Wikipedia talk:Edit warring/Archives/2011/July

RfC on proposed new 3RR exemption

 * Should the following proposed exemption to the three-revert rule be adopted? Three or more edits removing separate, unrelated material which was recently added and unquestionably violates Wikipedia’s core content policies on original research and/or verifiability. 10:20, 8 June 2011 (UTC)

Comments from TreasuryTag (proposer)
Hi, I’d like to propose an additional 3RR exemption which would read as follows: Three or more edits removing separate, unrelated material which was recently added and/or and unquestionably violates Wikipedia’s core content policies on original research and/or verifiability.

The most frequent problem I’ve run into when editing Wikipedia, in all the years I’ve been here, is dealing with material such as     which – while not technically vandalism, because it’s not a wholly deliberate attempt to downgrade the encyclopaedia – are often made in not-the-best-faith and should undoubtedly and indisputably be removed under even the most liberal reading of our content policies. In my view, it’s pointless imposing bureaucratic restrictions on editors who wish to make good-faith edits which unambiguously improve articles.

I do think that repeatedly removing the same piece of information from an article – even if it’s definitely original research – could cause potential problems and disruption as an edit-war, and that is why I’ve included the provision that they reverts must all be different. Quite frequently, editors are surprised to find 3RR templates being slapped on their pages for making three discrete reverts, naturally assuming that edit-warring has to pertain to the same material each time. I would agree that this is a bit of an anomaly and should be rectified as I’ve suggested above.

3RR blocks should be to prevent disruption in the form of edit-wars, not used to stop good-faith editors making helpful and constructive edits. I strongly recommend that we amend the policy. ╟─TreasuryTag► Tellers' wands ─╢ 10:20, 8 June 2011 (UTC)

*Support I would hope that it would be clear that the offending editors (the person or persons adding the material that violates WP policy) may face penalties for edit warring. Erikeltic ( Talk ) 10:57, 8 June 2011 (UTC)
 * Support I've often found myself passing 3RR in good faith under the terms TT outlines above. Although I've never deliberately flouted the rule, and can be hard to stick to it when you're actively interested in an article that is being repeatedly subjected to unsourced/false/speculative information. And this, as I have recently brought attention to at Wikipedia talk:Template messages/User talk namespace is not edit warring and it's hard to see the benefit in repremanding users for it. Personally I don't like that 3RR is applicable to any edits on the same article, no matter how unrelated, but I understand that confusion/ambiguity could arise if the rule wasn't set as such. Also perhaps an exemption for reverting back to a clear and agreed consensus, as I also came a cropper to recently? U-Mos (talk) 10:43, 8 June 2011 (UTC)
 * Oppose This would make it too difficult to enforce the edit warring guidelines -- especially since two of the diffs TreasuryTag supplies above were proven to be factually correct (yes, believe it or not, "Let's Kill Hitler" is the next episode of Doctor Who...). I also don't think that ascribing motives to other contributors, as in "not-the-best-faith", is at all productive.--SarekOfVulcan (talk) 11:07, 8 June 2011 (UTC)
 * Actually one-and-a-half were subseuqently proved correct, but that's not the point. At the time they were added, they were both guesswork and given that there was a clear note saying, Put nothing here without a reliable source, it seems fairly clear that it wasn't a completely good-faith edit. As for your only sentence relevant to this proposal, "This would make it too difficult to enforce the edit warring guidelines," you've not actually provided any basis for that assertion, so it's rather difficult for me to respond. ╟─TreasuryTag► assemblyman ─╢ 11:10, 8 June 2011 (UTC)


 * Question What about a situation like FC Bayern Munich where the squad will not change until July 1, 2011 and consensus is to leave the article as is until then, but various editors keep making the changes over and over again. It would be possible that an editor doing reverts could have 3, 4, or even 5 reverts (reverts from multiple editors) to put the article back to its factual status.  Am I correct in assuming that 3RR would not apply in a situation like that?   Erikeltic ( Talk ) 11:37, 8 June 2011 (UTC)
 * As I underatand, that would be a similar situation to what I was blocked and then un-blocked for, not currently exempt from 3RR by policy but (at least in my case) handled leniently by the community. Hence my closing suggestion above. U-Mos (talk) 12:43, 8 June 2011 (UTC)
 * Sarek - What is your interpretation of my question? Do you think the example would be a violation of 3RR?  Shouldn't disruptive edits (like that) be considered vandalism? Is not fixing vandalism outside the scope of the 3RR?   Erikeltic ( Talk ) 13:05, 8 June 2011 (UTC)
 * I'd say "not vandalism", and therefore not covered by the 3RR exemptions. In a case like that, I'd go to WP:RFPP, ask for it to be locked until July 1, and if it gets locked on the wrong version, put in an editprotected request pointing to the discussion that established consensus, like I did . -- SarekOfVulcan (talk) 22:01, 8 June 2011 (UTC)


 * I would support at least this; in fact, any revert of an edit that is unquestionably against our core principles, and in particular against explicitly established consensus, should be exempt from sanction (what's the point of having these principles if we're going to jump on people when they start enforcing them?) But it should come with a warning that what's "unquestionable" to one may be highly questionable to another, even if both are being reasonable and good-faithed. --Kotniski (talk) 11:57, 8 June 2011 (UTC)


