Wikipedia talk:Three-revert rule/Archive 3

Manitoba (disambiguation)
Not sure if this is the location to note this problem, but here goes: there is 1+ person who repeatedly deletes content on the Manitoba (disambiguation) page. The deletions always relate to Manitoba as the former stage name of musician Dan Snaith. Since 2004, Snaith uses the name "Caribou". I came to the disambiguation page to find the new (Caribou) name for Manitoba. I added this content and it was deleted a few hours later. Another user informs me that this has happened repeatedly on this page. Can someone assist? Thanks, Hu Gadarn 17:42, 2 June 2006 (UTC)

In excessive cases, people can be blocked for edit warring or disruption even if they do not revert more than three times per day.
I do not like this section. It gives admins blanket power to block anyone they want. All they have to do is arbitrarily call something "edit warring" and block users they dislike/disagree with. anyone have plausible checks to this? Otherwise it must be removed. —The preceding unsigned comment was added by MateoP (talk • contribs) 16:37, 1 June 2006 (UTC).


 * In practice it is a restating of the dissruption clause. The check is other admins.Geni 02:51, 10 June 2006 (UTC)

3RR confusion
In recent months, I have come across several instances of administrators holding different interpretations of 3RR policy. Looking over the history of this page, it's not difficult to understand why: significant modifications have sometimes been made without consensus discussions taking place beforehead. The current version is qualitatively different from that of 1 January 2006, and I am not certain that it would be accepted by all, or even most, administrators.

I would refer editors specifically to this particular entry from March of this year. This may have been intended as clarification of an existing rule; in practice, however, it may have established a contentious interpretation as official policy. I could list other examples.

It may be adviseable for the Wikipedia community to undertake a comprehensive overview of this policy in the near future. Until then, I would suggest that Wikipedians use extreme caution in blocking editors for ambiguous cases involving perceived 3RR violations. It is simply not reasonable to expect every regular contributor to follow the frequent changes to this page. CJCurrie 02:42, 10 June 2006 (UTC)


 * On a related matter, I'm genuinely curious where the consensus was established to add Note: There is no requirement for the reverts to be related: any four reverts on the same page count. to this policy. As long as I've been an administrator, 3RR has applied to reverts of the same material; unrelated reverts were explicitly excluded from the rule. When was this change discussed, and how exactly were people planning to ensure that longstanding administrators knew that there'd been a policy change this significant? Bearcat 02:55, 10 June 2006 (UTC)


 * That became consensus when that was the way I inforced the rule short after its creation.Geni 11:05, 10 June 2006 (UTC)


 * I propose we simply eliminate it, and post a notice on WP:AN to the effect we've done so. Thusfar, it's led to very subjective interpretations of who should or should not receive a block (and if it makes you feel any better, Bearcat, HOTR is hardly the first user to be blocked under this interpretation.)  Using extreme caution is acquiescence to the introduction of a whole new dimension of subjectivity that simply needn't be here, and will likely lead to more wheel-warring (ahem) in the future.


 * If 3RR is to be tightened - and I'm certainly open to it - it should be to make it the Two-revert rule. Such a change would be immediately comprehensible to all.  This one, while well-intended, plainly isn't, and has generated an unacceptable level of confusion on the noticeboard.Timothy Usher 03:12, 10 June 2006 (UTC)


 * To your "dimension of subjectivity" point, I can only respond "not so". The policy does not indicate editors must be blocked for any 3RR violations, let alone for ambigious situations involving recent rule changes.  CJCurrie 03:24, 10 June 2006 (UTC)
 * Well, editors already can be blocked for one revert, depending on what it is. 3RR is meant to be an "electric fence", not an interpretive judgement call.  As in touch it, brrzzzt.  Not, maybe brrzzzt.Timothy Usher 03:34, 10 June 2006 (UTC)


 * That depends on what you mean by "brrzzzt". The 3RR policy stipulates that more than three reverts by a single editor to a single page in 24 hours is forbidden.  It does not stipulate that the editor in question must be blocked.  CJCurrie 03:45, 10 June 2006 (UTC)

Another matter:

I've looked over the 3RR page history, and far as I can tell this is the first instance of the phrase "in whole or in part" being added to the policy. I cannot find any prior discussion for this change, which was made less than a year ago.

Did such a discussion occur in a different forum? If so, could someone please direct me to it? CJCurrie 03:24, 10 June 2006 (UTC)


 * I don't think "in whole or in part" is the problem here. 3RR should most certainly apply if there is common text which is being restored or undone, even if it's buried beneath a bunch of other changes.  SlimVirgin was on the right track; unfortunately the language about "undoing other editors' work" is hopelessly ambiguous.Timothy Usher 03:34, 10 June 2006 (UTC)


 * Hmm ... you may be right. I suppose "in whole or in part" could be read two different ways (actually, that might be the problem).  CJCurrie 03:35, 10 June 2006 (UTC)


 * Here's where the "undoing the actions of another editor" clause was added on August 29 last year, but it was added to clarify what admins were already doing: "Reverting doesn't simply mean taking the previous version from history and editing that. It means undoing the actions of another editor, and may include edits that mostly undo a previous edit and also add something new ..."
 * This was intended to include complex, partial reverts, where editors revert war by first reverting one part, then changing tack and reverting another, and mixing and matching to game the system. (I'm not saying this was done in the case in question, because I haven't looked at it: I'm making a general point only.) SlimVirgin (talk) 03:57, 10 June 2006 (UTC)


 * That may have been standard practice before August last, but I'm not certain it was formally agreed to as policy.


 * Anyway, I'll reiterate my previous point -- given the level of confusion concerning the *current* 3RR, it may be useful to revisit the issue in a comprehensive matter. I think TU's point about an announcement of policy may be useful as well.  CJCurrie 04:01, 10 June 2006 (UTC)


 * The other case is where added material is simply rewritten - e.g. "Hitler was a lifelong Catholic", "The Fascist dictator never renounced the faith of his childhood", "The Führer remained loyal to the Pope of Rome", etc. (although I suppose "The" is present in all three...) Conversely, if other editors are rewriting their passages, continued removal should also count.  However, this merits more specific language than "undoing other editors' work".Timothy Usher 04:07, 10 June 2006 (UTC)


 * Hi CJ, it's been standard practise ever since I became an admin, as I recall. It would make no sense to have a 3RR policy that only included simple reverts of the same material, because it would be too easy to game. All you'd have to do to repeatedly change a section but avoid 3RR, is restore (or delete, depending on what you're trying to achieve) one sentence from it, making a few other edits at the same time; then restore another sentence, also making a few other changes during the same edit, and so on. There would be no end to it if the editor was tricky enough. That's why admins look at the number of times an editor undoes another editor's work on the same article within 24 hours. The point of 3RR is to stop continual reverting of people's work, not only to stop deletion/restoration of exactly the same material. SlimVirgin (talk) 04:10, 10 June 2006 (UTC)


 * I understand this point, and I'm not opposed to it in principle -- I'm just not certain it was ever formally accepted as policy.


 * The thing is that policy is what admins are doing in practise. If admins are doing it, and the policy page states it, then it becomes policy by definition. If you want to propose a change, that's fine, of course. Bear in mind too that there were many discussions on AN/3RR during the first few months of its existence, and the policy was moulded by those discussions, and by what was happening in practise as people got used to enforcing it, and seeing what worked and what didn't. It's more of an organic thing than having something formally accepted. SlimVirgin (talk) 04:18, 10 June 2006 (UTC)


 * The problem is that the current language (useful though it may be) is considerably different from the one I familiarized myself with when I became an admin. I'm concerned that certain assumptions or interpretations by particular editors may have become "policy" by default, without due process for other views.  In any event, I'll also reiterate my other suggestion that admins use discretion in blocking editors for perceived 3RR violations, at least in circumstances where the violation may have been accidental or based on an outdated interpretation of the policy.  CJCurrie 04:27, 10 June 2006 (UTC)

FWIW, I think the clarification helps. The spirit of the 3RR has always been "the 3RR is intended to stop edit wars" - sometimes its not the exact text being reverted, but another change elsewhere that continues the conflict. The wording as it is now may help prevent some edit wars by clearly stating that gaming the system isn't going to get you anywhere. Shell babelfish 04:22, 10 June 2006 (UTC)

I'm not at all opposed to "any reverts count" if that's the consensus; what I'm saying is that as long as I've been an administrator it hasn't been what's been communicated to me in 3RR situations. I can't and won't accept blame for not having previously been familiar with unwritten conventions that differ from the actual practice that I've actually seen undertaken in the specific situations I've been personally involved in. Whether it's always been the spirit of the policy or not, it hasn't been what people have told me when I've been involved in discussion or clarification of the policy. It hasn't been what people have done in the situations I've seen.

Basically, I have to agree with CJCurrie that certain assumptions or interpretations by particular editors may have become "policy" by default, without due process for other views (or, for that matter, any attempt to ensure that people were actually all on the same page.) I'm being faulted here for acting precisely as I'd previously been told was correct, and that's simply not acceptable. Bearcat 05:38, 10 June 2006 (UTC)


 * Bearcat, let's keep a sense of perspective: this is much bigger than your alleged fault, and even bigger than twelve editting hours of HOTR. Not that you don't see this, but just emphasizing, this is about what rules should apply to all editors of Wikipedia.  I'm for Homey's block, and agree that it proceeds from a straightforward understanding of current policy (i.e., this page) but opposed to exactly that understanding of policy.


 * It was no big deal, apparently, to change it in May, and I don't see why it's any bigger a deal to change it it June. Let's keep SlimVirgin's improvements while solving the misundertandings that have been the unintended results thereof.  Whosoever does this - maybe me - posts on WP:AN.  We've some semblance of consensus on this talk page, and that's all we need.Timothy Usher 06:42, 10 June 2006 (UTC)


 * Oh, I'm not failing to keep things in perspective — "this is about what rules should apply to all editors of Wikipedia" was exactly my point in the first place. There's been a lack of clarity about what rules applied, but certain people are wrongly dismissing that as the individual misunderstandings of individual editors, instead of realizing that there actually needs to be significantly more clarity in how the rules are communicated. Bearcat 08:59, 10 June 2006 (UTC)

Since I added some of the disputed text, I'll comment. I'm still in favour of the bits I added. Though 2RR sounds good too. There is, still, even with stricter interpreation, too much edit warring. The overriding rule, IMHO, has always been "don't get close to 3RR unless you want to risk trouble"; if you're sailing close to the wind then you need to be very sure of the rules William M. Connolley 10:56, 10 June 2006 (UTC)


 * Your statement strikes me as somewhat odd in light of this edit, recalled by (now de-sysopped) NSLE's recent action, .    The problem with the vague wording is precisely that complaints against editors who routinely sail too close to the wind are upheld or dismissed based on inconsistent interpretations of 3RR, arguably depending on the attitudes towards reporting or reported editors held by whoever is working the noticeboard at a given time.Timothy Usher 18:38, 10 June 2006 (UTC)


 * Timothy, I'm not sure I see any of the current wording as vague. Could you say which phrase(s) exactly is/are vague, and could you give an example of when it would not be clear what they meant? SlimVirgin (talk) 18:49, 10 June 2006 (UTC)


 * I don't understand TU's comment. 3RR is still not punitive; with the page protected, a block is pointless. Thats pretty std too William M. Connolley 19:01, 10 June 2006 (UTC)

I'd like people to not approach this legalisticly. Lately I've seen several variations of, "okay, that's my three," and "woops, rv self - have to wait another ninety minutes." I'm afraid making the "rule" more explicit will make that problem worse. My feeling is, if you have undone another's work three times in the last day, you are wrong. That is not to say someone else may not be more wrong. It's not that I block for 3RR; I act to stop disruptive edit warring. I might block one malignant troll or two happy warriors, or I might lock the page, or I might ask two reasonable people to not edit the page for a day. If we want to craft a careful definition of "revert", and make a block after four reverts automatic, then let's just do it in software. Tom Harrison Talk 19:27, 10 June 2006 (UTC)
 * "My feeling is, if you have undone another's work three times in the last day, you are wrong."- the problem is when you say "work", it could be just a very minor edit, but an edit INTENDED to be reverted by another user which that editor is hoping to catch out, such as mentioned by Homey below, adding a dispute tag, adding a link, even just changing spelling from British to American english. Arniep 19:34, 10 June 2006 (UTC)
 * This comment looks like an assumption of bad faith without any supporting evidence whatsoever. Pecher Talk 19:45, 10 June 2006 (UTC)
 * I think there is a good reason to suspect bad faith on Zeq's part considering Homey had tried to implement a permanent block on him for his violation of Arbitration Committee probation only 9 days previously. Arniep 20:04, 10 June 2006 (UTC)


 * That's an astonishingly conspiratorial view of another editor. One could as easily make the opposite argument, that Homey was trying to entrap Zeq. Jayjg (talk) 19:59, 11 June 2006 (UTC)

This "policy" needs urgent review
I think we need to realise that certain users or groups of users may be using 3RR as a form of attack to intimidate or silence editors who oppose their position. I think 3RR should take into account the actions of all users involved in reverting, otherwise a user or group of users could make a number of edits deliberately designed to stir up another editor. Then, if a user reverts more than three of these they will get blocked, but that person or persons can then continue to make aggressive edits. This kind of behaviour goes on all the time and this must be taken into account before blocking the supposed "guilty" party. Arniep 12:11, 10 June 2006 (UTC)
 * 3RR was always a sticking plaster, both to policy and to individual disputes. Somewhat effective though.  Rich Farmbrough 14:36 10  June 2006 (GMT).
 * The policy is completely flawed as the person who makes the original edit can revert the 3 reverts of their edit by another user, but if the other user then reverts again to restore the article as it was before the first user's edit, they will have violated the 3RR "rule" and get blocked. Arniep 16:17, 10 June 2006 (UTC)
 * There is indeed a built-in favor of the one who is adding new material. If material which has been removed is being in some fashion re-added, the first edit is a revert to a previous version.  There's no formal timeframe into which previous versions must fall.Timothy Usher 18:24, 10 June 2006 (UTC)
 * There shouldn't be a "built in" favor to anyone. An editor can use the 3RR "rule" to protect themselves by making a provocative edit which the user knows full well will be reverted by another user and the user will "win" either way by reverting back to their version or getting the other user blocked. Arniep 19:16, 10 June 2006 (UTC)

In my case, I reverted 3x, I then added an "original research" tag. The tag was removed, I put it back and whammo, I was technically in violation of 3RR and banned, unbanned, and then rebanned. Seems a bit extreme to me and seems to me like something 3RR was not intended for. Homey 19:05, 10 June 2006 (UTC)


 * That's exactly what 3RR was intended for, Homey: to stop editors constantly reverting to their preferred versions, whether or not it involves adding/deleting the same or different material. 3RR says each editor may do this three times within 24 hours only, but even then, should try to avoid it). SlimVirgin (talk) 19:36, 10 June 2006 (UTC)
 * As it stands the 3RR "rule" can be used maliciously and in a petty manner, because it is policy, not a guideline, a user knows that if they catch another editor unawares, they can trip them up so they fall onto the trip wire and WOOOO BANG, look what you've done, now your blocked and you can't even post on the noticeboard about it. You are automatically labelled the party in the wrong even if the other user made the edit in a deliberate attempt to start an edit war which the user knew that they cannot lose. Arniep 20:21, 10 June 2006 (UTC)


