civility disputes so as to get some sort of decision. My preferred solution would be to to have RfCs which had a freezing effect for about three months which could be used when an edit felt things were going nowhere and was just generating heat rather than light. When such an RfC is passed the decision would be fixed till the end of the term. Saying it should be changed can be counteracted by pointed to the RfC rather than any arguments of substance. Persistence in asking for or trying to change the decision could be treated as disruptive behaviour.
This would allow for a breather in intractable disputes and allow people to get on with their lives elsewhere in Wikipedia until the term ends. Some editors would be frustrated that they can't get their edit in but things don't have to happen immediately. Overall I think except for the warrior caste we'd be much happier and more productive. Dmcq ( talk) 08:04, 29 May 2012 (UTC)
Delaying the problem doesn't fix it. North8000 ( talk) 11:58, 29 May 2012 (UTC)
RfC/U has pretty much been reduced to a paper trail creating exercise in preparation for some higher tier of the DR process.
And I might point out that creating an rfC takes almost as much work as creating an RfArb, and is almost always typically drama filled by its very nature.
This much work (and stress) should at least possibly have more of a resolution than: "the user says they will try to do better in the future" (especially when that's an easy lie to tell).
No, I don't want lynch mobs to have the ominpotent power over individual editors, but we need to do something to fix this process. - jc37 11:44, 29 May 2012 (UTC)
RFC/U needs teeth. Period. It is farcical how ineffective it is. You reduce ArbCom cases by bolstering RFC/U by providing oversight by either one Arb or one admin from a pool of preselected, rotating admins. It needs to be a "lower court" in some ways, with consequences, simple formality, and the ability to be appealed in AN, similar to how we do DRV. What I see is that it often emboldens editors who are problems, instead of fixing problems. They can smirk, play nice for a few weeks, knowing nothing will come of the process, yet the process is needed to provide this paper trail. RFC/U should be retooled and used where AN/x is improper, the case is more complicated, and the problem isn't a single event but a pattern of overall concern. There should be a limit to the sanctions passed, clearly defined, but it needs "clerks" with teeth, who can choose to get involved (pass sanctions to anyone involved) or choose to not get involved (community handled event, no action required). As it currently stands for dealing with problem users, we have AN/I and we have ArbCom, and nothing effective in between. Dennis Brown - 2¢ © 11:17, 31 May 2012 (UTC)
We need a way for semi-automated editing tool usage to be revokable. If bot owners were doing what I constantly see these tool users doing, they and their bots would likely be blocked. And as these tools provide a way for not only great disruption, but also essentially fait accompli, this is clearly a major problem. - jc37 11:44, 29 May 2012 (UTC)
The need for this is demonstrated at least weekly at AN/I. Having the ability to disable automation of another editor would be an effective tool to deal with problem users before a situation escalated to the point that a block becomes the only remedy. Editor A:-)Brunuś is but one recent example. Dennis Brown - 2¢ © 13:21, 29 May 2012 (UTC)
The Committee should be more proactive, being more liberal about accepting cases which involve minor but egregious aggressiveness and incivility.
The Arbitration Committee should work toward taking effective action when the problem is small. My experience both as an editor and as an arbitrator is that it is very difficult to get anything done with respect to particularly nasty situations when they first start. Eventually, of course, action is taken because the behavior begins to affect many people and many articles.
The downside is that doing heavy things regarding problems when they are small is hard to explain both to other arbitrators and to the community. And, of course, sometimes the problems are small or can be mitigated without getting heavy, and making a issue of something is inappropriate.
As to leaving Wikipedia, I created a fork because of aggressive and incivil behavior by other editors, came back to Wikipedia and helped form the arbitration committee as a way of dealing with aggressive and incivil behavior by other editors and now feel that while a great deal was accomplished the system is ineffective. "Laws, like cobwebs, catch small flies, great ones break through before your eyes," Benjamin Franklin. User:Fred Bauder Talk 13:42, 29 May 2012 (UTC)
I don't know if Arbitration has that much of an impact on editor retention as suggested by ArbCom itself. In my perspective, Arbitration affects a comparatively small number of articles and editors, whereas the rest of the encyclopedia are free to continue along its way. On the other hand, the cases that ArbCom deal with highly controversial topics and highly controversial editors.
That being said, the editing community needs to retain the prerogative – as well as needs to remain proactive – in handling complex situations collectively on their own, without Arbitration whenever they can. I support the status quo in that ArbCom remain as the last resort in dispute resolution, to be employed when all other attempts at community-level resolution have failed. Whenever possible, collaboration and consensus-building, not leadership or its accompanying bureaucracy, need to resolve the many issues that they face.