 * My concern would be that it would be far far too easy for people to believe or claim to believe they were enforcing core policies when the fact of the matter was a rather ordinary content dispute. Much drama has spilled on the BLP exemption over this breadth of interpretation already, and I would be very reticent to add another exemption to a rule already struggling with being an effective deterrent because of them. The 3RR was supposed to be a clear, no-nonsense line in the sand – a crude mechanism to halt edit wars, but one that everyone understood and that was effective precisely because there was initially little wiggle room in enforcing. While I appreciate the sentiment that protecting living people from defamation is worth weakening or at least complicating our dispute resolution ecosystem, I cannot say the same for this proposal as written. Skomorokh  12:11, 8 June 2011 (UTC)
 * I agree that any wording such as "unquestionably" leads to a potentially controversial value-judgement being made, but no more than (say) WP:IAR does, and that's also a relevant policy here: as says, why the hell are we blocking established editors for making non-disruptive helpful edits in good faith? My proposal is an attempt to answer that question; I dare say there are other possibilities and I'd be more than happy to consider any of them. (I should also point out that the separate-material-provision is supposed to prevent the use of the exemption on 'content disputes' – which invariably involve multiple reverts on the same material.) ╟─TreasuryTag► constablewick ─╢ 12:30, 8 June 2011 (UTC)
 * This is a reasonable response but I don't think it resolves the tension. Two points:
 * 1. 3RR is intended to be deployed routinely; IAR only exceptionally. This is because a principle that might produce good consequences in local cases (such as "it's ok to make a lot of reverts to an article in a short period of time if you're defending core policies"), won't necessarily produce good consequences overall; ask yourself if what it would be like if good-faithed edit-warriors were "allowed" routinely invoke IAR for instance. We are "blocking established editors for making non-disruptive helpful edits in good faith" because doing so is less disruptive and better for the encyclopaedia in the longer view than allowing good faith editors to be the judge of when it's okay to edit-war.
 * 2. Yes, making the effort to try different, referenced material is better behaviour than trying to re-insert the same stuff, but if it's being reverted it's not benefiting the article and the edit-war is preventing intermediary improvements and driving others away from the article. The purpose of the 3RR – to intimidate the belligerents into using the talkpage/WP:DR – is still necessary in these cases in order for de-escalation of tempers and for any progress to be made on the article.
 * Speaking as an editor quite comfortable with robust compromise-orientated editing, if you mean to loosen the bonds on cases like "two well-intentioned and policy-aware editors try different iterations of satisfying their own concerns and after three or four mutual rejections settle amicably on a final version", I could get behind that, but only if you dropped the invocations of specific content policies (it's the dispositions of the adversaries that matters, not what they are claiming to defend). That said, I'm having a hard time thinking of ungameable wording for these cases even. Skomorokh  12:54, 8 June 2011 (UTC)
 * Hmm... I think part of the problem here is that we're using different definitions of 'edit-war' and of 'improvement'. I would not consider, Editor A adds content X. TreasuryTag removes it per policy. Editor B adds content Z. TreasuryTag removes it per policy. Editor C adds content Y. TreasuryTag removes it per policy, to be an edit-war, because that term implies (at least to me!) a struggle between specific parties over a specific issue. Secondly, I would simply have to dispute your claim that if it's being reverted it's not benefiting the article. Deleting unreferenced speculative content, for instance the "Stephen Fry" diff I linked to above which turned out to be a complete lie, in my opinion is a significant benefit to articles. ╟─TreasuryTag► prorogation ─╢ 13:00, 8 June 2011 (UTC)
 * I must agree with Skomorokh, despite sympathy for the ideas of this proposal. If it were possible, I'd like to have other exemptions, too, such as "blatant POV pushing". The problem with this and with the exemption proposed here is that they're just not objective enough and are too gameable. I also want to point out that admin discretion should apply in cases where an editor breaches 3RR while clearly enforcing content policies by removing original research and POV pushing. The rule never requires a block, after all. Heimstern Läufer (talk) 14:23, 8 June 2011 (UTC)
 * Oppose, the existing policy WP:V allows for the removal of un-sourced content (clarity in edit message implied), an editor wishing to replace the content must include references. There is a multitude of discussion on the removal and replacement of un-sourced content. If the 3RR is coming into consideration for these action then there potential of edit-waring is there, and should be enforced with due consideration. JeepdaySock (AKA, Jeepday) 14:54, 8 June 2011 (UTC)
 * The existing policy WP:V allows for the removal of un-sourced content. I think you've misunderstood the proposal, which is to exempt such permissible edits from counting towards the 3RR. ╟─TreasuryTag► Clerk of the Parliaments ─╢ 14:56, 8 June 2011 (UTC)
 * There should not be an exemption, if it looks like edit waring it probably is. The current exeptins listed are so obvious that they hardly need mentioning. "Reverting your own actions ("self-reverting")." & "Reverting obvious vandalism", What admin is going to block a user on 3RR for these exceptions? JeepdaySock (AKA, Jeepday) 15:03, 8 June 2011 (UTC)
 * I still don't think you get it, but whatever. ╟─TreasuryTag► Subsyndic General ─╢ 15:05, 8 June 2011 (UTC)
 * I don't think that weakening 3RR (or the rule against edit-warring in general) in this way is a good idea. Whether something "unquestionably" violates OR and V is frequently disputed. People who repeatedly add content flawed in this way should be convinced of their error or, if that is not possible, blocked for disruption. The proposal's wording is also flawed, as the "and/or" allows reverting material simply because it was recently added, which makes no sense.  Sandstein   16:42, 8 June 2011 (UTC)
 * Sorry, you're right about the wording, which was a mis-type. Corrected. ╟─TreasuryTag► collectorate ─╢ 16:44, 8 June 2011 (UTC)
 * TT, I'm not sure your correction actually fixed the wording - it doesn't make sense to me now. I am sympathetic to your aims, but I have to oppose based on as others have said how hard it is to objectively define these things sometimes. On the other hand, maybe a proposal to make page protection easier to obtain in these kinds of cases might be better? E.g., a lower bar to get say, a semi-protection for an announced (but as-yet unaired) TV episode?  umrguy 42  17:56, 8 June 2011 (UTC)