 * Ridiculous are the accusations that people misuse the 3RR to intimidate others. The policy exists to be enforced; if it isn't, it might just as well not exist in the first hand. Pecher Talk 19:43, 10 June 2006 (UTC)
 * Yes ridiculous isn't it the idea that Zeq had absolutely no reason to feel hostility towards Homey who had tried to implement an indefinite block on him 9 days previously per Zeq's Arbitration Committee probation on Israel related articles. It seems quite clear that Zeq's reporting of borderline 3RR was retaliation for that. Arniep 20:01, 10 June 2006 (UTC)
 * It doesn't matter whether Zeq wanted to retaliate or not; the policy must be enforced regardless of the motivations of the people involved. You may be right that Zeq didn't and doesn't harbor any good feelings towards Homey; I assume nobody would harbor good feelings after being treated in such a manner. However, there is no clause in WP:3RR that a report must be discarded because the reporting editor has bad feelings towards the offender. If there were such a clause or if it were an acceptable practice among admins, the policy would be a joke. Pecher Talk 20:30, 10 June 2006 (UTC)
 * Sorry but of course it matters. People can easily trip people up with this "policy" because it is a "policy" not a guideline which is what it should be. The policy as it stands is just a way that certain users try to use to get other users blocked with whom they disagree. Arniep 20:46, 10 June 2006 (UTC)


 * Whilst I think reverting is in general a bad idea, except in the case of vandalism ,I do feel this rule encourages full page reverts as an editor can save their "allowance" and hit all the edits they don't like in one go. I have seen it used to penalize an editor trying to work for concensus whilst another editor stayed out of the way for an hour or so and just wiped the lot in one go. Also I have had a situation where I have been editing and another editor has made an edit and saved before I did so are below me in the edit history. This would break my run of edits unknown to me and so I could then violate 3rr unintentionally. The best rule is to revert once if you feel you have to and then make a lot of noise on the talk page to attract other editors and so get a more varied input into the situation. Sophia  19:53, 10 June 2006 (UTC)


 * Hi Sophia, I agree about other editors editing behind you during a long run of edits, leading to an inadvertent 3RR violation. Admins who have experience administering 3RR are usually able to see when that's happened. I agree that it's annoying. SlimVirgin (talk) 20:08, 10 June 2006 (UTC)


 * If you made four reverts to different sections in a row, say at 18:11, 18:13, 18:16, and 18:18, and if they were reverts that you could have done as a full page revert but didn't want to because you wanted to leave some good changes in place, and if someone made a change to a different section at 18:17, but there was no edit conflict because it was a different section, then I would treat those four as one revert if it were reported at WP:AN/3RR. If, however, you made three consecutive reverts (counted as one), went off and had dinner, came back two hours later, and edited a few more articles, and then went back to the original article (which several other people had edited in the meantime) and reverted something that you could have reverted earlier, then I'd count that as a second revert.  I don't think any admin would knowingly block someone who had reverted different sections one after the other as an alternative to a full page revert (which would have reverted some good changes), and who had been unaware that another editor was editing at the same time. AnnH ♫ 20:19, 10 June 2006 (UTC)


 * I've had a situation where I was copy editing, and was doing it bit by bit, so I had a long run of edits in a row. One of the things I was doing was deleting red links because the article had a lot of them. Unknown to me, someone was restoring the red links as I was editing. We weren't getting edit conflicts because we just happened to be saving at slightly different times. I kept seeing the red links back in place and I assumed there was a server problem, or else that I thought I'd deleted them but in fact hadn't, so I kept removing them again. Eventually, I realized someone else was restoring them as I was going along. :-) I don't think any admin would block in that situation because it's pretty clear it's an error. SlimVirgin (talk)  20:24, 10 June 2006 (UTC)
 * The more I think about this "policy" the more ridiculous it seems. If I remove 4 pieces of information from four different users from an article in one day is that a violation of 3RR (would I have to check if the four pieces of information were from different users?)? Does it make a difference if I remove them in one edit or 4? If I removed it in one, could I then be reverted three times and then revert to my version again and not violate 3RR? Yet, if I had removed the same information in two edits, I could not revert to my version without violating 3RR, quite frankly the whole policy is a nonsense. Arniep 20:46, 10 June 2006 (UTC)
 * Well, nonsense would be an inaccurate term to use here. The whole policy seems grossly inadequate in the description of the rules and conditions which are presented here. -- S iva1979 Talk to me  20:54, 10 June 2006 (UTC)


 * The way I understand it is that if you remove the edits of four other people one after the other with other editors making changes in between then you are in violation of 3RR. If you remove all four editors work in one edit (easiest done as a revert to an earlier version) then you are not in violation and still have two reverts left (if you look at things in that way). Hence my concern that full page reverts are encouraged by the way this policy is currently written.  Sophia  22:12, 10 June 2006 (UTC)
 * Exactly the policy is silly because it can be gamed, i.e. users who are "wise" know not to make many edits but to make edits "en masse" which gives them more power to retain their version against another users reverts. IMO the main problem is that it is a policy not a guideline, which means it can be used as tripwire in disputes to silence a critic, whereas in truth, everyone who engages in an edit war are equally responsible for it as more often than not it occurs without discussion or seeking mediation. Arniep 22:24, 10 June 2006 (UTC)


 * Well, in the example I gave above, if you revert the edits of four people in close succession, but as four separate edits, and one or two editors manage to slip in an edit in the middle of your reverts without causing an edit conflict (because it's a different section) an administrator would look at the history and would see that you intended it as an alternative to a full page revert. But if you went off after three real reverts (the kind that would be counted) and had dinner, came back and edited a few different articles, returned to the one you had reverted on, saw a recent addition that had been there when you were reverting earlier, thought, "Oh, I don't like that either," and reverted it, you might get caught. The difference is that in the first case, you're doing separate reverts in preference to a full page revert which would count as one, but which would undo some good changes, whereas in the second case, you're just deciding to make another revert. I don't think we need to have it absolutely laid down in stone, as there is room for the administrator's judgment. Remember you can be blocked for one revert, and you can be let off for four. What we need clarity on is the fact that reverts of different sections count, and that reverting the removal of a tag counts. By the way, you can also do a full page revert with an edit summary "Rv to Str1977. Will re-instate the good changes in a moment." AnnH ♫ 22:42, 10 June 2006 (UTC)
 * "there is room for the administrator's judgment. Remember you can be blocked for one revert, and you can be let off for four." - this seems to be even worse than a fixed rule, please see my proposal below. Arniep 23:40, 10 June 2006 (UTC)
 * AnnH, it would make sense not to ask only if the reverts were successive, but also if they were contested between the relevant sessions.Timothy Usher 03:07, 11 June 2006 (UTC)

A brief history of WP:3RR as it relates to the matter at hand
Having gone through the history, I can offer the following report describing the evolution of this policy page as it relates to the matter at hand. Please feel free to correct any errors or add relevant points that I’ve missed.


 * Let’s look at the very first version of this page . It’s clear that the writer had simple version reverts in mind.   Over time, the writing of the policy evolved to prevent gaming of the system.  We’ve not the space to discuss all these changes, so I’ll highlight those which seem most relevant to this discussion.


 * Jayjg’s edit of 18:12, 6 January 2005 18:12, 6 January 2005 changed "reverting the same page” to "adding/deleting a specific piece of article text", reflected what is now the minimal version of the policy. This was removed within minutes.


 * Tony Sidaway’s edit of 01:12, 2 February 2005 added "It does apply to all other reverts on that page during the twenty-four hour period, not just carbon copies of the original revert". This, too, was removed in short order.


 * Radiant!’s edit of 09:59, 29 August 2005 added "Reverting doesn't simply mean taking the previous version from history and editing that. It means undoing the actions of another editor, and may include edits that mostly undo a previous edit and also add something new, page moving, admin actions such as protection, etc.". Unlike Jayjg and Tony Sidaway’s previous attempts, this stuck, and was reinforced by SlimVirgin on 02:06, 30 August 2005, which changed "Don't revert any single page more than three times..." to  "Don't revert any single page, in whole or in part..." and on 08:10, 3 January 2006.  At this point, there can be no doubt that what is now considered the minimal interpretation of 3RR, including any related reverts is established policy.  Several attempts to remove this language were reverted by William M. Connolley and Tom harrison on  on 21 February of this year.


 * William M. Connolley’s passage of 09:53, 13 April 2006, which states, "Note that, for the purposes of counting reverts, successive edits by the same editor are considered to be one...", suggests that he had arrived at a novel interpretation of the Radiant/SlimVirgin policy, as this clarification is completely unnecessary if the reverts must be related (and indeed, as per Sophia’s objection, has some odd side-effects). The interpretation stems from the ambiguity of "Don't revert any single page, in whole or in part..." when what was really meant was, "Don't revert any part of any page..."


 * SlimVirgin’s edit of 04:14, 15 April 2006 reads, "A revert may involve as little as adding or deleting a few words or even one word. Even if you are making other changes at the same time, continually undoing other editors' work counts as reverting. "Complex partial reverts" refer to reverts that remove or re-add only some of the disputed material while adding new material at the same time; this is often done in an effort to disguise the reverting." This wording seems intended to prohibit gaming the system through such complex partial reverts, but inadevertantly reinforced the maximalist interpretation with the passage “...continually undoing other editors' work counts as reverting."


 * William M. Connelly’s edit of 15:09, 18 April 2006 added the passage at issue, "There is no requirement for the reverts to be related: any four reverts on the same page count." We now have the maximalist interpretation of 3RR written into policy.  SlimVirgin’s edit of 03:47, 20 April 2006 might be construed as an endorsement of Connelly’s change.

Summary: The original version of the page addressed simple version wars. The inadequacy of this policy to deal with partial reverts was recognized, and several attempts were made to change it before one version, by Radiant! and SlimVirgin stuck. The ambiguities of the language used to prohibit partial reverts created misunderstandings, and ultimately to an expansion of 3RR to include unrelated reverts.

Current problem: This maximalist interpretation is neither broadly understood nor broadly accepted, and has led to the inconsistent application of the policy, an undesirable level of debate on WP:ANI/3RR, wheel-warring and accusations of unfair treatment.

Recommendation: Remove the language counting unrelated reverts, address ambiguities in the partial-revert language which might be construed as supporting the broader interpretation, and post a notice of the changes on the relevant noticeboard(s).Timothy Usher 21:12, 10 June 2006 (UTC)


 * Thanks for writing this out, Timothy, although I still don't understand how "Don't revert any single page, in whole or in part..." is ambiguous. Also, it's worth noting that at no point in its life did the policy say the reverts had to be simple reverts of the same material; if you look at early discussions about the policy when it was first introduced, you'll see that partial reverting was meant to count as reverting too. SlimVirgin (talk) 21:30, 10 June 2006 (UTC)
 * The question is not whether the wording is ambigious, it is whether the policy is a good policy or not. Per my comments above, it is not. Arniep 21:44, 10 June 2006 (UTC)


 * It's ambiguous between "Don't revert any parts (related or not) of any page." and "Don't revert any particular part of any page (even if this reversion is only part of your edit)." Perhaps I'd wrongly read your intent in the passages related to "complex partial reverts", but the second interpretation seemed to me most natural, and it seems I'm hardly alone in this.


 * If the intention of this change were clearer, it might not have stuck.


 * I've been aware of the maximalist interpretation for some time, and in fact reported another editor for it, only to have it dismissed as not counting because the last revert wasn't related - which after seeing other editors get blocked for precisely this, seemed rather unfair.


 * Note that if we're to maintain the maximalist interpretation, we might need multiple "previous version" lines on WP:ANI/3RR, for if the reversions are additions to the article which had previously been removed, these sections may never have coexisted (as came up in the dismissed report to which I'd referred).Timothy Usher 21:45, 10 June 2006 (UTC)


 * I'm sorry if I'm being dense, but I still don't see the difference between: "Don't revert any parts (related or not) of any page," and "Don't revert any particular part of any page (even if this reversion is only part of your edit)." SlimVirgin (talk) 21:54, 10 June 2006 (UTC)


 * The latter would require that some version of the same material is being reverted (incidentally, it should be specified that, in the case of disputed additions, creative rephrasing still counts).Timothy Usher 21:59, 10 June 2006 (UTC)


 * Okay, thanks, I think I see now. There was never anything in the policy, implied or explicitly stated, that said it had to be the same or related material. The policy was always simply "don't revert more than three times in 24 hours on any page." SlimVirgin (talk) 22:03, 10 June 2006 (UTC)


 * I accept that this was your intent, but first, it clearly wasn't policy throughout most of 2005, or Jayjg and Tony Sidaway's edits wouldn't have been immediately removed without further contest. Radiant1's change of 09:59, 29 August 2005 does not, by the most straightforward readings, appear to support the maximalist interpretation (though you may know what he was thinking?), but is only clarifying that page moves etc. are also version reverts (as they obviously are), and that partial reverts of the same material count ("may include edits that mostly undo a previous edit and also add something new.")  Your subsequent change of "Don't revert any single page more than three times..." to  "Don't revert any single page, in whole or in part..." is most naturally construed as an attempt to summarize the partial-revert rule, as is your "complex partial reverts" change of 15 April.  It wasn't at all clear until William M. Connolley’s change on 18 April that unrelated reverts were to be counted.Timothy Usher 22:28, 10 June 2006 (UTC)


 * Unrelated reverts are still reverts. Only a second rate and below edit warriour would be silly enough to try and rule lawyer a case otherwise.Geni 03:09, 11 June 2006 (UTC)
 * Several editors have avoided blocks through precisely such lawyering, and this entire discussion was prompted by wheel-warring over HOTR's most recent block, a complaint which was initially dismissed on the basis that the reverts were not related. Second-rate and below arguments they may be, but they've not been brushed aside as you'd imagine.Timothy Usher 03:18, 11 June 2006 (UTC)


 * I can show that arbcom have decided that unrelated revets count. Or at least endorsed my actions based on them doing so. So some admins are not fully skilled at dealing with rule lawyers. that is hardly the fault of the rule.Geni 03:21, 11 June 2006 (UTC)

I think there is too much attempt to pin down precisely what the rules are. The rules inevitably are, and indeed should, remain somewhat vague - otherwise you get people going just up to the boundary. 3RR is intended to prevent edit warring. Also, I think the level of problems this entire matter is causing, either way, is exaggerated.

As for my 18th April edit... that went in because it was how I was interpreting policy, and how I thought others were too. Had people disagreed, then we would have discussed it. But it stuck. If people *now* disagree it could come out again... but leaving what? Even more ambiguity. I admit that my two paras are somewhat contradictory. William M. Connolley 22:40, 10 June 2006 (UTC)


 * It did stick, but clearly there are many who aren't aware of it, and several valid reports were dismissed because of this. I propose a notice on WP:AN clarifying the policy as it stands, and inviting comment.


 * As for ambiguity, I can rewrite it to be crystal-clear, whichever interpretation is to be used. It's easy to specify that reverts must include common material or creative recastings thereof.