Lastly, what many people (even me) forget is that dispute resolution is no quick fix. Highly controversial topics are going to require lots of time and legwork to make progress. Simply passing the buck to ArbCom is not going to reduce the amount of time and legwork needed. -- MuZemike 18:42, 29 May 2012 (UTC)
ArbCom isn't the problem. The problem is that all too often, ArbCom is the only solution. The closest other mechanism for fulfilling the most common need need (RFC/U) is toothless. The first sentence was said by Dennis Brown in a discussion above, the most brilliant summary I've seen on this page.
The processes and delays at ArbCom do, indeed, make it seem like it is functioning as a "court" using "evidence" and "findings."
Therefore I suggest:
The aim in all of this is to totally reform ArbCom, and thus to improve editor retention. It will reduce the elements that affect editor retention, while allowing ArbCom to focus on greater matters than whether editor Gnarph has improperly edited on the topic of "NASCAR championship incidents" or the like. Collect ( talk) 13:02, 30 May 2012 (UTC)
Much good work/ideas in there on structure & procedure, but I tend to disagree with the underlying premise that such is the problem. North8000 ( talk) 11:03, 31 May 2012 (UTC)
My personal opinion is that Arbcom is not the disease but a symptom of it. When cases come to Arbcom its usually because its a contentious site wide issue that has gone unresolved in other venues. When dealing with editors, once the case gets here Arbcom historically has the view that the accused must be guilty and tend to favor banning the user or sanctins against them and they are presumed guilty on arrival with the attitude that if they hadn't done anything wrong they wouldn't be here. Additional problems include a general lack of consideration of discussions, the results are inconsistent and frequently poorly and vaguely written, they make unprecedented decisions about how to interpret different policies and perform a variety of other actions that make the community generally uneasy. I would say there is a general feeling of distrust amongst the community towards arbcom for many of the aforementioned reasons. Many of the cases (both editor and non editor related) that come here are very high profile and attract well known and experienced users so the conduct and decisions by the Arbcom are viewed far outside the temple of Arbcom. A poor decision or comment ripples out across the pedia and causes discussion in other venues and editor talk pages. For these reasons I feel that yes Arbcom decisions do affect editor retention because people see the decisions they make and it causes them to leave. Not in droves necessarily but it does affect it. Kumioko ( talk) 14:47, 30 May 2012 (UTC)
Requiring the panel of arbitrators, who are also very active day-to-day admins, to vote on findings and penalties leads to very frequent questions of prior involvement and conflicts of interest, actual and apparent. It results in a star chamber with all of the inherent injustice that entails. Frequently diffs shown in support of findings do not support them.
After the proposed decision is drafted, Randomly select a jury of ordinary editors, without regard to admin status, to vote on findings and penalties, and toss out any penalties which don't match the approved findings. This has been a bedrock principle of Western jurisprudence for centuries, and there is no reason that dispute resolution in Wikipedia should continue to discourage disputants and onlooking editors by flying in the face of it.
71.212.251.217 (
talk)
04:46, 31 May 2012 (UTC)
Yes, but what does it mean to be constructive in collaborative projects? I think it means mostly two things: (a) helping each other; and (b) openly discussing any problems, whatever they might be. Therefore, speaking about the retention of editors and Arbcom, I would suggest the following policy changes to encourage collaboration and communication in the project.
I have little to add. But wanted to lend my weight to some points other people have raised.
That's it. Shooterwalker ( talk) 07:21, 11 June 2012 (UTC)
I think the most common situation that really harms retention is this. Editor A (who could be a newbie or not) makes a change that is not unreasonable [4]. Very soon an experienced editor X comes to blindly revert his edit without explanation [5] instead of making a compromise edit, as he easily can [6] (I am giving the simplest example; note the presence of nationalistic POV, "Jewish colonel in the Soviet army" in last version). Things like that really harm desire of people to contribute. In the more complicated cases, the "reverter" frequently explains his revert like this: "I am not sure that your changes represent an improvement compare to the previous version" (then editor A must provide a long explanation, but he might not have enough time and patience). Why revert if you are not sure or not really familiar with the subject, and why revert if you can easily make a compromise version? Unfortunately, I do not know how to fix it (except not writing wrong essays like WP:BRD that encourage exactly such behavior). I know that Arbcom started punishing people for edit wars more seriously, but this is not a good solution. It actually forces reasonable people (who do not want to be sanctioned) to leave political/history subjects to " tigers" who are willing to take risks and edit war. It only helps that we do not have a lot of articles currently owned by "tigers"... My very best wishes ( talk) 13:14, 12 June 2012 (UTC)