Comment Having had occasion to exercise administrative discretion in otherwise clear 3RR cases, I'm not convinced that hard-coding another specific exemption into the policy is the best solution here. I appreciate the principle that TT is describing here, but (unfortunately) it can lend itself to abuse, even if said abuse is clearly unintentional and in good faith. Perhaps a compromise might lie in adding some language to the admin notes to emphasize and clarify that blocks should reflect apparent intent. --Ckatz <sup style="color:green;">chat <sub style="color:red;">spy  17:18, 8 June 2011 (UTC)
 * I'd definitely support something like that. <font color="#00ACF4">╟─TreasuryTag► Chancellor of the Duchy of Lancaster ─╢ 17:31, 8 June 2011 (UTC)
 * While that wording would probably be a good idea, it does seem to open up extra potential for gaming unblock requests.-- SarekOfVulcan (talk) 17:50, 8 June 2011 (UTC)

This proposal seems like a good first pass at addressing the issue. It's clear to me that some sort of clarification to reflect what's really happening on the ground (so to speak) would be good. User:Ckatz's point above is well made, and something that I would support as well. I don't see "gaming unblock requests" as being an issue to be particularly concerned about, since wikilawyering blocks is rarely effective. — <span style="font-family: Courier New, monospace ;font-style:italic">V = IR (Talk&thinsp;&bull;&thinsp;Contribs) 18:24, 8 June 2011 (UTC)


 * Oppose - This opens up a huge can of worms for edit warring... "Unquestionably violates...original research"??? Ask 20 Wikipedians what constitutes "original research" and you will get 20 answers. If implemented this would be a wet dream for people who get their jollies with ANI drama... Three RR is a very simple concept and there should be damned few exceptions to it. Carrite (talk) 18:38, 8 June 2011 (UTC)
 * I agree with many of the comments already made, that one can be sympathetic to the concerns giving rise to the proposal, without wanting to complicate a rule that needs to be clear and difficult to game. We actually have, already, a workable alternative to the need to make multiple reversions in cases like this: post on an appropriate noticeboard, and get other editors to revert what needs to be reverted. If it's really a case where there would be consensus that core policies require reversion, then the reversion will get done, and if it takes a few extra steps, well, the sky does not fall in the mean time. Or, as already noted above, get page protection. --Tryptofish (talk) 18:54, 8 June 2011 (UTC)
 * That's a real good point, and is actually the reason that 3RR was created in the first place. People need to realize that, despite Wikipedia's high profile, problematic content on Wikipedia isn't going to change the world. We all need to step back a little and take our time (within reason). That being said, there certainly is a reality here that administrators tend to use discretion in the application of 3RR, which is what User:Ckatz was getting at above I think. Codifying the common behavior of administrators in the admin notes of the policy to explain what is going on is a good thing to me, because it clarifies what can easily be confusing to newer users ("3RR will get you blocked", but don't mind these examples over here where we've let people slide on the issue in the past...). — <span style="font-family: Courier New, monospace ;font-style:italic">V = IR (Talk&thinsp;&bull;&thinsp;Contribs) 19:13, 8 June 2011 (UTC)
 * Thanks, yes. In fact, I've gradually come to understand that treating Wikipedia that way (stepping back, etc.) is pretty much The Path to WikiZenTM. --Tryptofish (talk) 19:25, 8 June 2011 (UTC)
 * Oppose, essentially agree with and  - from above comments. -- Cirt (talk) 20:01, 8 June 2011 (UTC)
 * Sarek didn't actually provide a substantive reason that I've yet found... <font color="#A20846">╟─TreasuryTag► estoppel ─╢ 21:34, 8 June 2011 (UTC)
 * Separate edits removing unrelated material are separate edits. To block for 3RR when the material removed is different in each case would be rules-lawyering. I don't see a need for this, and it would encourage gaming by arguing over just how different the edits were. Guy (Help!) 20:53, 8 June 2011 (UTC)
 * Separate edits removing unrelated material are separate edits. To block for 3RR when the material removed is different in each case would be rules-lawyering. Myself, U-Mos and many others have been warned and blocked on that basis multiple times. I agree that it's a fairly dumb idea, and that is what my proposal seeks to prevent. <font color="#A20846">╟─TreasuryTag► estoppel ─╢ 21:34, 8 June 2011 (UTC)
 * Yea, well... I'm honestly not trying to personalize this, but I've just got to wonder how much of that (being warned/blocked over 3RR) is actually because of 3RR violations and how much comes from "any excuse for a block". That's the one negative side effect of having "bright-line rules", is that it's possible to use them for... other reasons. — <span style="font-family: Courier New, monospace ;font-style:italic">V = IR (Talk&thinsp;&bull;&thinsp;Contribs) 21:55, 8 June 2011 (UTC)
 * I've always interpreted 3RR as applying to substantially the same material. It's perfectly possible to have an edit war going on with one user about one thing on a page plus a sequence of reasonably collaborative to-and-fro editing about something else with another; counting the two together towards 3RR doesn't encourage good editing. This is made worse by the sometimes very expansive notion of "revert" such that it's barely distinguishable from "edit after somebody recently changed something". Rd232 talk 12:22, 9 June 2011 (UTC)
 * Oppose The examples given do not require unlimited reverts to fix. There is no great urgency to getting right the title of a future TV show. The only types of edits which have that level of urgency are vandalistic edits and those which add unsourced negative material about BLPs. Everything else can be corrected using talk page discussion consensus, notice boards, and other, relatively slow, remedies.     Will Beback    talk    22:23, 8 June 2011 (UTC)
 * Oppose per WP:WhatTheHeckIsThisSupposedToFix?. Poorly worded, and a Wikilawyer's delight. The appropriate response to multiple episodes of edits, each of which "unquestionably violates Wikipedia’s core content policies" is to revert them, and then justify it afterwards per WP:IAR. I've done this myself on occasion, though I've never needed to justify it - it was "unquestionable" after all. If you are that confident, break the 3RR rule, and rely on common sense. AndyTheGrump (talk) 03:07, 9 June 2011 (UTC)
 * Oppose - Well intentioned proposal. However, just as "one man's terrorist is another man's freedom fighter", the terms "unquestionably", "separate unrelated material", and original research, as examples, are open to differing interpretations and wikilawyering. 3RR is intended to stop edit warring, and edit warring is edit warring, except for the few serious and time critical exemptions as listed in WP:NOT3RR. Oppose per Sarek, Carrite, and the well reasoned arguments of Skomorokh. Also oppose per WP:CREEP & WP:BEANS. — Becksguy (talk) 05:12, 9 June 2011 (UTC)
 * Oppose - unworkable. Slightly more plausible would be an exemption for removing entirely unsourced content (i.e. provision of any source, good/bad/indifferent/irrelevant, would mean the exemption wouldn't apply). Rd232 talk 12:18, 9 June 2011 (UTC)
 * conditional support The way *I* understand the 3RR it's meant to avoid "no you didn't" "yes I did" "no you didn't" style edit warring where one editor puts something in, another takes it out, they put it right back, and so on. Firstly I have an issue with the fact that this makes it so that the editor removing the matieral always wins (they have the last revert available).  I support this proposal in one case: when the material is being added by multiple editors.  So if you have one editor who's insisting on inserting unsourced material into the biography section and another who keeps adding a spurious WP:COATRACK "controversies" section, you can revert BOTH of them up to 3 times.  Also, I would support a policy that says removal of unsourced material (that should have a source! not just reverting changes to formatting!) is always encouraged at all times, be it 1, 3, or 50 times a day from the same article. HominidMachinae (talk) 20:11, 9 June 2011 (UTC)
 * Sort of support I also think that the community's intention involves multiple removals of "substantially the same material". Imagine an not-very-good article with four sections.  A newbie makes one low-quality change to each section.  I remove each change, using a separate edit in each section.  My removal is either reverting the hopelessly bad change to the section, or re-writing the section.  This is four edits, and all four removed the other editor's work.  But I'd expect a note of thanks at the end of that, not a block for "removing the other editor's work four times on the same day in one article".   WhatamIdoing (talk) 00:38, 10 June 2011 (UTC)
 * If you made those four edits without interruption then they would not count as a 3RR violation.   Will Beback    talk    01:24, 12 June 2011 (UTC)
 * TreasuryTag seems to believe that at least a minority of admins don't agree with you, and I think that suggests that we need greater clarity. But I'm not sure that I understand what "without interruption" means.  If I make the first two edits, then someone else changes a category on the page, and then I make the last two edits, is that somehow a 3RR violation—whereas if the third person hadn't invisibly (thanks to the software's handling of section editing) "interrupted" my work by changing the cat, then my edits are fine?  That doesn't make any sense at all.  It doesn't even make sense if the person making the change is the original low-quality editor:  So long as I only make one change to each section, it shouldn't make the least bit of difference if that person is working on some other part of the article at the same time.  WhatamIdoing (talk) 04:37, 28 June 2011 (UTC)