 * We might also ask if the number of reverts allowed should be reduced to two, particuarly if a stricter standard is to be applied to what counts.Timothy Usher 23:15, 10 June 2006 (UTC)


 * The interpretation that has always existed and been enforced is that any reverts count, and 3 as a "reasonable" upper limit (though not an entitlement) was promoted by Jimbo and voted on by a large consensus. Jayjg <small style="color:darkgreen;">(talk) 19:44, 11 June 2006 (UTC)

Proposal to scrap 3RR
I propose that 3RR be scrapped entirely. The problem which we have is edit warring, which is something which all users engaged in it are equally responsible for. At the moment an edit war can be deliberately precipitated by a user, knowing that they will be able to revert 3 times back to their own version, or get other users to do it, causing the person who opposes their view to be blocked. Alternatively a user can for example in a separate session, knowing that a user has made 3 reverts earlier in the day which they have perhaps forgotten then make a deliberately provocative edit bordering on vandalism, knowing the person is working on the article at the time, in the hope the user may trick the other user into reverting to cause the user to be blocked. To avoid 3RR being used vindictively in this manner, I propose that revert war templates be created that can be posted on the talk pages of all users involved and the article in question which will encourage users to discuss the matter on the talk page before editing the article, the disputed material will then be removed or kept by the "edit war" admin depending on how long the information had stood in the article, i.e. if it stood in the article for more than a month, it should be included, if less, removed. Reverting of the "edit war" admin then becomes a blockable offense. Discussion should then take place on the talk page, possibly with the "revert war" admin mediating. If that does not succeed, it should go to one of the other mediation channels. Arniep 23:30, 10 June 2006 (UTC)


 * We don't have the rescourses and in any case that kind of disspute resultion isn't realy an admin task. The 3RR can be though of as an antibiiic that kills of a certian very pointless form of edit waring. Our more skilled edit warriours have of course adapted to deal with it over time.Geni 23:55, 10 June 2006 (UTC)
 * But the status quo isn't acceptable as it is often people who are acting in bad faith who are the ones to report 3RR after having precipitated the dispute, and as they made the first edit on that day, they are "in the clear". Maybe we should have a different rule in that if an edit results in an edit war, the article must be returned to it's established version while the dispute is being resolved either through discussion or mediation. This would stop the silly number games of counting whether this or that edit counts as revert or not and would stop the 3RR being used as a "trap" to catch people out by bad faith users. Arniep 01:02, 11 June 2006 (UTC)
 * The three revert rule is prehistoric. Originaly it was not infoceable by block. Before then it was ignored from time to time the results were not pretty and would be far worse today. Ultimately the three revert rule works as a block on blunt force reverting which is needed.Geni 01:43, 11 June 2006 (UTC)
 * So basically you are saying it's fine that the 3RR rule is used aggressively? Arniep 07:54, 11 June 2006 (UTC)
 * It is not reasonable to suppose that an editor may "trap" another editor into a 3RR violation.Timothy Usher 08:32, 11 June 2006 (UTC)
 * It is reasonable as it happens all the time. A user knows that if they make an aggressive controversial edit which they know another editor with whom they have been in dispute will object, they will succeed in restoring their version without breaking the 3RR "rule", whilst if the other user tries to restore the article to before the controversial edit was made they will be blocked. At the moment this policy favours aggressive editing and using 3RR as a group bullying tactic. Arniep 12:04, 11 June 2006 (UTC)
 * That's not trapping anyone into a violation, as they're always free to let the changes stand (just as they were before they learned about Wikipedia), but it certainly would gain an advantage in a content dispute. That happens all the time. This would seem a minor fix compared to, say, a request for page protection.  Of course, in the latter instance we can say we've "stopped an edit war", but only vacuously so.  If 3RR is too often the final arbitrer of content, it's because one can't currently force editors to join the talk page, or to be reasonable once they're there.  It only takes one such user to make compromise impossible.  Conversely, many pages have been violating policy for so long that if another editor comes along and tells them to use reliable sources (to name just one policy that's generally unenforced, say as opposed to WP:3RR or WP:NPA), he'll be told to take a hike.  In both cases, the way it's likely going to be settled is by 3RR.  Whether we'd describe this as a "group bullying tactic" I think would likely depend on our view of the content dispute.  The problem would seem to lie not in any deficiency in WP:3RR, but in the lack of enforcement of content policies and the non-existence of common-sense user conduct policies sufficient to guarantee that problem editors will be bumping up against something other than 3RR.Timothy Usher 18:52, 11 June 2006 (UTC)
 * But don't you see what I am saying- as it stands at the moment, the 3RR encourages bad faith users to make aggressive edits WITHOUT ANY DISCUSSION BEFORE MAKING THE EDIT. I propose that the we count the original controversial edit as a revert, so if the other user feels strong enough the article will be returned to its original state before the controversial edit, and then the users will have to discuss on the talk page. Arniep 22:23, 11 June 2006 (UTC)

It's no point arguing the toss, Arnie. The whole thing is utterly ridiculous. We have content decided by commitment, not by whether it's actually any good. You simply will not succeed in any attempt to change that because settling content by commitment suits the committed, who are a/ empowered and b/ more likely to turn up to votes on it and most importantly c/ more concerned with "winning" than the outcome's actually being good. -- Grace Note.

A more modest proposal
Putting the broader debate aside for a moment, may I recommend that we exempt non-content edits (NPOV templates, citation requests, footnoting corrections, etc.) from the 3RR? It seems to me that the rule was created to prevent content disputes from getting out of hand, and I don't believe that technical adjustments need be counted toward this end. CJCurrie 03:34, 11 June 2006 (UTC)


 * What would then stop editors from fighting over the tags all day?Timothy Usher 03:41, 11 June 2006 (UTC)

Common sense, in most cases. I can't say that I've seen many tag disputes since I joined Wikipedia. If you think it would be a problem, we could always create a separate revert rule for non-content edits.

My point is simply that it seems foolish to prevent someone from making technical corrections even after reaching the three-revert limit. CJCurrie 03:48, 11 June 2006 (UTC)


 * This might be viewed as a side-effect of the interpretation under which unrelated reverts count.


 * I've seen quite a few tag disputes, usually related to the POV tag.


 * Is it foolish to restrict it? Well it's foolish to prevent someone from making any correction or improvement after hitting the 3RR limit, but so long as policy (excepting vandalism) studiously avoids referring to the substance of content disputes, we're in no position to determine what constitutes a correction, or an improvement.Timothy Usher 04:03, 11 June 2006 (UTC)

Perhaps we should specify separate revert-rules for content and non-content edits. I still believe that the current system can lead to absurd situations. CJCurrie 04:33, 11 June 2006 (UTC)


 * I haven't seen any absurd situations. The only except to 3RR at the moment is blatant vandalism, and that seems correct. Creating awhole new pile of different rules would be a bad idea. Rv'ing POV tags is pretty common William M. Connolley 07:28, 11 June 2006 (UTC)
 * You don't have to create a "whole new pile of rules" you just have to get rid of this one. Unfortunately this is an absurd bureacracy that benefits those who are wise to the system like a lot of the "rules" on Wikipedia (and no they don't help to make a better encyclopedia). Arniep 07:56, 11 June 2006 (UTC)


 * I've seen plenty of revert wars over non-content edits. SchmuckyTheCat 15:03, 11 June 2006 (UTC)

William's "change"
It has been claimed or implied a few times at this page and elsewhere that William M. Connolley "changed" the policy on 18 April by adding:
 * There is no requirement for the reverts to be related: any four reverts on the same page count.

In no way was that a change to policy. It was a clarification. I was already aware that different reverts counted, and had been, for as long as I can remember. That was what was in practice, and the wording of the policy page prior to that did not say that the reverts had to be related; it was at least possible to guess that you could be, but extra clarity was needed. If the policy says you mustn't revert more than three times, and it doesn't say "you musn't carry out the same revert more than three times, and admins are blocking for unrelated reverts, and then someone makes the policy page clearer by specifying that the reverts don't have to be the same, and there's no protest, then it shouldn't be called a change in policy. I explained that particular aspect of 3RR on numerous occasions before 18 April, and I can recall people being blocked for different reverts long before 18 April. William's edit was helpful in ensuring that people would not misunderstand the policy; it was not a change to policy. AnnH <b style="font-size:medium;">♫</b> 07:32, 11 June 2006 (UTC)
 * "I was already aware that different reverts counted, and had been, for as long as I can remember. That was what was in practice" how are mere ordinary users (even admins!) who are not in the right circles meant to know what some admins always knew in their own head? I fell foul of this myself, I read the policy some time ago, and I regularly watch the pump and relevant pages for new policy information, yet it was never clear that 3 reverts on anything at all counted towards the 3RR " rule". Arniep 07:50, 11 June 2006 (UTC)
 * I would say that before William made it more explicit, it was perhaps not quite as clear as it should have been. However, I read it a year ago, and have read it many times since then, and I have always known that reverts to different sections counted. And I knew that simply from reading the policy page, not from some secret info that only the elite have access to. I also knew from the 3RR noticeboard that that was the practice.  William's addition did not contractict an old rule; it did not change an old rule. It just meant that people would have even less excuse than they had had before to claim that they didn't know. AnnH <b style="font-size:medium;">♫</b> 09:58, 13 June 2006 (UTC)
 * Errrm... so is all this talk just cos you're annoyed about being blocked? William M. Connolley 08:01, 11 June 2006 (UTC)
 * Please don't try and deflect criticism by attacking another user, you were the user that added the text without making any announcement so that other users could see a change to the policy (yes, a change as users shouldn't be expected to read the minds of admins). Thanks Arniep 11:58, 11 June 2006 (UTC)
 * How could that possibly be a change unless the rule before that said that no editor may perform the same revert more than three times in a 24-hour period, or no editor may revert the same section more than three times in a 24-hour period? It didn't say anything like that, so there was no change in policy. If there had been, his change would certainly have been reverted, as many, many admins watch this page, and would certainly oppose an attempt to change policy without discussion. It's unfortunate that his excellent clarification is being used to imply that the rule was changed. Let me state this clearly: when people were blocked for making four unrelated reverts to the same article before William's so-called "change", as indeed frequently happened, there was nothing in the policy page that they could have appealed to. AnnH <b style="font-size:medium;">♫</b> 09:58, 13 June 2006 (UTC)
 * Perhaps, AnnH, but the history of practice is not as easily verifiable than the history of the policy page: my report was explicitly based upon the latter. To the extent that the policy page didn't reflect prevailing understanding and practice, my report is thus by nature deficient.Timothy Usher 08:29, 11 June 2006 (UTC)
 * When I wrote my post in this section, Timothy, I didn't have your report in mind. I was thinking of Bearcat's claims here that the rule was changed two months ago. I noticed William's change at the time, and I thought, "that's a good idea to make it clearer;" I didn't think, "that's good to bring in that rule." In fact, the three clarifications that William has made in the last few months (replacing or removing POV tags is not exempt from 3RR, reverts to different sections count, and consecutive reverts with no intervening edits count as one) have all, in my view, been excellent additions to the page, reflecting actual practice, as understood for over a year by administrators who help out at the 3RR noticeboard. People who intend to revert a lot, and admins who intend to step in and unblock really do need to ensure that they're familiar with the policy. AnnH <b style="font-size:medium;">♫</b> 17:43, 11 June 2006 (UTC)
 * Sorry but it doesn't matter what some admins thought the rule should be or always was, if admins don't write it down on this policy page it is not reasonable to expect ordinary Wikipedians to know "what it really was". Arniep 17:48, 11 June 2006 (UTC)


 * However, in this case, the user is a longstanding editor and admin, who had the rule explained to him again just days before he was blocked; thus there is no excuse here. Moreover, inexperienced editors are always warned before being blocked; it is only experienced editors who are not afforded such niceties.  In addition, it always was written down on the page; the fact that it has since been made even more explicit does not contradict the fact that it has always been there. Finally, it is a complete abdication of personal responsibility to state that someone can be "goaded" into 3RR, not to mention a rather conspiratorial view of Wikipedia that violates WP:AGF, another fundamental Wikipedia policy. Jayjg <small style="color:darkgreen;">(talk) 19:48, 11 June 2006 (UTC)
 * Firstly I was referring to my case, when I was reported for violating 3RR twice, and only the second time was it actually explained to me that any 4 reverts counted and this was never obvious until William Connelly wrote on this policy page in April whereas I had read the policy some time earlier. The very fact that William had to state it in in bold must have meant that it was also not obvious to many other users, (even admins!). Re:AGF, I don't see why good faith should be assumed for a user on probation on the very articles in question, and which he has tripped two users into 3RR, and then run off straight away to report them like a little child. Arniep 22:15, 11 June 2006 (UTC)
 * I'm glad you added "like a little child", that really helped the tone here. Once again, no one can "trap" or "trip" others users into violating 3RR.  Let's try to steer the discussion away from this particular dispute.Timothy Usher 22:30, 11 June 2006 (UTC)
 * Well, sorry but the fact Zeq is on arbitration committee probation doesn't exactly give an indication that they have a very good record of adhering to policy does it. I see very little "assuming of good faith" by admins in other cases where users have been as troublesome as Zeq. Arniep 23:10, 11 June 2006 (UTC)

Yet another bit: does a block "clear" all previous violations?
OK, so rather than add this directly I'll discuss first. Subject: does a block "clear" all previous violations? Example: user X reverts 4 times; is blocked (for less than 24h); returns and rv's again, such that this gives a new set of 4R in 24h, though 3 of them are previous to the block. Is this blockable? I would say yes. Opinions? Is it worth adding (one way or other) to the page? William M. Connolley 18:25, 13 June 2006 (UTC)
 * Yes, I think it should be added to the page in the as unambiguous manner as possible. The mere fact that this question has been raised on talk shows that there may be some ambiguity on this issue. Pecher Talk 19:31, 13 June 2006 (UTC)
 * I would say it does violate 3rr. I have no sense that a block resets some counter of allowed reverts. As I've said before, I'd like to move away from that knid of legalistic thinking. I worry about instruction creep exacerbating what I see as 'revert-counting' to game the system. Tom Harrison Talk 20:05, 13 June 2006 (UTC)
 * I would say definitely yes. Yes to "Is this blockable?", that is, not to "Does a block 'clear' all previous violations?" And where it says that you can be blocked even if you technically stay within four reverts, I would say that the scenario that you've presented is a clear case where such a block would be appropriate. So if someone is blocked for the full 24 hours for making four reverts, comes back and immediately reverts that page again, I think s/he shouldn't be surprised at being blocked a second time. I'm sure I've seen it happen &mdash; perhaps with Chooserr and the BC/BCE dating system, though I may be mistaken. AnnH <b style="font-size:medium;">♫</b> 20:16, 13 June 2006 (UTC)

Complex reverts
I have made a clarification on the policy page.

It doesn't change policy at all, it just makes one aspect of it clearer. This came after User:KimvdLinde explained to me how the policy as it stands could be used to block me.

Consider the following situation:


 * User:Bad removes the date of birth from a biographical article without explanation.
 * User:Good restores the DoB (his 1st revert) with the summary rv unexplained deletion (possible vandalism)
 * User:Bad removes the DoB (his 1st revert) with the summary removing unsourced assertion
 * User:Good restores the DoB (his 2nd revert) with a summary explaining that the DoB is uncontroversial, etc.
 * User:Bad removes the DoB (his 2nd revert) with the summary removing controversial unsourced assertion
 * User:Good restores the DoB and includes a note of what the source of information is.
 * User:Bad removes the DoB and source (his 3rd revert) with the summary removing badly sourced DoB
 * User:Good restores the DoB and includes even fuller references regarding the source of the information.
 * User:Bad removes the DoB and full source info (his 4th revert).
 * User:Good reports User:Bad for 3RR, secure in the knowledge that he himself has reverted only twice, whilst the other guy have reverted four times.
 * Both users are blocked under the 3RR and declared to be as bad as each other.

The policy stated that "‘Complex partial reverts’ refer to reverts that remove or re-add only some of the disputed material while adding new material at the same time; this is often done in an effort to disguise the reverting. This type of edit counts toward 3RR."

The above wording allows the blocking of User:Good in the above situation. However, User:Good could be forgiven for believing that complex reverts only meant reverts that include token extra alterations in order to disguise the revert. (The above paragraph certainly has the intention of insinuating that the purpose of adding new material to an edit is to disguise the revert.) This might confuse good-faith editors into thinking that if new material is added in an effort to reach a compromise or acquiesce to other users' demands, then there is no 3RR violation. That was certainly my belief, and would bet that most editors would assume the same.

In order to make it crystal-clear that good faith is no defence, I have edited the policy page to read as follows: "‘Complex partial reverts’ refer to reverts that remove or re-add only some of the disputed material while adding new material at the same time; this is often done in an effort to disguise the reverting. Note that this type of edit counts toward 3RR whether carried out in bad faith or good, i.e. whether it is done to disguise the reverting or in the spirit of compromise."