 * Wha? - "Three or more edits removing separate, unrelated material" isn't prevented by, or even covered by 3RR. You can use 15 different edits removing content if you want. 3RR only kicks in if someone reverts your 15 edits, and you revert them, and they revert you again, etc. Second, if someone's removing blatantly inaccurate material and starts edit warring over it, 3RR doesn't apply. If the exemption was "removing blatantly inaccurate information", I would be more inclined to support, in theory.  <font face="Old English Text MT">Swarm  <font face="old english text mt">X 20:37, 13 June 2011 (UTC)
 * I think you'll find you're wrong. 3RR applies even if the content being reverted is different each time. That is why I made the proposal. Obviously. <font color="#00ACF4">╟─TreasuryTag► pikuach nefesh ─╢ 20:40, 13 June 2011 (UTC)
 * So by "Three or more edits removing separate..." you mean "three or more reverts to remove separate..."? Your proposal is unfortunately worded; an "edit removing content" is not the same as a "revert" in any sense.  <font face="Old English Text MT">Swarm  <font face="old english text mt">X 20:49, 13 June 2011 (UTC)
 * A revert "undoes the work of another editor." Removing content is undoing the work of the editor who put it there. Therefore, removing content is a revert. I don't know why you're commenting here if you don't understand the basic issues. <font color="#A20846">╟─TreasuryTag► Not-content ─╢ 20:52, 13 June 2011 (UTC)
 * I very much appreciate your explanations, Treasury. My apologies for your frustration.  <font face="Old English Text MT">Swarm  <font face="old english text mt">X 02:06, 14 June 2011 (UTC)
 * Oppose Due to, as stated above, the highly subjective nature. The terrorist/freedom fighter idiom is a perfect parallel. There's no way to prove something is "unquestionably original research", so naturally this exemption will start crossing into highly debatable situations. I can already imagine the unblock requests from everyone who was edit warring to remove article content. Edit warring is, and always has been a two way street. We can't have every Randy in Boise thinking he's exempt from 3RR just because he thinks he's removing "unquestionable original research".  <font face="Old English Text MT">Swarm  <font face="old english text mt">X 02:03, 14 June 2011 (UTC)
 * Oppose Clearly too open to interpretation, and would definitely lead to abuse. Cenarium (talk) 13:08, 14 June 2011 (UTC)
 * Oppose. Good idea in principle. But as an ex-admin who has worked AN3, my firm view is that it will be totally unworkable, gameable, and lead to disputes. --Mkativerata (talk) 22:47, 25 June 2011 (UTC)

A narrower question
Much of the opposition to the original proposal has focused on the policies cited and asking for exemptions to be based on editors' assessment of what is "unquestionable". As some of the opposition to the proposal has been sympathetic, and it is still attracting some support at the time of writing, it might be helpful to separate out the issues somewhat.