This is because I believe that people should be made fully aware that they will be immediately blocked if they act like User:Good in the above situation. — Gulliver <big style="font-family: 'Arial Unicode MS';">✉ 05:03, 14 June 2006 (UTC)
 * I agree with the change and am going to try adding a similar example to the page itself. I have seen this aspect of the policy interpreted very differently by or in reference to different users and that shouldn't be the case since 3RR is one of the few things on Wikipedia that doesn't involve subjective evaluation. --CBD 10:53, 16 June 2006 (UTC)

Repeat reversions vs. different reversions
The policy as it stands seems to tie the hands of editors when dealing with rogue editors. Where an editor can currently make a dozen bad edits, only 3 of those can be removed by another editor taking care of the page. This unnecessarily gives editors going against the community an advantage, where 3RR referring primarily to repeat reversions would remove this asymmetry.--Nectar 20:36, 19 June 2006 (UTC)


 * This is true. You can insert junk, as long as its different, numerous times. However, this rarely occurs, and when it does can be dealt with as vandalism. Finding a situation where people add non-vandalism repeatedly but differently is much rarer; it requires more editors (as you say) or patience William M. Connolley 20:44, 19 June 2006 (UTC)


 * A small number of unreasonable editors are a reality on Wikipedia, but few bad edits can be easily classed as vandalism. In the case where a persistent unreasonable editor remains active on an article for years, it's one thing to put that editor on par with Ph.D.s in the subject, but it's another thing to give that editor a policy advantage over Ph.D.s.
 * Taking into account the argument Raul654 makes in his "First Law" (a small number of dedicated contributors "are the most valuable resource Wikipedia has" and most who leave do so because of trolls/POV warriors) it seems the most reasonable position here would be to err on the side of not increasing the burden on those editors.--Nectar 21:01, 19 June 2006 (UTC)

3RR violation by dispute tag insertion?
Should a 3RR block occur after an editor chooses to insert a dispute tag instead of reverting, when reverting would violate 3RR, inserting the dispute tag returns the article to a previous version, and the dispute tag in question has never been removed except through reverts?

Would such a block be within the letter of WP:3RR which states that reverting "means undoing the actions of another editor"? Why or why not?

Would such a block be within the spirit of WP:3RR? Why or why not?

My opinion is allowing 3RR blocks in such cases would forment more ill-will in the project than allowing dispute tags to be placed on disputed versions even when doing so happens to return the article to an earlier state. I think it is an abuse of the rules (and the definition of the word "action") to consider insertion of such tags "undoing the actions of another editor" when the dispute tag has only been removed through reverts. Others feel that "a revert counts even if the revert is regarding the addition or removal of a tag," and, "any revert, including re-addition of tags counts towards 3RR."

I have asked the same questions of the arbitrators on Requests for arbitration Publicola 18:55, 20 June 2006 (UTC)
 * See Wikipedia talk:Three-revert rule above.Timothy Usher 19:41, 20 June 2006 (UTC)
 * That proposal was a different situation, involving non-content edits. My question is only about inserting a dispute tag, not removing one. Publicola 20:19, 20 June 2006 (UTC)


 * The way you ask the question seems to admit of a simple and clear answer: Insertion of a tag previously removed by someone else, when such insertion returns the article to exactly the state it was in previously, would qualify as a revert in my view. Further, removal of a tag previously inserted by someone else, when such removal returns the article to exctly the state it was in previously, (the opposite direction of what you asked) would also qualify as a revert in my view, it is symmetric. That part's clear enough. In my view any revert, including re-addition or removal, of tags counts towards 3RR. If two or more editors are alternatly adding and removing tags, that strikes me as revert warring within the meaning of the guidelines. I've see Timothy Usher's proposal and I don't think that tags should be exempt from the spirit of not reverting or the text of 3RR. Hope that helps, although I suspect it's not the answer you wanted (especially since you seem to be forum shopping by taking this to several places at once). Rather than pursueing this in depth, consider just not doing it... next time there's a tag war about to break out, take discussion to the talk page and try to work out your differences, or ask an admin or other interested party to step in and try to mediate, help, warn, block, whatever it takes. Edit warring seems to never work out well, and all alternatives should be pursued.  + +Lar: t/c 20:10, 20 June 2006 (UTC)
 * I think you've misunderstood my question. I'm talking about inserting dispute tags, not both inserting and removing them, which would just cause edit wars on the tags.  Moreover, you seem to think that Timothy Usher made the above proposal, when I think you mean to say that he was arguing against it. The idea of "forum shopping" with two messages linked to each other is interesting, but my understanding is that ordinary discussions are discouraged on the Arbitrator's page, although they have the authoritative say in the matter -- I was trying to make sure there was a place that everyone interested could see and discuss this without getting in trouble for cluttering the arbitrator's page (it happens.)
 * Finally, I'd like to point out that often it isn't really feasible to check whether inserting such a tag returns the article to one of its previous states, if the article has more than a few dozen edits. It's wrong to punish those who are making a good-faith effort to comply with the rules. Publicola 20:25, 20 June 2006 (UTC)
 * I did not misunderstand your question, I merely answered it in a broader scope than you asked it, to make a point, repeated: Anything that returns an article to a previous state is a revert. It doesn't matter if it was returned to the previous state by insertion, or by deletion. It doesn't matter if it was returned to the previous state by hand, or with the assistance of a tool. It doesn't matter whether you knew it was returned to a previous state or not. You may not KNOW it was a revert, but it nevertheless was one. If you suspect you might be reverting something (certainly if you are editing the same thing!! what a giveaway!) more than once in a day... check on it, that's what history and diff are for. Ignorance is no excuse. As I understand it, without digging into your situation deeply, since I'm more interested in the question than the situation, you got a block for reverting and now you want to see if you can argue that it's not actually reversion? Is that about the size of it? There is a limited exception for reverting simple vandalism. I do not think there should be any other exceptions at all. Including the case you raise. You should, after the first time you insert it and it doesn't stick, take it to the talk page. Simple principle, really.  + +Lar: t/c 01:39, 21 June 2006 (UTC)
 * The policy doesn't define reverting as returning an article to a previous state, it defines it as "undoing the actions of another editor...." Those are not at all equivalent.  Indeed this was something that happened to me, but I have another reason for asking about this. I am working on a policy proposal which might need to depend on people being able to freely insert certain kinds of templates including dispute tags which might cause the article to get back to an earlier state.  I'm not ready to share the details.  I will say, however, that I came about this block while following an earlier draft of my proposal as if it was policy. I have since revised it, but there is still a problem with this issue.
 * There are actually four different exceptions in article space at present (simple vandalism, self-reverts, removing posts by banned or blocked users, and potentially libellous material) and a couple more for special circumstances. I am not proposing a new exception, only that people interpret the policy the way that its text defines it, in terms of the specific actions of editors, instead of the way that you and many others define it in terms of returning the article to a previous state, or, in the alternative, change the policy to refer to the state of the article instead of the actions of editors.  I continue to believe that allowing someone on the 3RR borderline to insert a dispute tag would blow off a lot of steam. Publicola 02:53, 21 June 2006 (UTC)


 * Tags are text, and so count like any other text. People should be more reluctant to remove a tag than any other text, but thats a different matter. If you think you might be close to 3RR, then stop: don't get close is the best protection: WP:1RR is good too William M. Connolley 20:26, 20 June 2006 (UTC)
 * Simple vandalism is text, too, so I don't really see the analogy. I think if we could tell people that they can diffuse their frustrations from multiple reverts by adding a dispute tag instead of reverting, that would calm a lot of situations which otherwise end up on AN/3RR -- and thus lessen you and your fellow admins workload there. Publicola 20:33, 20 June 2006 (UTC)
 * Its not an analogy. Its a literal truth. SV has a specific exclusion, so doesn't count. Allowing tag inserts... nice idea, but no. What would happen is that people who couldn't get their pet version on would insist on the tag, regardless of consensus otherwise William M. Connolley 20:40, 20 June 2006 (UTC)
 * I bet you have plenty of experience with that kind of thing happening on the heavily-edited articles in your subject area, but my question anticipates the situation you describe -- if the tag was only removed by reverts to previous versions (i.e., something other than the tag also got changed and the article ended up in the previous state) then it would be okay for someone on the 3RR borderline to add it. But if the next person takes out the tag as a specific action, then adding it back would be a violation. Publicola 00:25, 21 June 2006 (UTC)


 * I'm increasingly disenchanted with tags. I often see them placed in a kind of tag-and-run by someone who hasn't edited the article or commented on the talk page, but who (presumably) just read the article and is unhappy about something. Otherwise, I see them used to hold an article hostage by people who have been unable to gather consensus. I rarely see them used in a productive way. I would not be inclined to treat tags differently from other text. Tom Harrison Talk 20:43, 20 June 2006 (UTC)
 * I too think they are abused that way, and in fact I had been involved with a dispute which took months, with someone who thinks a tag belongs on an article about a nonprofit website because it has too many external sources that he thinks are spam. I and almost all of the other editors disagree, but I'd rather have the tag on the top than see him start editing out sources. Publicola 00:23, 21 June 2006 (UTC)

If the consensus turns out to be that these blocks are valid, then the definition of "undoing the actions of another editor or editors" should be changed to "returning the article to a previous state". Otherwise, I believe, the page technically contradicts itself in any situation where a portion changed by a revert is subsequently changed to an earlier version of that portion. In such cases, there was no "action" undone. Publicola 00:23, 21 June 2006 (UTC)
 * I am not sure I follow this. Removal is a kind of action. If I remove paint from the side of my house, I performed an action (I know it is because I have the sore muscles to prove it, work was expended). If you put it back on my house, you undid my action. I think trying to get policies to cover every possible contingency, exception, special case, twisted definition, etc. is counter productive and instruction creep-ism. Better to assume good faith, to read policies and guidelines for intent and spirit, and to evaluate based on the general idea embodied instead of seeking exact precision and skirting the envelope. This is a wiki, but I don't see these words as needing vast changes, they are fine enough as is, in my view anyway.  + +Lar: t/c 22:52, 21 June 2006 (UTC)

Policy needs Administrator Enforcement Rules
I've been interested in 3RR for a while. And i've been watching how some admins actually enforce the rule, and i've noticed something, some admins simply look at the 3rr listing and block the person without ever looking at the actual situation.

A great example is The Turkic People's History Page anyone that takes 2 seconds to actually look at the history of the page can clearly see that 2 editors were envolved in an edit war, both of them have clearly violated 3rr. However lets look at the outcome of the "war", JWB posted a 3rr notice on Eiorgiomugini and instead of both being banned for 3rr violations as they should have been, the blocking admin did not do his job and simply saw the 3rr notice and blocked Eiorgiomugini, not JWB.

I'm not trying to name any names, but this is happening way way way too frequently, by some admins who seem to use the 3rr page as a way to get their block count up. This is extremely damaging to wikipedia. The 3rr is NOT suppossed to be a punishment, but simply a way to end edit wars. If the admin is neglecting to look at the actual page and get a feeling for the situation the admin is neglecting his duties as an admin and has no business touching the 3rr page. I've seen cases where an edit war was over a comprimize was reached, and then 20 hrs later after the war on the page was ended, an admin finally got around to looking at the 3rr page and decided to toss out a block, even after the warring was over and concensus was reached, which goes against the idea that 3rr is not a punishment.

Another issue i've noticed that is even more disturbing (and has happened to me) is that the blocking admins not only don't look at the page to get a feel for the situation, they don't even make sure all of the listed "reverts" are actually reverts. Here's an example, Editor 1 makes a change to the article, editor 2 reverts it, editor 1 puts it back, editor 2 reverts it, editor 1 puts it back, editor 2 reverts, (for those keeping track, editor 1 now has 2 reverts, and editor 2 has 3) editor 1 then edits a completly seperate section of the article, changing some text that hasn't been changed for weeks. Editor 2 then puts editor 1 up on the 3rr report page, listing the 4 edits I described above, and the blocking admin ignorantly clicks through the edits, sees 4 changes made within 24 hrs and issues a 24hr block to editor 2. This has happened, and seems to be more common them people think. I'm trying to not name any names, however if you don't believe me that something like this has happened i'll be glad to provide some concrete examples.

The 3rr policy must be changed to include rules for the admins issuing the 3rr blocks. I propose that admins be required to get a 2nd admin's approval before any 3rr block is done (unless it's vandalism), and that they are required to read through and understand the situation before issuing any blocks. There also needs to be a system put in place where admins are held accountable for neglecting their duties and being trigger happy on the 3rr page, unblocking someone 20hrs after you blocked them for 24hrs because they pointed out that they didn't violate 3rr is not an acceptable situation. It's getting out of hand and needs to be brought under control. Seraphim 05:56, 21 June 2006 (UTC)


 * True. It's supposed to be an impersonal "electric fence," but essentially all adminstrators will be prejudiced against someone with a 3RR block on their talk page, while those they were reverting against, even if they are clearly more guilty of not discussing their changes, or even more egregious behavior which led to the 3RR situation, go on freely without any real deterrent to their bad behavior.
 * It seems to me that making 3RR easy for administrators to enforce, by limiting common sense exceptions and literal interpretation of the text of the rule in favor of easy-to-remember generalities which don't really reflect the letter or the spirit of the rule, pretty clearly results in an easier time for admins, but much more discontent and ill-will against the more numerous editors who bear the brunt of this corner-cutting. This situation is not good for the project, even if the admins to prefer the ease which it affords them. Publicola 23:20, 22 June 2006 (UTC)

Loophole
"User A" and "User B" are collaborating on an article. Both A and B make several technical adjustments to each other's work. They aren't concerned with the 3RR, since all of the adjustments are consensual.

"User C" has a grievance with A and/or B, and tries to use the edits to the collaborative page to have one or both banned.

Sounds stupid, doesn't it? The problem is that under the current wording, "User C" would be perfectly within his rights to register a complaint.

I don't know if anyone has ever been blocked under these terms, but it might be a good idea to close this loophole all the same. CJCurrie 06:12, 23 June 2006 (UTC)


 * I was just thinking along the same lines. While I agree with the intent of the, "Note: There is no requirement for the reverts to be related: any four reverts on the same page count" wording I think it could be rephrased slightly. If 'any four removals of one or more words' count as a 3RR violation then copy-editing the same page four times in twenty-four hours (if working by sections for instance) would be a 3RR violation... unless no one else edited the page between those efforts. Obviously this is not intended, but it is technically what the page says. Perhaps the 'multiple reverts are treated as one if performed with no intervening edits by someone else' exception should be expanded to 'multiple reverts are treated as one if they have not been objected to (or reverted) by other editors'. The concept that it is 'one revert' to make consecutive changes to five different sections of a page, but becomes 'two reverts' if a bot adds an interwiki link to a foreign language version (or other completely unrelated edit) in the middle of them has always seemed arbitrary anyway. --CBD 10:38, 23 June 2006 (UTC)

This is absolutely true, and it has happened to me. I was working collaboratively with another editor, back-and-forth on the new reference citation styles. Eventually, another editor who didn't like my recent deletions reverts them back in, I revert back to the new citation styles followed by an edit of a summary of the insertions the third editor wanted, thinking I'd have the support of all of the other editors. Half an hour later, bam! Blocked for 3RR! The four "reverts" in question were three cooperative changes that just happened to reverse "other editors work" (from months ago!) and the revert of the third-party, which by that time had already been enhanced with a summary of the text he wanted included. I put up an, but it wasn't taken seriously, and nobody even got to it until 3 hours before the block expired. It sucked, and I thought about griping to the admin who blocked me, but I'm glad I just kept quiet because maybe I would have pissed off the admin even more and had something else happen, because too many people complain about their fair 3RR blocks and that makes the admins cranky. Please fix this abuse loophole thank you! --Supplicant 04:33, 25 June 2006 (UTC)


 * Good lord ... I wouldn't have guessed that this actually happened to someone.