Setting aside the practicalities of enforcement for a minute, to what extent should successive reverts of the same material be judged differently from edits removing separate, unrelated material in principle? <font face="New York">Skomorokh 12:53, 10 June 2011 (UTC)
 * It's critical that in addition to being separate, unrelated material, the material is from different editors. On a busy page, one could easily rack up 4 unrelated reverts of non-vandalism, but non-helpful drive-by additions during the course of a day if monitoring it closely. (Presumably this instance is unlikely to be reported in the first place, for most people.) 4 reverts of the same editor, even over unrelated items, should not be happening - there should be a conversation happening there and a revert-heavy editing environment will discourage it from taking place. Christopher Parham (talk) 19:50, 10 June 2011 (UTC)
 * The point of 3RR is to give priority to stopping edit wars over getting perfect content.[xkcd.com/386/] If the content added does not rise to the level of vandalism or BLP violation or libel or copyvio, it can wait 24 hours. If there is a real problem, it's very easy to ask another editor for assistance and thus avoid 3RR.   DGG ( talk ) 00:11, 12 June 2011 (UTC)
 * That approach is fine if the article isn't very good. But if it is good, there needs to be more emphasis on talking before making changes, and the corollary of that is greater latitude to revert undiscussed changes, especially if they're of obviously poor quality (eg unsourced or poorly sourced, etc). In short, your attitude was fine when WP was young, but the more we manage to get articles up to a decent standard, the more emphasis there needs to be on protecting content from unnecessary change. After all the articles are ultimately there for the benefit of readers. Rd232 talk 01:19, 12 June 2011 (UTC)
 * The sentence "protecting content from unnecessary change" gives me a serious case of the heebie-jeebie's. If this isn't a backhanded route to article ownership, I don't know what is. I have yet to see a Wikipedia article that is written so well that it can't stand some editing, and we're certainly not protecting articles that meet some mythical quality standard (yet... that seems to be where this sort of thing is headed, though). The "that anyone can edit" part is the first principle of Wikipedia, you know. — <span style="font-family: Courier New, monospace ;font-style:italic">V = IR (Talk&thinsp;&bull;&thinsp;Contribs) 15:03, 13 June 2011 (UTC)


 * I don't think it's critical that the material be from different editors. Imagine a newly created article on a garage band:
 * At the moment, it has one editor, ever. You remove four plainly inappropriate sentences (e.g., "The band is the best goth-muzak group in the nation") individually, explaining exactly why you've removed each sentence in the edit summaries.  This should not be treated differently than removing all four in one go, with an edit summary of "rm garbage".  WhatamIdoing (talk) 04:45, 28 June 2011 (UTC)

Edit warring and sockpuppets
The policy currently says that :
 * "Reverting actions performed by banned users, their sockpuppets and by tagged sockpuppets of indefinitely blocked accounts"

is not considered edit warring. Obviously, the policy means the edits made by these users before the fact of sockpuppetry has been revealed, because after that the sock becomes blocked and it cannot edit any more. In connection to that, what should we do in a situation when some editor had already been sanctioned for edit warring against a user who was later found to be a sockpuppet? In my opinion, it would be correct to edit his block log accordingly (i.e., to remove the record about the block from the log), because the 3RR refers to the users, not not the accounts, so removal of any edits of the blocked user's sock is not a edit war. I suggest to add to the policy that the user who was sanctioned for the edit warring against a user who has later been found to be a sock of indefinitely blocked or banned user can request for removal of the record about his block from his block log.--Paul Siebert (talk) 00:49, 27 June 2011 (UTC)


 * Edit waring is about a behavior not about the edits that are being reverted, so in your example above per WP:AGF the person who was reverting the sock edits was wrong. Per current policy, "Reverting actions performed by banned users, their sockpuppets and by tagged sockpuppets of indefinitely blocked accounts" is a category of vandals who happen to also be identified as banned or socks, which makes the identification of the vandalism more clear. A known sock or banned user who is engaged in an edit war, is wrong and knows they are wrong. This sentence encourages new editors to continue reverting edits pending an admin block for the banned or sock user. JeepdaySock (AKA, Jeepday) 16:14, 27 June 2011 (UTC)
 * Edit waring is a behaviour that, among other things, manifests itself in reverting of the edits made by others. Reverting of the edits made by socks of banned or blocked users is not edit warring. The socks that have already been tagged are not able to edit, because, as a rule they have already been blocked by that moment, therefore, under "Reverting actions performed by banned users, their sockpuppets..." the policy means the edits made by socks before the fact of sockpuppetry have been revealed and the socks have been blocked. In connection to that, I do not understand why the editor who reverted the edit made by the blocked user via its sockpuppet account (which had been revealed and blocked later), and who has been blocked for that, cannot request for removal of the record about that block from his block log: in retrospect, he did an absolutely correct thing, so the revealed fact of sockpuppetry makes the block unjustified.--Paul Siebert (talk) 16:54, 27 June 2011 (UTC)