 * This is just another reason why a more lenient 3RR policy might be in order. Please see my latest proposal, listed below.  CJCurrie 04:39, 25 June 2006 (UTC)

Technically, under the "undoing the actions of another editor or other editors in whole or part" definition, which I have disputed, any deletion of any text that you did not personally insert into the article is a revert, by definition. If you delete four passages from any one article in seperate edits with intervening edits by other editors, within 24 hours, then unless you wrote those passages yourself, then you have violated the letter (but not the spirit) of the 3RR. Any admin who tells you that they do not judge edit quality when making 3RR block decisions needs to think about what they are saying more carefully. Again, I believe the correct definition of a revert should be "returning the article to a previous state." Accordinly, I am adding a contradict tag to this policy. Publicola 04:47, 25 June 2006 (UTC)


 * However, your own personal definition of a revert does not accord with Wikipedia practice or policy. Jayjg <small style="color:darkgreen;">(talk) 23:23, 25 June 2006 (UTC)


 * Wait a second -- how would four such deletions not qualify under the policy's definition? If they were text written by others, and the deletions had intervening edits from others, then that would be a 3RR violation.  Are there some unwritten rules which the admins use when judging what is or is not a 3RR violation?  Are there any such rules that would be fair?  Don't such rules require a value judgement on the part of the reviewing admins, contrary to what is claimed above? 71.132.131.237 01:07, 27 June 2006 (UTC)

Suggested change: Only block as a last resort
Several discussions have taken place about the future of the 3RR since my first "3RR confusion" post a few weeks ago. If there is any conclusion to be drawn from these discussions, it's that there is no concensus as to the future of the 3RR.

While most contributors agree that the 3RR is a useful policy for cracking down on vandalism and tendentious editing, there is no similar consensus on the specific rules that govern the policy. Many believe that it has become applied too rigidly, and often punishes well-intentioned editors on technical grounds. Others believe that the current regulations are appropriate, and that any significant changes would be detrimental to Wikipedia's integrity. I suspect that it would be nearly impossible to reach consensus on these matters.

For that reason, I'm going to try a different approach. My proposal is to leave the regulations as they are, but to change the punishment meted out to offenders. This is not necessarily a radical departure from the current approach. The 3RR does not require that offending users be blocked; it simply grants administrators the power to carry out such blocks. My recommendation is that such blocks be used very sparingly.

Specifically, I propose:

(i) That administrators be required to examine the nature of specific disputes. If the 3RR violation is simply a spelling/grammar fix or something similar, then the offender should not be blocked. Instead, s/he should be encouraged to find someone else to fix the page should the problem recur in the future. (In extreme situations, the offender should be permitted to claim justification for his/her changes.)

(ii) That offenders be given a chance to correct themselves prior to a block occurring. That is, editors who violate the 3RR should be given the opportunity to self-revert their changes in order to avoid being blocked.

(iii) That blocks be reserved for chronic offenders, or for contributors who refuse the opportunity to self-revert.

What say others? CJCurrie 03:18, 25 June 2006 (UTC)


 * CJ, the whole point of the policy is to stop admins from judging quality of edits, and to stop editors from thinking that, just because they prefer their edits, they have the right to revert war endlessly to retain them. Your changes would overturn the whole ethos of that, and there is no consensus to do that, as the recent thread on the mailing list made clear. People can be blocked at the moment even before they get to 3RR, if an admin is so inclined. SlimVirgin <sup style="color:purple;">(talk) 03:25, 25 June 2006 (UTC)


 * Yes, and I disagree with that rule as well. Returning to the present dispute, I'm not suggesting that admins involve themselves in content wars, but they should be allowed to determine if the change is essentially content-neutral (eg. a spelling or grammar correction, or something similar).  CJCurrie 03:34, 25 June 2006 (UTC)


 * Without judging the quality, the loophole above exists, and will continue to be exploited. Supplicant 04:39, 25 June 2006 (UTC)


 * Agreed. Complementary or consensual edits should not count toward the 3RR.  CJCurrie 04:41, 25 June 2006 (UTC)


 * Admins aren't mind readers, how are they going to know what two other editors do and do not agree on? In many cases they both might be unsure. Publicola 04:55, 25 June 2006 (UTC)


 * Admins may not be mind readers, but surely they can ask if two editors were working co-operatively on a page. CJCurrie 05:34, 25 June 2006 (UTC)


 * So, if more than two people are involved in the recent edit history, then the admin (or the 3RR complainer) would have to get another recent editor, to "second" their motion for a block. That seems fair -- get at least two people to have to sign on to a 3RR request, unless there were only two people edit warring.  That would have saved me from my 3RR block, I suspect.  Publicola 07:41, 25 June 2006 (UTC)

I disagree with pretty well everything CJC says. 'If there is any conclusion to be drawn from these discussions, it's that there is no concensus as to the future of the 3RR.'' is quite wrong: the conclusion is, that blocking admins are quite happy with the current rules and see no need for these new proposals William M. Connolley 09:03, 25 June 2006 (UTC)

I also disagree with changing this policy. WP:3RR sends the right message: revert more than three times in 24hrs and you will be blocked. Period. Easy to understand and easy to follow. Don't want to be penalized, don't do it. ≈ jossi ≈ t &bull; @ 22:58, 25 June 2006 (UTC)


 * Exactly. These constant attempts to introduce loopholes only increase the likelihood of edit-warring; 3RR is intended to decrease that likelihood. Jayjg <small style="color:darkgreen;">(talk) 23:19, 25 June 2006 (UTC)


 * I agree. There's a strong consensus in favor of measures that reduce edit warring, and that includes this policy, so it shouldn't be weakened. SlimVirgin <sup style="color:purple;">(talk) 04:58, 26 June 2006 (UTC)


 * I'm not proposing a change in policy, so much as a change in the punishment meted out to offenders. CJCurrie 05:09, 26 June 2006 (UTC)


 * No, you're proposing a fundamental change to the policy viz. that whether it's a blockable violation should depend on the quality of the edits. You wrote: "I propose ... [t]hat administrators be required to examine the nature of specific disputes." But the whole point of the policy is to limit edit warring, period, and also to create a strong culture against editing-by-revert, regardless of the quality of the material being reverted. If you're fixing someone's spelling mistakes and they keep reverting you, the thing to do is not to keep on reverting them back, because then there's no end to it. When it reaches the point where both have reverted three times, you should seek help from another admin, who can look to see whether the reinsertion of spelling errors has reached the level of disruption or vandalism, and if it has, other steps can be taken. But if it hasn't, if it remains at the level of a content dispute, then both editors have to be treated equally. I realize that it makes no sense sometimes, but most of the time it makes a lot of sense, because editors unfortunately disagree as to what constitutes a high-quality edit. There is a strong consensus in favor of the policy as it now stands. SlimVirgin <sup style="color:purple;">(talk) 08:06, 26 June 2006 (UTC)

Another proposed enhancement
How about if we require that the person complaining of the 3RR violation be required to point it out on the offender's talk page, and admins be required to give them, say, 30 minutes to self-revert? Any problems with that? Publicola 07:41, 25 June 2006 (UTC)


 * How about you stop this pointless attempt at tweaking of the rules driven purely, I assume, by your own recent blocks. Admins will in general warn-not-block newbies who genuinely don't know the rules; the page already advises people to issue warnings William M. Connolley 09:15, 25 June 2006 (UTC)


 * I've only been blocked once. I have pointed out that the definition of 3RR is flawed and arbitrary, reports are often capricious, and enforcement requires making judgement calls, contrary to your assertion.  Unable to address the technical substance of my complaints, you then "assume" that I am acting "purely" in bad faith.  I note without comment that you spend much of your time on 3RR enforcement. Might I ask that you assume good faith and help to explore the shortcomings of the rule instead of threatining people with punitive measures for complaining about its shortcomings? Publicola 17:12, 27 June 2006 (UTC)


 * No, the defn is sensible. *Reports* may be capricious... I suppose we could add a rule saying "no capricious reports" but I doubt that would help. Enforcement indeed requires judgement, which is why admins get to do it. I have had a lot of experience with 3RR... and I find that the rule itself works quite well. It maybe could do with tightening to 2RR, perhaps, but in general 3 seems to work OK. People complaining about the rules after being blocked is alas not unknown, but they give up and go off to more useful things after a while... William M. Connolley 17:39, 27 June 2006 (UTC)


 * William, I think Publicola is suggesting that these warnings be mandatory, not optional. This suggests that he does know what the rules currently say but happens to disagree with that rule, as is his right.  I realize you may not agree with him, but it's important that you do your best to understand his point and not to discourage him from participation.  The ideas of a new user are not automatically wrong, nor are those of a seasoned admin automatically right.  What matters is the merit of the ideas, not the person proposing them.  Al  17:49, 27 June 2006 (UTC)

Agree with William. Do not want to be penalized? Don't revert more than three times in 24 hrs. Easy. ≈ jossi ≈ t &bull; @ 23:00, 25 June 2006 (UTC)


 * "Blocking is always preventative, not punitive.", if the edit war is over, there is nothing to prevent. The idea that the 3rr is a punishment goes against the actual policy.  If there is an active edit war, then I have no problem with someone going in to block one of the 2 warring editors.  However if an editor makes say 2 reverts on one section, and then earlier in the day there were 2 edits elsewhere in the article, and there is no war going on at all, the admins should not be allowed to block the user, same goes for admins reading 2+hr old 3rr reports and blocking the user.  If the war already stopped you should NOT block the user.  It's PREVENTATIVE and not a punishment, and some admins seem to be forgetting that and instead trying to rack up their ban count.  Seraphim  23:13, 25 June 2006 (UTC)


 * Actually, in practice its use is prohibitive; specifically, attempting to prohibit further edit-warring. Editors who consistently get blocked for 3RR either learn to stop edit-warring, or eventually have to leave the project.  If editors did not get blocked, they would not learn that edit-warring is bad, and would edit-war with impunity. Jayjg <small style="color:darkgreen;">(talk) 23:21, 25 June 2006 (UTC)


 * Surely you must admit, that blocking a user that did 4 reverts, 3hrs after the edit war is over is not preventative, especially since the edit war ended without an admin stepping in, which shows that all the editors involved managed to come to a comprimise, which is exactly how wikipedia works. Attempting to claim that it's preventative because they will learn that edit warring is bad is simply not how the policy is written, if it was it wouldn't say "Blocking is always preventative, not punitive" which makes it clear that blocking a user under the 3rr is not intended to be a punishment. Which means that it was not intended to be preventative through repeat punishments.
 * The facts are that some admins are blocking people sometimes up to 10 hrs after the edit war has ended, an edit war that may simply have been a user getting caught making 2 reverts, of a seperate section of an article, of complex vandalism like linkspam, 20hrs after he reverted the link 2 other times the previous day. Admins need to actually take the time to look at the article, and if needed block all the people who violated 3rr on the article, not just the people that were reported, and if the edit war is over, there is nothing for them to prevent, and they should move on.  What is going on now is clear abuse, and the casualty is usually the new editors who don't know what the 3rr is. Seraphim  23:29, 25 June 2006 (UTC)


 * Not at all. People revert 4 times, and hope to get away with it. If they discover that it inevitably results in a block, and repetition merits ever longer blocks, then they will stop doing so. If they find that they are never blocked because no-one gets to it "in time", or they assert that "the edit-war is over", then they will continue edit-warring. People are almost always blocked as a result of announcements on WP:AN/3RR; the notices wouldn't go up there if people really thought the edit war was over. Jayjg <small style="color:darkgreen;">(talk) 23:33, 25 June 2006 (UTC)

Unless they are being vindictive.Homey 04:51, 26 June 2006 (UTC)


 * I have to say that SeraphimXI is right. No matter how it's being spun, the current policy is cleary punitive and not preventative.


 * To be preventative, blocking has to occur during an edit war, not for reverts in the dim past. Likewise, if the goal is to prevent, then a warning suffices.  All you have to do is say "You have reached the limit of 3RR; one more revert and you will be blocked".


 * Besides being punitive instead of preventative, the current policy also isn't effective. The threat of force is better than force itself for preventing an edit war.  As a nice benefit, this would avoid the possibility that the user simply lost track of edits on a busy, perhaps controbersial article or that there are other circumstances.  For example, due to a misunderstanding by an admin, Infinity0 was recently blocked on 3RR violation and then unblocked.


 * For these reasons, I support a policy of warning before blocking, each and every time, and not blocking for historical events where no such warning is possible. Al  17:38, 27 June 2006 (UTC)

exceptions - editsummary
Would a note in the editsummary be more effective than a talkpage entry? Agathoclea 20:32, 25 June 2006 (UTC)


 * People tend to claim the exception - rv banned user for example - but this then usually requires something on the talk page as evidence for whatever ban, block or whatever is claimed, since there generally isn't room in the summary William M. Connolley 20:38, 25 June 2006 (UTC)


 * Yes it can be rather tricky - I have been fideling with that and finally thought I cracked it - [ block log] - but I failed. Agathoclea 21:14, 25 June 2006 (UTC)

Cleanup
The policy page was becoming a bit messy with exceptions listed in three places - and the three lists not agreeing. I have consolidated these into the 'Exceptions' section other than a short list in the 'Detail' section which includes the most common types and instructions to see 'Exceptions' for more info. Hopefully that will be uncontroversial, but in the process I also made a few changes which not everyone may agree with; Please comment / question / revise as needed. --CBD 10:47, 26 June 2006 (UTC)
 * 1) I said that 3RR does not apply if there is no edit warring. I'm not sure why this would be controversial, but there hasn't seemed to be agreement on it.
 * 2) I changed a statement about "spam and vandalism" to just "vandalism" (with a link) since spam is listed as a type of vandalism.
 * 3) I removed a statement that 'removing warnings from your talk page is not allowed' because that is a proposed policy change which has been disputed.
 * 4) I moved the 'Administrator involvement' sub-section from 'Exceptions' to 'Enforcement' because it isn't an exception to when something counts as 3RR like the others, but rather an exception to who can enforce 3RR.
 * 5) I variously moved and reworded some things while attempting, possibly unsuccessfully, to retain the same meaning/intent.


 * CBD, you seem to have added new material, or limited some of the exceptions, in your copy edit, so I'm removing them and trying to restore it to the way it was. SlimVirgin <sup style="color:purple;">(talk) 20:44, 26 June 2006 (UTC)


 * Fine by me. I'm not a big fan of the 'reverting blocked users may be harassment if they are not being disruptive', but included it based on two prior ArbCom rulings to that effect which I have seen. The other two sections you edited were actually present before my update. --CBD 23:02, 26 June 2006 (UTC)


 * Hi, yes, I'm sorry, I just noticed they were there already. Someone added, for example, that users can't delete libel (beyond 3RR) if they're involved in editing the page, which is of course wrong. They can and must. Also, reverting blocked users is never harassment. And a few other things like that have been added, but I don't have time to look through it all carefully right now. Will do so later. Sorry for blaming the changes on you. SlimVirgin <sup style="color:purple;">(talk) 23:08, 26 June 2006 (UTC)


 * No worries. --CBD 23:15, 26 June 2006 (UTC)

3RR as vendetta
I am getting quite concerned about Zeq using 3RR as a way of persuing a vendetta. Compaints by me resulted in his being banned from a few articles and being blocked for vote stacking. Since then he has become obsessed by me and wikistalks me incessently. He opened an article on wikipedia under my real name, even. Now, he is poring over edits in an attempt to construct 3RR cases. His current one cherrypicks edits from Israeli apartheid from more than a 24 hour period. Also, since browser limitations make it difficult for me to edit more than one section at a time I tend to edit in small batches (which I agree I should try not to do). He has taken three edits made in a a row, with no intervening edits from anyone else, and is counting each of them as three (or is it four) different reverts. He is showing extreme bad faith and vindictiveness and is posting complaints against the spirt of 3RR which his friends amongst admins have obliged on at least one occasion. Surely this sort of stalking is not what the 3RR was intended to permit.Homey 12:15, 26 June 2006 (UTC)

Did he or  did not he  violated 3RR multiple times ? The facts are what counts - not Homey's "explnations" above (which are not accurate) see 3RR board. Zeq 12:38, 26 June 2006 (UTC)
 * Please ignore Homey accusations and just check the facts:

When one of your claimed reverts is restoring a quotation that was removed weeks (not hours, weeks) ago it's clear you are trying to manufacture allegations. Same with taking a cluster of uninterrupted edits as "multiple reverts". Homey 13:18, 26 June 2006 (UTC)


 * Folks, this talk page is meant to discuss proposed changes to the policy, not to resolve personal disputes. Pecher Talk 13:41, 26 June 2006 (UTC)

The policy problem is editors who tourque 3RR for purposes of a vendetta. I'm afraid it does happen, Zeq is a prime example and it is a problem that will continue unless something is done to stop editors from abusing the complaint system to make bad faith and specious complaints. Homey 13:55, 26 June 2006 (UTC)


 * So, what's your suggestion regarding the policy? Pecher Talk 13:57, 26 June 2006 (UTC)

A 48 hour ban for editors who make specious or bad faith 3RR complaints. An acknowledgement that a series of uninterrupted edits in a row should not count as seperate reverts since the purpose of the 3RR is to stop or prevent revert wars ie where "first editor makes an edit, a second editor reverts it, the first editor reverts the second editor, the second editor reverts the first" etc as opposed to "first editor editing, first editor editing, first editor editing, no editor reverts in response" which is what Zeq is somehow claiming to be 3 reverts. For purposes of 3RR having one editor make several edits in a row, uninterrupted should really be counted as a single edit.