 * It depends on what you mean by "know". I encounter a particular sockmaster fairly frequently, and I usually know as soon as I see the change that she's back again.  But what's obvious to me isn't going to be obvious to anyone who just happens by the articles.  I could easily see someone who hasn't encountered her before not knowing what's going on, and warning, or possibly even blocking, editors who really are doing the right thing, while waiting for CU to get all their ducks in a row.  WhatamIdoing (talk) 04:53, 28 June 2011 (UTC)
 * Well, let's consider the following sequence of events:
 * You revert the edit that is not supported by the cited source;
 * A user X re-adds it back without discussion;
 * You revert this edit and (on the talk page) request to provide a reference;
 * A user Y re-adds the text back;
 * You explain the problem on the talk page and remove the unsourced statement again;
 * A user Y re-adds it without any explanation;
 * You remove the unsourced text - and get blocked for violation of 3RR.
 * A user Z starts an SPI against the user Y, who appears to be a sock of the indefinitely blocked user AAA.
 * Per policy, removal of the edits ##4,5 is not counted as revert now, so, in retrospect, your actions were not a violation of 3RR. However, despite the fact that all traces of the sockpuppet activity should be removed from Wikipedia, the record about your block, which de facto is a direct result of the activity of a sockpuppet of the blocked user, stays. In my opinion, that is incorrect. A good faith users, who have been sanctioned as a result of the edit wars with sockpuppets of blocked users should have a right to request for cleaning of their block logs as soon as the fact of sockpuppetry has been revealed.--Paul Siebert (talk) 20:42, 28 June 2011 (UTC)
 * Sure seems to a lot of hypothetical here, are there real edit examples? JeepdaySock (AKA, Jeepday) 10:52, 29 June 2011 (UTC)
 * To the best of my knowledge there is no way of removing an entry from a block log. JamesBWatson (talk) 14:34, 29 June 2011 (UTC)
 * Well, if this is technically impossible, the whole discussion has just an academic interest. However, the situation I described is not too hypothetical, taking into account that the indefinitely blocked users are usually blocked for edit warring, so their sockpuppets inherit their behaviour. As a result, a probability to be involved in the edit war with sock (who, despite its newbie's appearance is an experienced edit-warrior) is relatively high.--Paul Siebert (talk) 14:38, 29 June 2011 (UTC)
 * In the hypothetical example above, "You" was either edit waring with 3 separate editors, and sanctions would be appropriate. Or if they had reason to believe the edits were organized vandalism (meat puppets) or the actions of a single sock they should have reported at to WP:AIV and/or WP:SPI if the edits were judged inappropriate by the community another editor would have reverted edits by "Y" and/or "Z". WP:CRYSTAL applies to rules as well as articles, There is no scenario where edit warring is appropriate. WP:OWN might also be a consideration the hypothetical question. JeepdaySock (AKA, Jeepday) 14:54, 29 June 2011 (UTC)
 * The idea that "in retrospect, your actions were not a violation of 3RR" makes it justifiable is silly. Your actions have to be judged on the basis of information available to you when you took those actions, not "in retrospect" on the basis of information you didn't know about. JamesBWatson (talk) 16:29, 29 June 2011 (UTC)
 * Since the editing of the block logs is technically impossible, the discussion has just a theoretical interest now. Let me point out, however, that your rationale is somewhat illogical. The 3RR is a quite formal thing: it is based on bare count of reverts, which are defined quite formally. If new evidences indicate that "your" revert appeared not to be a revert, the decision about the block should be re-considered.
 * Clearly, since the 3RR cases are judged by just a revert count, the actions of the users violating the 3RR are not judged on the basis of the information available for them. However, if we take into account such information, for instance, in the hypothetical case I discuss, we need also take into account that the user Y clearly violated the WP:BURDEN policy (repeatedly added the unsourced information), so we have here a conflict between two policies. In addition, the user Y was clearly gaming the system, so the user "You", who believed that he deals with a good faith editor was in unfair conditions. This situation is general, because the sock of banned users are not acting in good faith, and this information is not available for their opponents. In summary, the behaviour of good faith users towards their peers and towards those who have been found to be the socks should be judged according to different standards, because the latter never behave in good faith.--Paul Siebert (talk) 17:15, 29 June 2011 (UTC)


 * There is not a conflict between two policies, this is a community project. Had "you" involved the community in the edit dispute instead of taking upon "yourself" to battle three different user ID's to remove content, "you" would not have violated 3RR. The behavior of 3RR is the issue not the edit count, see Edit_warring even if it was clear at the time that three different user ID's were socks, the action was wrong.  It is not a battle for one volunteer in a community. JeepdaySock (AKA, Jeepday) 10:44, 30 June 2011 (UTC)
 * I would say, the only issue with the "You's" behawiour is the edit count, because he acted in a full accordance with other policies: he request to provide a ref, removed the unsourced text, he initiated the talk page discussions. By contrast, the behaviour of his opponents is much more problematic. Regarding the involvement of community into the dispute, this is not always a panacea, because even if you request for such an input, you do not necessarily get it.--Paul Siebert (talk) 14:02, 30 June 2011 (UTC)
 * There is probably no editor on Wikipedia who is more engaged in improving references to articles without respect to their subject, or who has been involved in more controversy for the removal of unreferenced content, then myself. So I feel confident in saying there is no instance where WP:BURDEN is so clear that a single individual should assume the enforcement of it unilaterally. For every rule there is generally an exception. But no place does it say anyone editors interpretation of the rules is final.  "You" violated the trust of the community by attempting to enforce a single interpretation of WP:BURDEN which is what 3RR is intended to discourage. When Paul Siebert describe "You's" conduct with his "opponents" Paul validates the engagement of a war, and validate that 3RR is correct to sanction the action. In the example given "you" must have been watching the page, as it is unreasonable to suspect any other method of noticing the edits.  "You" had potential issues with WP:COI and/or WP:OWN and should have involved the community. When the community is involved they will either agree with "you's" assessment and remove the added text, or not. If the community does remove the text, then "you" has no risk of violation 3RR, if the  community does not remove the text, then community decision is to allow it to stay and if "you" would removing the content against community judgement and should be subject to 3RR. Realizing that I have made several responses on this chain, I will desist from further comments. JeepdaySock (AKA, Jeepday) 15:37, 30 June 2011 (UTC)