Taking two edits that are, literally, 23 hours and 59 minutes apart and saying this is a revert is not what a good faith editor would do. Cherry picking from a series of a dozen or so uninturrupted edits and taking 3 of them as 3 different reverts is not what a good faith editor would complain about. Taking edits that are weeks apart, as Zeq has done, and saying they are a revert is definitely bad faith. Homey 14:07, 26 June 2006 (UTC)


 * No-one is going to change the rules for this. People have been blocked for multiple vexatious complaints (:-) but you're not there yet. How about sticking to WP:1RR and all your worries will disappear... William M. Connolley 19:32, 26 June 2006 (UTC)

Grammar and copy editing
Following up from the above, Zeq filed a sucessful 3RR against me following his posting completely ungrammatical paragraphs to an article I had started Boycott of Israel. I attempted to correct his grammar in some places and in others hid his paragraph with a note to him to please rewrite. As a result he slapped me with a 3RR complaint which one of his friends banned me for. If copy editing and removing or hiding incomprehensible writing is now considered edit warring and a 3RR offence we might as well give up the ghost. Homey 18:57, 26 June 2006 (UTC)


 * Read the rules William M. Connolley 19:32, 26 June 2006 (UTC)

I've just reread them. I don't see anything in them that refers to dealing with ungrammatical or incomprehensible edits. Homey 19:57, 26 June 2006 (UTC)


 * Exactly. SlimVirgin <sup style="color:purple;">(talk) 20:03, 26 June 2006 (UTC)


 * HotR, I think the point is that edit warring over grammar is still edit warring. The 3RR policy is non-subjective by design. That doesn't mean we have to 'give up the ghost' though. There are other ways of resolving issues than just edit warring. Post a note on the talk page and discuss until you agree... or if that doesn't work then take it to dispute resolution. --CBD 20:29, 26 June 2006 (UTC)


 * If an editor is reasonable that would be fine but if an editor is looking for excuses to post fatuous 3RR complaints this provides a rather large loophole. Simple copy editing, such as correcting spelling or grammar should not count as a revert at least not initially. In my case Zeq posted one sentence, I corrected it, he then posted a second sentence which I corrected and then two more which I corrected. He then put forward a successful 3RR complaint when all that was happening was that I was either correcting or commenting out paragraphs that had fixable errors in the former cases or were incomprehensible in the latter.


 * I agree, if you end up getting two editors reverting each other over spelling or grammar then it's a revert war and should count to 3RR but if you have an editor who is simply going through someone's submissions and copy editing them in a single instance those instances should not count as "reverts" towards a 3RR penalty. Homey 21:04, 26 June 2006 (UTC)


 * Yes, it does sound like he may have gamed you into violating the 'stricter interpretation' of the policy... you thought you were ok because you were editing different sections each time and he didn't contest any of your changes. That sort of '3RR without any edit warring' situation doesn't seem valid to me. If nobody is reversing your changes or saying they disagree how can it be called an 'edit war'? That's why I added the 'if no edit warring has taken place' exception. --CBD 22:16, 28 June 2006 (UTC)

Also, as I recall, I did actually post a note on User talk:Zeq asking him to please proofread and use spell or grammarcheck. My request was ignored. Homey 21:15, 26 June 2006 (UTC)

Suggestion: Mandatory warnings before blocks.
Police are trained to, whenever possible, first warn an alleged criminal that failure to comply with their commands will result in the use of force. Only when the person ignores this warning is force then used. This allows someone to choose to cooperate instead of resisting, and results in a lot fewer civilians getting shot. There is a lesson in this that we can apply here.

Sticking with the police analogy but shifting to a smaller scale, firearms have both a trigger and a safety, where the latter must be intentionally defeated for the former to function. In our context, warnings are the safety, blocks are the trigger.

I propose that, in order for an editor to be blocked for 3RR violation, they must first be warned on their talk page.

Here is the specific policy I suggest:
 * No editor may be blocked for violation of WP:3RR without prior warning.
 * If an editor is at or near the 3RR limit, a standard warning tag may be placed on their talk page by any editor. Once this warning is in place, if that editor chooses to revert past the limit then they may be blocked.
 * If the editor's reverts are not noticed until they actually pass the 3RR limit, then their last edit may be reverted as if it were vandalism and a warning placed on their talk page. As before, once the warning is in place, choosing to revert further may result in a block.

The net effect is to reduce the number of blocks while still preventing edit wars by using the threat of punishment instead of actual punishment. Editors will only be blocked for willfully and intentionally violating 3RR, not for any sort of accidental violation or misunderstanding (such as the recent case of Infinity0 being mistakenly blocked when he was reverting text by a banned user or certain other cases where the rules were ambiguous and the editor did not intend to edit war). Nobody will ever be surprised by a 3RR block again.

Note that, while this policy is no harder to enforce, it is no less effective. In fact, since any editor can place the warning, admins are only needed in the few remaining cases where the offending editor ignores the warning. This will reduce the workload of admins and shrink the AN/3RR page.

I recommend this policy on the basis of its fairness, efficiency and adherance to the spirit of WP:BLOCK. Of course, I would not think of just changing the text of the project page without a consensus of support, so please respond with your feedback. Al 20:13, 27 June 2006 (UTC)


 * I am in agreement with this idea. However, I think that only one warning per user is sufficient.  What I mean by that is that a person may be new on Wikipedia and make an error, never knowing about the 3rr Rule and get blocked.  A warning would be right.  But only one warning.  Get in another war and you get blocked.  I have never been blocked so I am not saying this as someone with a problem, but I am in agreement with this proposal.  I think it just makes good sense.--Anon 64 02:50, 28 June 2006 (UTC)


 * As it stands, there is already some mention that newbies should get a warning, not a block. But if it works for newbies, why not apply it across the board?  The net result is fewer blocks while still stopping edit wars.  Better yet, warnings prevent oversights, like the Infinity0 incident and avoid unnecessary admin involvement. I'd like you to take another look at the full proposal and reconsider it in light of this.  Al  04:04, 28 June 2006 (UTC)


 * Well, I think the main reason to limit it to ONE TIME is that otherwise, people may take some license with it. "I will do it and maybe I won't get caught and if I do get caught, I will just get a warning and a revert, Oh well.".  --72.13.168.149 05:48, 28 June 2006 (UTC)

Uhm, our goal isn't to catch and block people. At best, that is a means towards an end, and it's not the best means. Our goal is to stop people from edit-warring. Under the proposed rules, if someone starts to edit war, they'll get a warning on their talk page that opens them up to a block, which makes it rather unsafe to ignore. Likewise, having their changes reverted directly undermines their motivation for edit-warring, since it means they didn't get their desired version. The proposed rules would lower edit-warring while also lowering the need for admins to block. It's a win-win. Al 06:00, 28 June 2006 (UTC)


 * This proposal sounds very well thought out, and the terms make a lot of sense. In fact, I am surprised that such a policy is not alreay in place. I support this move. rom a rin [talk ] 21:32, 27 June 2006 (UTC)


 * I see no pressing need for this. I don't block unless I am confident the user is aware thsat what they are doing is wrong - if they have had multiple previous blocks for 3RR I see no need to enforce a "no warning so you can't block me" clause; also, it is an added incentive to blanking warnings from user talk (which does not always get spotted).  Better to educate the people reporting than add more instructions.  Just zis <span style="border: 1px; border-style:solid; padding:0px 2px 2px 2px; color:white; background-color:darkblue; font-weight:bold">Guy you know? 22:04, 27 June 2006 (UTC)


 * Well, for one thing, even editors who've violated 3RR in the past are capable of unintentionally (or controversially) violating it today. With a warn-first policy, admins will be able to determine if it was intentional and avoid punishing otherwise.  Better yet, a warning by any editor will carry weight, as it would enable blocking.
 * As for hiding the warning, I don't see how that's going to fool anyone for an instant. Moreover, any such attempt merits a block, which should be sufficient disincentive.  Consider that if we see someone violating 3RR and therefore leave a warning, we're going to notice if a warning was previously left, particularly if the edit comment was something obvious, like "3rr warning for ArticleName".
 * I should also mention that adding a warning is a lot easier than filling out a 3RR violation report, with all those diffs and history snapshots, so editors will be further encouraged to be on the lookout, without adding to admin workload.
 * With all due respect, I don't see anything you've brought up that argues strongly against this suggested policy. Al  00:37, 28 June 2006 (UTC)


 * It's not as if the "three revert rule" block was some surprising and unorthodox idea. If you are ever blocked for the "three revert rule" it follows that you must have been edit warring, and so shouldn't be surprised that something is done to stop it.  Complaining that you were not warned is somewhat missing the point.  --Tony Sidaway 22:19, 27 June 2006 (UTC)


 * As it turns out, that doesn't quite follow. Consider the recent case of Infinity0, who was blocked for 3RR violation despite not having been edit warring.  If he had been warned first, he could have left a brief response explaining that he was only reverting text by a blocked user, saving him the trouble of being incorrectly blocked then having the block removed.  This is not a unique case, and there are other examples on this talk page of how 3RR could be violated without edit-warring.
 * The purpose of warning is to avoid the need for blocking. It is a minimal force approach that works better than full force. I don't quite see how your comment addresses this virtue.  Al  00:37, 28 June 2006 (UTC)


 * I think it's useful to warn people, but I oppose this as instruction creep. The more process we put on 3RR the more people will see three reverts as an entitlement. Tom Harrison Talk 23:04, 27 June 2006 (UTC)


 * I would say that this is the opposite of instruction creep, as it simplifies the procedure. It is less work to drop a warning tag than to fill out a full 3RR ANI notice, much less block someone and leave an explanatory comment.
 * Keep in mind that, at the moment, editors can do 3 reverts without getting any feedback at all. Under my suggested policy, edit-warriors will quickly accumulate warnings on their user page, which serves as corrective feedback. I would consider this a further advantage.  Al  00:37, 28 June 2006 (UTC)


 * Before being reported on WP:AIAV for vandalism the editor is supposed to have had a test3 or 4 warning. Most 3rr blocks come after reports to the 3rr noticeboard so we could require that a warning be given before the 3rr report is logged to bring this into line and ensure consistency. This will give editors the chance to self revert in the case of a mistake and will avoid editors getting caught out by not understanding the "edit's don't need to be related" rule. Sophia  23:12, 27 June 2006 (UTC)


 * Indeed. Under my suggested rules, someone who violates 3RR without a warning won't get blocked, but they'll have their last edit reverted as being equivalent to vandalism, which takes away anything they hoped to achieve by edit-warring.  And, of course, if they then revert, the block is automatic. I think this handles all outcomes well.  Al  00:37, 28 June 2006 (UTC)


 * We're not compelled to warn before blocking for vandalism, either. In particularly egregious it's not uncommon to impose a block without warning. As with edit warring, a user should realize that the conduct isn't appropriate, whether they've read the specific policy or not. TenOfAllTrades(talk) 23:21, 27 June 2006 (UTC)


 * I'm not sure if this is at all parallel, as vandalism is defined as intentionally damaging an article, while 3RR violation can be inadvertant or only apparent. Vandalism occurs with the very first edit, while 3RR takes time to build up, making a warning particularly effective.
 * Also, please note that, as per the text above, 3RR violation does not necessarily involve actual edit-warring. I think you'll find that the suggested policy handles 3RR in a way that's appropriate for 3RR, which is quite different from what's appropriate for vandalism. Al  00:37, 28 June 2006 (UTC)


 * Again, I think it is a good policy with a ONE TIME rule. I do not think it counts as instruction creep.--Anon 64 03:28, 28 June 2006 (UTC)


 * It is instruction creep. It adds another requirement, another layer of red tape. I don't see the pressing need for this added rule. - Will Beback 04:22, 28 June 2006 (UTC)


 * I do not see the red tape, but I am not an expert in this. Is it easier to block someone than to put a warning on their page?--72.13.168.149 05:49, 28 June 2006 (UTC)

Under the current rules, a 3RR block usually starts by someone reporting an editor on WP:ANI/3RR. Because you need a historic link and four diffs with matching timestamps, this is a lengthy process. Then an admin has to see it and carefully confirm (we hope!) the validity of the report. At this point, they can review the history of the editor who they believe has violated 3RR to see if they're newbies, regular users with clean records or repeat offenders. Depending on what's found, the admin either leaves a warning or leaves a notification of block and then sets a block. If the blocked editor disputes the block, another admin has to come in and review it. Alternately, the block can skip the WP:ANI/3RR step if the admin notices the reverting first-hand, but the rest of the steps remain, and now the admin has to post elsewhere on ANI to alert other admins of the block.

In contrast, under the proposed rules, most cases would go no further than a regular editor placing a warning tag on a user page. The warning tag makes the target vulnerable to a block and shows that someone is watching them, ready to block. This prevents anyone from accidentally or mistakenly violating 3RR, and is likely to prevent the need for a WP:ANI/3RR report and all the steps that follow. In short, it's a bloodless solution to the problem of edit wars. Al 06:07, 28 June 2006 (UTC)


 * The aim of the 3RR block is to stop an edit war. I think all types of warnings, etc, are good if they stop the edit warring. But more rules don't necessarily help stop the edit wars. Instead they give those with bad faith another opportunity for wikilawyering. -Will Beback 05:55, 28 June 2006 (UTC)

I'm sorry, but I don't understand how you get to the conclusion that it gives the blocked user an opportunity to wikilawyer. Quite to the contrary, under the proposed rule, the blocked user has necessarily been warned before the block, so they know better than to continue reverting. I think it's a safe bet to say that most people who violate 3RR do not do so with the intent of getting caught and blocked. A warning would therefore be highly effective. Al 06:09, 28 June 2006 (UTC)


 * The need is that it removes all those unnecessary blocks, replacing them with something just as effective but less forceful. As a side effect, it makes the process more fair.  Al  05:34, 28 June 2006 (UTC)
 * Which unnecessary blocks? Is there a list of them somewhere? -Will Beback 05:55, 28 June 2006 (UTC)

Yes, actually, there is: it's called WP:ANI/3RR. With warnings that have teeth, blocking becomes rare and largely unnecessary. Remember, our goal is to stop edit warring, not block users. Blocking is no more than a means to an end, and I argue that it is not a particularly good means. Al 06:11, 28 June 2006 (UTC)
 * So does that mean you think that every 3RR block is unnecessary? Do you think that warnings can entirely replace blocks? -Will Beback 06:16, 28 June 2006 (UTC)

That depends how stupid and self-destructive people are. If you've got a warning on your talk page and you know that someone is watching you and will get you blocked if you continue edit-warring, you're going to stop immediately if you have any sense of self-preservation. Think of the warning tag as a little red laser dot on your forehead that shows that the police sniper has you in their sights.

Will some people be too stupid to take the hint the first time and find themselves blocked? Perhaps, but I believe that the next time they see a warning, they'll know what it means and be properly deterred. People violate 3RR either because they don't know better, by accident, or they think they can get away with it. A warning fixes all three.

So are all the 3RR blocks unnecessary? Perhaps we can't avoid every last one, but I believe we can reduce the number of blocks substantially with this policy, lowering the workload of admins while increasing the fairness of the system. If you disagree, I would genuinely like to understand why. Al 06:25, 28 June 2006 (UTC)


 * So again, how much of a problem is this? How many unneccessary blocks are there? You haven't shown that this is a problem. I personally think that folks should always be warned as a good principle, but I don't think it should be a rule. Lastly, I'm surprised how many times I've warned someone about the 3RR using the standard tempate and they've lashed back that they haven't broken the 3RR and how dare I accuse them of it. Perhaps all we need to do is change the 3RR warning. -Will Beback 06:32, 28 June 2006 (UTC)


 * Right now, there is no incentive to warn. In fact, there is a disincentive, as I've seen experienced edit warriors double-team an opponent and lure them into miscounting reverts so as to get them to violate 3RR, allowing a 3RR report to be filed so that their opponent is blocked.
 * Since there is no incentive to warn, and some incentive to avoid warning, most 3RR blocks come without any warning, giving the editor no chance to terminate the edit war. This is counterproductive.
 * By requiring a 3RR warning, we've changed it into something that has bite. It not only enables a block, it informs the editor that blocking is now enabled and that their edit pattern on this article is being monitored.  This is a strong deterrent.
 * In this way, the suggested change not only enshrines a good principle as the rule, it empowers the mundane warning template with teeth. Even if someone disputes the validity of the warning, they've lost any ability to deny that they were warned, which removes a whole lot of wikilawyering.
 * In the end, I'm arguing that, in principle, all the blocks on WP:ANI/3RR are unnecessary, and in practice, most can be avoided. All we have to do is apply some basic game theory and recognize that limiting ourselves (by disallowing unwarned blocks) can actually be more effective.  Al  06:45, 28 June 2006 (UTC)


 * Unless there's evidence that this is a real problem, I don't see any reason to change the policy. The concerns appear based on theory rather than actual incidents. -Will Beback 22:02, 28 June 2006 (UTC)


 * I would think that the continued presence of a non-trivial number of 3RR blocks each day is overwhelming evidence that the policy could be improved. All 3RR blocks qualify as "actual incidents". Al  22:17, 28 June 2006 (UTC)

I think fairness, and being seen to be fair, is the key. Homey 05:50, 28 June 2006 (UTC)


 * Indeed, and warning before blocking is extremely fair. It takes away the surprise factor, offers deterence instead of punishment, and avoids a whole class of errors.  Al  06:11, 28 June 2006 (UTC)

they give those with bad faith another opportunity for wikilawyering
 * I think, in fact, mandatory warnings reduce the opportunities for wikilawyering by removing the "I didn't know about 3RR", "I didn't realise this was 3RR" and "it was an accident" defences. If someone continues to revert after they've been warned it's an open and shut case. Homey 05:58, 28 June 2006 (UTC)

Indeed. Al 06:11, 28 June 2006 (UTC)


 * I agree, this is a very well thought out and helpful idea. Wikibout-Talk to me! 16:57, 28 June 2006 (UTC)

I too am against this idea, for all the obvious reasons people have given above. Newbies generally do get a warning; others usually not. Increasing the workload for the admins, and the barrier to being blocked, is not a good idea William M. Connolley 19:14, 28 June 2006 (UTC)


 * While you're free to disagree, I would prefer that you not do so on the basis of a misunderstanding. Under the proposed rule, there would be reduced workload because most edit wars would be solved without admin intervention.  Yes, it's a barrier to being blocked, and that's what makes it work.  Frankly, you make it sound almost as if you think that blocking is the goal of WP:3RR, when in fact it's just one way to end edit wars, and not the best one.  Al  19:36, 28 June 2006 (UTC)
 * I don't think mandatory warnings would help end edit wars. If leaving warnings would help, people would do it (and indeed, often they do) William M. Connolley 20:13, 28 June 2006 (UTC)


 * That's because the current warnings have no teeth. Give me a minute and I'll write up my idea of a warning tag, ok?  Al  20:21, 28 June 2006 (UTC)


 * I also oppose it. Newbies should be warned once and asked to review the policy. After that, it's often a good idea to offer someone the chance to self-revert before reporting them, but very aggressive editors simply use that as another opportunity to keep on fighting, so it certainly shouldn't be mandatory. SlimVirgin <sup style="color:purple;">(talk) 19:20, 28 June 2006 (UTC)


 * I don't quite see how your conclusion follows from your argument. An editor who gets a warning now knows that any further reverting will get them blocked, which is a strong incentive to stop.  In contrast, blocking people after they've already passed the limit causes more disruption but no additional benefit.   Al  19:36, 28 June 2006 (UTC)

Hell no. The goal is to stop edit wars, not to provide a framework for wikilawyering. --Carnildo 19:30, 28 June 2006 (UTC)


 * Presumably, you have some argument for how the proposed rule would fail to stop edit wars or increase wikilawyering, but you have not stated it thus far. Since you offer no reasoning, I can offer no rebuttal. Then again, the absence of reasoning is rebuttal enough.  Al  19:36, 28 June 2006 (UTC)

If I may make an observation, it seems as though everyone so far who has stated that they are agaist this proposal has not even read it. There is a lot of talk about "making more work for admins" and "wikilawyering", none of which even remotely applies to this idea. I would suggest that anyone wishing to post their comments here actually read what they are commenting on, and think about the implications for a moment. At the risk of being accused of assuming bad faith, it almost looks like certain admins enjoy being able to block users, and will oppose anything that might hinder such blocks. Please, please prove me wrong; I don't wish to have such a negative view of the Wikipedia Administrators. rom a rin [talk ] 20:27, 28 June 2006 (UTC)


 * Romarin, as I'm the one proposing the rule change, it's up to me to be as clear as humanly possible. If all these editors are getting from this is "oh, look, another bit of red tape to get in the way of my job", then you're right: they're not reading this carefully.  Regardless, it's up to me to address any disagreement, whether it's based on the actual merits of the proposal or merely a misunderstanding of what the proposal is all about.


 * In an effort to clarify this, I'm going to make one point, then show the sample warning that I promised William just a minute ago. The point I want to make is that this proposal is about making most 3RR blocks obsolete, not making it harder to block inveterate 3RR violators.  If anything, blocking would be easier because any revert after the warning is an automatic block with no further explanation needed.


 * Anyhow, here's a strongly-worded version of the proposed warning-with-teeth. If this proposal is accepted, we would likely tone down the words slightly while retaining the same meaning.  For now, I'm leaving it at full force for maximum comprehensibility.


 * You are receiving this warning because you are about to violate the three-revert rule. If you revert : one more time in this 24 hour period, I will file a WP:ANI/3RR report that will get you blocked for at least 24 hours, with increasing blocks for repeat offenders and the risk of being permanently banned.  In addition, your last revert will itself be reverted, making it pointless.


 * Please read WP:3RR for an explanation of what qualifies as a revert, and what the exceptions are. If you believe that this warning has been placed in error, you are free to brielfy respond below with an explanation.  However, unless your explanation has been accepted by me or a third-party admin, and this warning is canceled, any further revert is grounds for immediate blocking.  Likewise, do not remove the warning while it is still in effect, or you will be blocked.


 * Does this help clear things up, William? Al  20:36, 28 June 2006 (UTC)

Warning is a good idea of course, and almost always done for new editors. However, there is no guarantee that this specific warning is accurate, and experienced editors should not require a warning of any sort, but should know better. What's worse, these kinds of notices have, in the past, been used as a means of harassing editors, who obviously know quite well that they have reverted 3 times, because the person warning them has also reverted 3 times. Also, the fact that so many people are being blocked for 3RR is not an indication that the policy is not working, but of exactly the opposite; if there is any failing, it is in people who persist in edit-warring to this extent. In any event, as has been stated many times, the idea is to stop edit-warring, not create loopholes for the same, which this proposal will do. Jayjg <small style="color:darkgreen;">(talk) 16:33, 29 June 2006 (UTC)


 * Editors can harass each other with the current warningss, so this doesn't make anything worse. For that matter, nothing in the policy I've suggested offers loopholes; I have absolutely no idea where you get that conclusion, so you'll just have to explain.
 * Frankly, it is unclear why experienced editors would not benefit from warnings. Just because someone is experienced doesn't mean they can't lose track of their edit history, and it's entirely possible to violate 3RR accidentally and without edit-warring.
 * If the goal of the policy is to stop edit wars, then anything we do in excess of what's necessary to achieve this goal is overkill, particularly when it causes additional harm. And, yes, blocks do cause additional harm.  A blocked editor can't go off and edit less controversial articles and remain a part of the Wikipedia community.  On a purely practical basis, a block is a lot more work for everyone, and that means admins are spread thinner and have less time to be careful and avoid errors.  It also pissed editors off and encourages them to leave the community.
 * Will we still need blocks with the proposed rules? Sometimes, yes, but only rarely.  This is good reason to move forward to a kinder, more gentle policy.  Al  06:29, 30 June 2006 (UTC)

Unnecessary and harmful. Creates an unnecessary step who's main effect will be to provide for a loophole allowing chronic edit warriors to get off the hook on a technicality. I won't support this in any shape. Doesn't help that some chronic edit-warriors and anons are pushing for it here. FeloniousMonk 17:06, 29 June 2006 (UTC)


 * With all due respect, your comment does not sugges that you read the proposal. You see, the warning is not an additional step, it is a replacement for a few steps.  The point is that a warning with bite will end the war without the need for any further intervention, much less a block.  Now, if you want to show that you understand this claim and address it directly, I would be interested in hearing your views.  As it stands, I don't believe you have addressed the actual proposal, so I'm not sure how relevant your comments are.
 * By the way, it turns out that you just got off from a 3RR violation on a technicality -- you reverted yourself after it your 4th revert was reported but before you were blocked. Does this mean you're an evil "edit-warrior" and we should disregard your views?  I think that's a silly question, amounting to poisoning the well.  Instead, I suggest we'd all do better if we actually addressed the merits of the arguments, not argued about each other's merits.  Al  06:29, 30 June 2006 (UTC)

Warnings would just add yet another layer of rules, another thing for people to wiki-lawyer about. There is already protection for new users who may be unaware of the rules; most Admins already try to avoid BITE-ing newcomers. This would open the door to even more abuse than we have currently; IMHO the goal is correctly to reduce edit warring, not increase wiki-lawyering. And a "warning with teeth"? No. Let's not add a loophole and then compound the error by making the notification nasty. KillerChihuahua?!? 18:49, 29 June 2006 (UTC)


 * Please see above for my response to FeloniousMonk, as the first section applies just as well to your comments. Thank you.  Al  06:29, 30 June 2006 (UTC)


 * I don't think a short-term block for 3RR is that horrible a thing. Unless the user is really asking for it, I suspect they won't get much more than a block that suggests a brief cool-down period. I don't see this change as necessary. --Improv 23:58, 29 June 2006 (UTC)


 * This might be true if blocks were always limited to 12 hours. However, admins increase block duration when there is a record of previous blocks, and this increase does not take into account the correctness of those previous blocks.  As a result, blocks can quickly reach levels where they are no longer preventative but instead punative, which is entirely against WP:BLOCK and the good of the community.  Al  06:29, 30 June 2006 (UTC)


 * This rule has served us admirably so far and one that is really simple to grasp: Don't want to be blocked, don't revert more than three times in 24 hrs; and if you do, you will enjoy a forced wikibreak, with a chance for you to assess if it is worth doing it again. Sweet and simple. I do not see the need to change it... ≈ jossi ≈  t &bull; @ 04:29, 30 June 2006 (UTC)


 * If it ain't broke, let's not fix it. Except that it's broke.  I've explained above in some detail just what's broken, and you have not addressed any of this.  Presumably, you have some refutation, but you have not chosen to share it thus far.  Al  06:29, 30 June 2006 (UTC)


 * I think it is a mistake to consider a 3RR block "punishment." For this reason, I think the police analogy is misplaced, and concern for people unfairly punished mislplaced.  Multiple reverts are symptoms of an over-heated argument.  The purpose of the 3RR block is to allow for some cooling-off.  It is not punishment, it is not like putting someone in prison or lashing someone.  It has no bad affects.  I feel very strtongly about this: not editing Wikipedia for 24 hours is not a bad thing, in fact, i think it is a good thing for everyone to do periodically.  I have another argument, which is that people who are new to Wikipedia have a responsibility to familiarize themselves with the rules, but this is not really my argument for keeping the 3RR unchanged, and if this were someone else's argument I would tell them I am not satisfied by the argument.  Here is why: even if there were no 3RR, if a newbie found themselves reverted twice and reverted someone else a few times, I would say it is just common sense - especially for newbies not familiar with our rules - to take a short (e.g. one day) break from editing, and use the time to get a handle on their own passion for their edit, the other person's response, and perhaps even take some time to learn more about Wikipedia.  I repeat my main point: this is not punishment.  It is constructive help. Slrubenstein   |  Talk 09:37, 30 June 2006 (UTC)


 * If so, it's a mistake shared by many admins. The idea that a block is just for cooling off might be valid if people were blocked only from specific articles, but they're not.  Instead, blocks are overkill in that they prevent any participation.  This is a bad effect.  If you disagree, go ahead and block yourself for a month, ok?  :-)  Please, a block is obviously a disruption to your ability to participate and is harmful to you.  No amount of spin can change this.
 * I've also pointed out a few times that 3RR violation is not the same thing as edit-warring, or even related, which is precisely why we shouldn't be so quick to use blocks when better ways are available. Unfortunately, you didn't address any of this.  Al  21:38, 30 June 2006 (UTC)


 * Nah, I don't think the idea has merit. Hiding Talk 11:49, 30 June 2006 (UTC)


 * This is a fine example of why voting is evil. Perhaps you have good reason to think the idea is bad, but we'll never know, because you haven't said.  For all we know, you're just being dimissive or have a terrible reason, such as one based on misunderstanding the proposal.  Al  21:38, 30 June 2006 (UTC)

Example revision
I noticed that under the recent changes to the 3RR policy and discussion here the example is no longer valid. Specifically the following:
 * Note that if User B's last change had been, "Some critics contend that they are the worst band ever.", then none of User A's text would have been removed and it would not have counted as a revert.

This is especially true when you consider the following sentence from the policy:
 * A revert may involve as little as adding or deleting a few words or even one word.

Seems to me that even if User B only added "Some critics content that..." and left "the worst band ever." in the 4th edit it still meets the definition of a revert and gets them block. --Bobblehead 20:52, 28 June 2006 (UTC)

Example
I've just cut the example to here:

Example
Supposing that the following edits were all made within one twenty-four hour period:


 * User A adds a new sentence: "They are the worst band ever."
 * User B removes this sentence with the edit summary, 'Not neutral point of view' (1st B revert)
 * User A reverts with no edit summary (1st A revert)
 * User B changes to, "Many consider them a poor band." with edit summary, 'more neutral'. (2nd B revert - 'They are the worst ... ever' removed)
 * User A reverts with no edit summary (2nd A revert)
 * User B changes to, "They are a very bad band." with edit summary, 'less severe evaluation' (3rd B revert - 'the worst ... ever' removed)
 * User A reverts with no edit summary (3rd A revert)
 * User B changes to, "Some critics contend that they are one of the worst bands now touring." (4th B revert - 'ever' removed)

Then 'User B' would technically have violated 3RR, by removing the word "ever" four times, and 'User A' would not. Note that if User B's last change had been, "Some critics contend that they are the worst band ever.", then none of User A's text would have been removed and it would not have counted as a revert. This focus on precise wording rather than the users' intentions may seem counter-intuitive, but the purpose of 3RR is to provide a non-subjective upper limit to reversions in order to stop out of control edit warring. Admins may, however, still make subjective determinations about edit warring and one evaluating the situation above might block both users if they felt that 'User A' had not been making a good faith effort.