 * It seems to me that there is a clear consensus against the suggestion. JamesBWatson (talk) 18:19, 30 June 2011 (UTC)

Wording unclear
As mentioned above the exemption reads:


 * Reverting actions performed by banned users, their sockpuppets and by tagged sockpuppets of indefinitely blocked accounts.

So, what is a tagged sockpuppet? How does it differ from a blocked sockpuppet? There is nothing hypothetical about this question: see Sockpuppet investigations/Vote (X) for Change. Jc3s5h (talk) 21:59, 20 July 2011 (UTC)
 * A tagged sockpuppet is one where the user page has been tagged to indicate that it is a sockpuppet of the blocked account in question. A blocked sockpuppet is one where the account has been blocked. The point of specifying "tagged sockpuppets of indefinitely blocked accounts" is that it is not enough for an individual editor to revert edits and simply say "I think it's a sock puppet account".
 * I have looked at the page linked, and don't see anything relevant. Can you be more specific? After spending some time searching, the only thing I could find that could be considered relevant was a link to this edit. Was that what you meant? (If so it would have helped to have given the link directly, and if not then I have no idea what you are referring to.) If that is what you meant then I don't see that rewording the passage you mention would help. The editor in question was clearly determined to keep on editing no matter what, creating sockpuppet accounts, editing without logging in, denying the validity of administrative actions taken against him/her, etc etc, and would have continued no matter what wording was used. JamesBWatson (talk) 08:12, 21 July 2011 (UTC)


 * Lets say I read an edit, and recognize it as being characteristic of an indefinitely blocked user, but the edit is made using some other account or IP address. So what text or template should I look for on some Wikipedia page to know that the account/address that made the edit is a tagged sockpuppet? For example, if the Socksuspectnotice or Uw-socksuspect templates are present on the talk page of the account/address that made the edit to be reverted,and refer to the indefinitely blocked user, does that mean it is "tagged"?

Improvement in wording needed
We had a situation occur recently that was a perfect example of improvement editing with different editors making minor changes in a give-and-take style. Brainstorming might be another way to put it. Each editor was comfortable with how the edits were proceeding, no editors were fighting, but really just trying to fine-tune or tweak the wording to make it better. Great flow, but an editor who I have had negative encounters with was watching the page and warned me *only* that I was violating 3RR. While several of us had technically reverted more than 3 times, none of us were 'warring'. It was a cordial and friendly exchange of ideas where all parties had the same goal, which was to say "how's this wording?" "good, but how's this?" "ah, great, now how about this version?" etc.

So the problem is '3 reverts doesn't always equal edit warring'. We all realize that edit warring is the issue we're trying to prevent, and technically you can do that with just 1 edit. I realize it is a 'bright-line rule' because it is easy to understand, but I would hope admins can see the difference, or that the policy might be re-written in a way to recognize that 3RR should be about edit warring, not just editing. -- Avanu (talk) 14:06, 6 July 2011 (UTC)
 * I think trying to change the wording in the way you suggest would be a mistake. Has there ever actually been a case of action on "edit warring" being taken in the situation you describe? Common sense should be sufficient. Years ago Wikipedia was simple and straightforward, and anyone could have a reasonable understanding of how things worked. Over the years there has been more and more instruction creep, with a shift away from just outlining general principles towards attempts to legislate for every conceivable situation, listing exceptions and then exceptions to the exceptions in some cases. This has been severely to the detriment of the project, which has become far less accessible. One of the five pillars of Wikipedia is "Wikipedia does not have firm rules", and that should be enough to cover this case. If common sense says "this is not a case where the edit warring policy should be invoked" then nobody needs to invoke it. In the (probably extremely rare) event of someone trying to invoke it inappropriately, that person can have it explained why it it inappropriate, and if there is consensus that no action is needed then that should be enough. JamesBWatson (talk) 07:57, 21 July 2011 (UTC)
 * Yes, I know the examples of such a situation. The users acting in good faith were blocked because another user made a change in between two consecutive edits. Although both users were acting independently, and they didn't interfere with each other, two consecutive changes made by the first user had been considered as two separate reverts, and this user had been blocked. In my opinion, the rule in its current version is written to make the admins' life easier (because it provides a strict algorithm for its implementation: as soon as three diffs have been provided, an admin has simply to check if these diffs are correct and to impose the sanctions), however, that has been achieved at cost of good faith editors, who write and maintain Wikipedia.
 * In my opinion, the solution may be to allow unlimited number of the reverts for the users when they are working with each other in concert. In other words, whereas a violation of the 3RR remains in force, only those users can report this violation who are directly involved in this chain of reverts (obviously, in a case of "a cordial and friendly exchange of ideas" that will never happen, and the third party acting in bad faith will not be able to interfere, unless he became a participant of the dispute). For instance, in the case I described above, the 3RR violation had been reported by the third user, who tried (successfully) to use the letter of the policy against its spirit. It deserves mention that this user had been subsequently blocked as a sockpuppet of the indefinitely blocked user, which is an additional proof of his bad faith...--Paul Siebert (talk) 21:49, 21 July 2011 (UTC)

What are "tagged sockpuppets"?
A class of the exempt reverts, which do not count for three-revert-rule purposes, is "Reverting actions performed by banned users, their sockpuppets and by tagged sockpuppets of indefinitely blocked accounts." In this statement, what is a "tagged sockpuppet"? Jc3s5h (talk) 19:21, 21 July 2011 (UTC)
 * Is there some reason for thinking the answer I gave above (when you last asked this question) is wrong? JamesBWatson (talk) 19:31, 21 July 2011 (UTC)


 * Three issues I have with your response are
 * You don't say exactly who applies the tag; could it be the same editor who reverts the edits?
 * You don't say exactly what the "tag" consists of. Should it be one of a small group of templates, or could it be any words that mention sockpuppetry and also the identity of the indefinitely blocked user?
 * You say "user page", but it is generally considered bad form to edit someone else's user page. Also, IP addresses don't have user pages, they just have contribution lists.
 * My comments might seem rather nitpicky, but when a penalty is applied by counting without regard to merit, in inherently contentious situations, I think the rules for counting should be quite clear. Jc3s5h (talk) 20:26, 21 July 2011 (UTC)

I've removed the 'request for comment' tag as this appears to be an editor asking for advice rather than requiring outside input for dispute resolution. <font color="#FFB911">╟─TreasuryTag► duumvirate ─╢ 19:41, 21 July 2011 (UTC)
 * subsequently replaced the tag and deleted my comment above. <font color="#00ACF4">╟─TreasuryTag► Counsellor of State ─╢ 21:26, 21 July 2011 (UTC)
 * My response:
 * accusations about sockpuppetry are only made if the sockpuppet is extremely clear and obvious; see this essay for an explanation.
 * While it seems unorthodox that someone would not use a template to note a sockpuppet, there is no requirement that a sockpuppet be noted with a template - any old text would do.
 * It is policy that such tags be placed on the (otherwise blanked) userpages of sockpuppets. Also, IP users DO have userpages; here is a random example. Furthermore, because everyone has an IP address, such users do not count as sockpuppets unless the user abusively uses another account.
 * Quinxorin (talk) 00:09, 22 July 2011 (UTC)
 * OK, that explains a lot, and increases my concern about the wording of the policy. I think I found the page that you, Quinxorin, were referring to: Sockpuppet investigations/SPI/Administrators instructions. By the way, it is not labeled as a policy, guideline, or anything like that. I have a problem with a policy using a term that is defined in a page that is addressed to administrators, and unlikely to be familiar to non-administrators.


 * It indicates that unlike registered accounts, IP accounts are only tagged on the talk page, and only if they are blocked. So this says that damage from IP socks in excess of 3 reverts per day can't be cleaned up if, by the time an administrator gets around looking at the matter, the sockmaster has stopped or moved on to another IP address, and the administrator does not issue a block. Also, no one who is not an administrator or working with an administrator can make use of this exemption, because the IP must be blocked before it can be tagged, and it one must be an administrator to issue a block. Jc3s5h (talk) 01:04, 22 July 2011 (UTC)


 * Actually, edits by sock IPs like this would appear to meet the definition of vandalism by falling under gaming the system. Any such edits would be exempt under the vandalism exemption.  So, probably the exemptions should be merged to clarify this.  I think there's no problem with tagging the sock yourself (in the case of the non-IP) but the real point here is if you act under an exemption to 3RR, you'd better be sure of yourself and probably ought to be involving an admin anyway, since someone should consider blocking the sock.--Doug.(talk • contribs) 09:12, 22 July 2011 (UTC)


 * Another unclear point is if an IP talk page has a notice of a block for sockpuppetry, but the block has expired, is the IP still considered "tagged". (Active sockpuppetry that raises this question may be found at User talk:188.220.57.163.) If Doug's argument is valid, it does not matter, but Doug's interpretation leaves the interpretation of whether or not the IP is gaming the system entirely with the reverting editor, which is a different tack than the current wording of the sockpuppet exemption. Jc3s5h (talk) 13:33, 22 July 2011 (UTC)


 * I think I just answered your question. See this page history and your stalker's talk page and contribs (or mine) for details. ;)--Doug.(talk • contribs) 14:05, 22 July 2011 (UTC)


 * Support concern I am a completely uninvolved editor, brought here by the RfC. I support Jc3s5h's concern over the current wording.  Can someone explain to me exactly what tag is being referred to and what the procedure is for adding such a tag. Martin Hogbin (talk) 09:28, 23 July 2011 (UTC)


 * Distinguish banned, indefinitely blocked, and blocked users? I don't understand why the exemption makes a distinction between banned users (where no "tag" is needed), indefinitely blocked users (where a tag is required), and users blocked for a definite period of time (where the exemption does not apply). No variety of blocked user is allowed to use a sockpuppet, and all such edits should be equally subject to removal. The issue isn't what kind of block does the sockmaster have, the issue is how strong is the evidence that the purported sockpuppet really is a sockpuppet. The whole exemption seems to be built on unsound logic. Jc3s5h (talk) 11:14, 23 July 2011 (UTC)


 * The thinking here seems to be that edits which benefit the project are allowed, but those of banned users are disallowed (simply because there would be no point to the ban otherwise). 78.145.26.194 (talk) 18:31, 27 July 2011 (UTC)