It only got added recently - I didn't notice at the time. If anyone re-adds it I won't re-remove. But... I feel that some of this is misleading. It should be headed "contrived example" or "worst-case example" - its not a common-or-garden example. Also I think some of the explanatory text is wrong/misleading... I'm not at all sure I would block B in this case... —The preceding unsigned comment was added by William M. Connolley (talk • contribs).
 * It was meant to be a 'worst case' example (indeed, I believe that was my edit summary)... because there have been blocks along these lines and people have contested that it was not the intent of the policy. Just above someone claims that it doesn't go far enough... by their interpretation just adding new words is a 'revert' (which makes no logical sense under the meaning of the word 'revert') - so anyone expanding an article by adding new info four times in a day is guilty of 3RR if there were intervening edits. This is why I think an example is needed. You say the example goes further than the policy intends, Bobblehead says that it falls short of what the policy intends... that's way too much 'wiggle room' for something which is supposed to be non-subjective. So far as I can see the example is directly in line with the policy... user B removed the word "ever" four times. How is that not a 3RR violation? Heck, under the clear wording of the policy if he had removed any four words once each it would be a 3RR violation. Mind you, I've never been a big fan of 3RR blocks... I've seen virtually identical situations receive completely different treatment far too many times. If we are going to have 3RR blocks the standards should be absolutely clear. --CBD 22:08, 28 June 2006 (UTC)
 * Perhaps I over spoke my confidence level that the example was no longer valid. I had originally intended it to show my confusion on the matter, but apparently in one of my revisions before submitting the text I removed that confusion. Heh. I'm perfectly willing to admit that I misinterpretted the meaning behind "A revert may involve as little as adding or deleting a few words or even one word." If adding "Some critics contend that" while maintaining the word "ever" would have saved User B from a block, then the example is valid. The confusion arises in that I have seen users blocked for doing exactly what the example suggests User B do, leave the entire addition made by User A and make additions of their own. The reason that has been given for the block is that 3RR is intended to stop edit warring and that User B was merely making the addition to game the system in order to avoid 3RR. This is especially true if User A came in later and removed "Some critics contend that" and User B submitted a 3RR violation. --Bobblehead 23:26, 28 June 2006 (UTC)

You know, the fact that there can be such ambiguous cases for 3RR violation gives a boost to my warn-first proposal, above. Al 23:27, 28 June 2006 (UTC)
 * I disagree. The warning suggestion is a separate issue; warnings will not clarify abiguous cases. KillerChihuahua?!? 18:45, 29 June 2006 (UTC)


 * Warning requirements are not the same thing as warning suggestions. A suggested warning is just another step in the process, while a required on is a replacement for the blocking process in most cases.  In other words, the sort of strongly-worded warning that I've given a sample of would be almost as effective as a block, with much less effort or harm.
 * Moreover, in ambiguous cases, the editor has no clue that they've done anything wrong (and, in fact, they may well not have). Under the current system, though, they have no warning: the first hint of trouble is that they're blocked, and then the harm's already been done.  They've been cut off from their work and their reputation has been permanently sullied, making any future blocks longer.
 * In contrast, if they get a warning first, they can stop and investigate the claim. Perhaps they'll find out that they were inadvertently violating a rule, perhaps they'll find out that the warning was in error (such as in the recent Infinity0 case). If it's the former, then we've succeeded in stopping them without a block, and if it's the latter, then it can be resolved without ever blocking.
 * In the end, warning first allows blocks to be avoided in a variety of cases, without losing our ability to stop edit wars. Al  21:38, 30 June 2006 (UTC)


 * KillerChihuahua is right. What we need to do is effectively communicate the spirit of the policy.  In the case presented above, the person reverting does not need to start with a "warning" about the 3RR but rather needs to take the time on the talk page or the other editor's talk page to explain the NPOV policy and how it applies to this case.  Also, any admin who blocks the reverter should pause and look carefully at what is going on and make sure that (1)she is enforcing the spirit of the policy, and (2) perhaps help out by taking time to explain NPOV.  But the problem here is not someone's lack of awareness of the 3RR rule, it is someone's lack of understanding of NPOV. Slrubenstein   |  Talk 14:14, 30 June 2006 (UTC)


 * It would be a fine idea for the block warning to direct the editor to the Talk page as well as offer other advice on avoiding edit wars. Of course, you're assuming that an edit war is involved, which is not always the case.
 * The reality is that admins who block on 3RR do so almost mechanically. There are so many blocks every day that there just isn't time to give each one personal attention and due diligence.  Instead, they check to make sure the links in the 3RR violation report look legitimate, then block.
 * The problem is not lack of awareness of the 3RR rule, but lack of awareness that they are about to break it and that this will result in a block. That's why the warning doesn't just say "Hey, check out 3RR, just in case you never heard of it".
 * In the end, nobody has offered any justification for blocking instead of warning. The warning stops the unwanted behavior without the harmful side-effects, so what is our excuse for shooting first?  Al  21:38, 30 June 2006 (UTC)

Interesting proposal. I think it would be a waste of time to warn at the 3rd revert, since many are not planning on going over 3 reverts anyways. But if the warning were given after the 4th revert, where a block might ordinarily be given, I agree that it would avoid a lot of headache, giving the person either the chance to explain in case of mistake, the chance to voluntarily withdraw from further reverts (while still being able to contribute elsewhere), or the chance to further reinforce the need for a block. Simply blocking a person with no discussion can seem so cavalier, almost rude at times. No wonder people get upset; it's uncivil. There needs to be an intermediate step. ^^James^^ 23:57, 30 June 2006 (UTC)


 * I think the proposal has solutions for both cases. For someone who's at three reverts but hasn't yet violated 3RR, the warning reminds them that they can't revert anymore without a block, and suggests that they avoid edit-warring.  For someone who's violated 3RR but hasn't been warned, the warning explains why their last revert has itself been reverted, and likewise puts them in the sudden-death position of a block after one more revert.  In both cases, nobody enters into sudden-death without warning, and this avoids all sorts of problems.
 * Having said that, you offer some good points. What if, in each case that we would block under the proposal, we offered an alternative?
 * This alternative is to be released on your own recognizance, with the understanding that you agree not to edit that particular article for 24 hours. If you do, you get blocked for twice as long as you otherwise would have.  Otherwise, you get to contribute positively to Wikipedia elsewhere while cooling off with regard to this article.  What do you think?  Al  00:09, 1 July 2006 (UTC)


 * Overly complicated, and unecessary: Don't want to be blocked? Don't revert more than three times in 24 hrs. See KISS ≈ jossi ≈  t &bull; @ 00:35, 1 July 2006 (UTC)

What's unnecessary is blocking someone from the whole of Wikipedia over a mistake on one article. You show a distinct lack of empathy. Al 00:53, 1 July 2006 (UTC)


 * I'd like to re-emphasize that the 3RR does not require that offenders be banned -- it merely gives admins the option. There's no reason why a more moderate system of punishment can't be encouraged.  CJCurrie 01:02, 1 July 2006 (UTC)

You're correct, of course. Unfortunately, if admins have the option to block without warning, they generally will, leading to surprise blocks where warnings would have sufficed. By requiring a warning, albeit one with some teeth, admins will use blocks as a last resort, not a first, and nobody will be blocked or even enter sudden-death without at least a warning. It's not a lot to ask for, but some admins won't grant it, which mystifies me. Al 02:22, 1 July 2006 (UTC)


 * For the record, I'm an admin and I generally give offenders a chance to self-revert. (I also don't block users that I'm involved in active disputes with, just as a matter of principle).  CJCurrie 02:27, 1 July 2006 (UTC)


 * To be frank, you're unusual. I have thus far been given ZERO chances to self-revert, even in cases where it's unclear whether I've genuinely violated 3RR, and I've even been blocked by an admin who was involved in an active dispute with me.


 * There will always been good admins and bad ones, and the former will be gentle, reasonable and fair no matter what the rules allow. However, the latter can only be held in check by rules that force even the worst admin to act just like the best.  The policy I'm proposing not only reigns in the "rouges", so to speak, but makes life easier for the rest.  More to the point, it makes Wikipedia a fairer place.  Despite this, the amount and (especially) type of resistance is staggering.  Al  02:35, 1 July 2006 (UTC)

I don't see much emphasis on warnings etc. on the blocking page, and I think that's the crux of the issue. Allowing admins discretion is important, but perhaps a formal intermediate step, that is strongly encouraged by policy except in the more overt cases, would be beneficial. Many admins already do this, so it's not that radical a change. ^^James^^ 02:48, 1 July 2006 (UTC)


 * Indeed. As it stands, there's too much discretion, yielding too much variability, hence an unpredictable experience for editors.  And it's more than that.  Even with good admins who only block when there are no other good choices, there's still an overreliance on global blocks to solve local problems.  I'm going to take a lot of the feedback I've gotten on this proposal and write up a new one, implementing a few of the suggestions and anticipating any misunderstandings.
 * It's come to my attention that most people who respond at all lack the time to read the proposal carefully, much less to follow up on any feedback, corrections and rebuttals. Instead, they treat it as a vote, which is highly counterproductive.  I'm going to have to find some way to deal with this problem.  Al  03:55, 1 July 2006 (UTC)
 * Not only that, some admins simply abuse their blocking powers on the page and don't adhere to the spirit of the 3rr. Some don't even really look at the page and attempt to figure out what's going on.  If you need a good example look at my block log specifically the following 2 blocks "03:55, March 21, 2006 William M. Connolley unblocked SeraphimXI (contribs) (Mistaken block. My apologies) 03:35, March 21, 2006 William M. Connolley blocked "SeraphimXI (contribs)" with an expiry time of 3 hours (3rr on Freemasonry)", in that case the blocking admin looked at the post, saw that someone posted 4 edits that I made on the page in less then 24 hrs and simply blocked me.  He didn't even bother to look at the page, or look at the comments I posted under the 3rr report pointing out that there was only 1 revert and that the edits were to seperate areas of the same page.  Only after he blocked me, did he bother to look at the case, and then 20 minutes later he unblocked me.  Stuff like this should NEVER happen.  Admins must be required to understand the situation before blocking, requiring a warning before the block would be a good step in the right direction, especially if that means that a 3rr violator can only be blocked if he continues after being warned, because i've seen alot of examples of people putting up the warning at the same time they put up the 3rr report, which is pointless.  Admins need to stop using the 3rr page as a way of boosting their block count, and they must be required to atleast spend 10 minutes to look over the case and understand what is going on, and they need to realize that if the edit war is over there is no problem and no block should be handed out.  I think I might start up a page keeping track of incorrect 3rr blocks, some people seem to believe that there isn't a problem.  Seraphim  04:20, 1 July 2006 (UTC)


 * I'm familiar with this particular admin, but I don't the issue is specific to them. Given how many 3RR blocks admins have to deal with daily, it's inevitable that most receive no more than a cursory review before enforcement.  You were actually lucky in that the admin actually undid their own block; this doesn't always happen.
 * The only practical solution is to shift workload away from the admins. In fact, that's one of the obvious goals of the WP:ANI/3RR system, in that it turns every editor into a 3RR-finder while making it so tedious to report a violation that people won't do it casually.  However, it still fails in that there is no distance between the search and destroy steps.  As you mentioned, a warning in and of itself offers no buffer unless (as per the proposal), blocks can only occur for actions taken after the warning.


 * Do you really believe that admins are into "boosting their block count"? What would that do to an admin? Absolutely nothing. Please stop all this anti-admin nonsense. Have a complaint about a specific sysop? Place a notice at WP:ANI. ≈ jossi ≈ t &bull; @ 04:35, 1 July 2006 (UTC)


 * If they aren't simply trying to boost their block count, why can they not spend the time to actually look at each situation before they block. Saying that they have alot of work to do so it's ok for the admins to just glance at the situation and block is junk. If the admin doesn't want to take the time to look over each situation and make sure a 3rr block is needed then they should not be handing out blocks.  Right now sure the 3rr page is designed to take the work off the admins, however the admins sometimes rely on the reports too much, forgetting that atleast 50% of the time the person filing the report is one of the heated participants in that edit war. What I want is simple, for the admins to take the time to verify that the 3rr was actually violated, look at the page history/talk to verify that there is still an issue and to block all the users on the page that broke the 3rr, not just the ones that were reported.  It's really not that hard, 10 minutes max for each block, and if done right there will be no more need for admins to point out that they allow other admins to review their 3rr blocks, if there is any question if someone should be blocked or not, they should not be blocked.  Seraphim  03:23, 2 July 2006 (UTC)


 * Seraphim, I don't know that they're particularly concerned about their block count, as such. Rather, they're just trying to keep up with a huge list of probably-true blocks, so they don't spend enough time on each.
 * Now, it's easy to tell the admins to just try harder, but this doesn't work. The right way is to make their jobs easier by ending edit wars before admins have to get involved.  That's the goal of the proposed rule change.
 * Nobody intentionally violates 3RR with the hope of getting blocked. If they know they're at their last revert and being watched, they'll just stop.  This means no block will be needed, and no admin will have to spend 10 minutes or even 2 minutes to implement the block.  Al  03:34, 2 July 2006 (UTC)


 * Another option would be to add a short version of the 3RR to the admonitions located at the bottom of every edit page. That way everybody is warned. -Will Beback 05:03, 1 July 2006 (UTC)


 * See, that's what I would call a toothless token warning. It offers no hint to an editor that someone thinks they're about to violate 3RR.  It therefore offers no protection from being blocked without any clue that it was going to happen. In short, it solves nothing.  Al  13:00, 1 July 2006 (UTC)

Above, CJCurrie wrote, "There's no reason why a more moderate system of punishment can't be encouraged." I strongly disagree with this characterization of the 3RR. I do not see it as a punishment, and I do not think it is constructive to present it as a punishment. Blocking someone from editing an article for 24 hours does no harm to anyone, and the purpose is to allow for things to cool down. We expect someone who has been blocked for 24 hours to return to the article in question and work on it, we just hope that he or she will be more effective in editing when they do. Hardly punishment. Slrubenstein  |  Talk 11:39, 1 July 2006 (UTC)


 * It's not supposed to be punishment, according to WP:BLOCK, yet that's what it is. A truly preventative response would stop just the edit war (if there is one) on that article and not affect the editor's ability to contribute productively on less controversial articles, much less participate in the community.
 * Read some of the comments people leave on their user pages when blocked. They're not cooling down; they're angry.  The longer the block, the angrier they get.  And when the block ends, they come back angry, particularly at admins for using excessive force.  And, you know, I don't blame anyone for being angry in such a situation.
 * Anger comes from frustration, and having someone lock you out of the system while you're trying to use it is designed to be frustrating. However, anger is hardly the worst effect.  If I had to choose, I'd say that the most negative consequence is that a block leaves people disenchanted with Wikipedia.  They feel that their efforts are wasted and lose a lot of their motivation to continue.  This is how we lose good editors.
 * We need to get past the double-speak and be honest here. Blocks were designed to lock out vandals, and that's about all they're good for.  Yes, if all else fails, then a block can always be used, but it's the last resort, not the first.  If you think the system isn't broken, prove it by blocking yourself for a week.  What?  Don't want to?  Didn't think so.  Al  13:00, 1 July 2006 (UTC)
 * Well, if we're going to "get past the double-speak and be honest here", then we need to recognize that the only reason for this suggestion is so that edit-warriors can avoid getting blocked for their 3RR violations. Jayjg <small style="color:darkgreen;">(talk) 03:04, 2 July 2006 (UTC)


 * Please, if we're going to "get past the double-speak and be honest here", then we need to recognize that you are the admin who once blocked me despite being actively involved in a dispute, and that the block was later removed, with much criticism laid at your feet over the extreme duration.
 * It is precisely this sort of admin sloppiness that leads me to propose that we should warn first, block later. As for edit-warriors, the big problem with 3RR is that real edit-warriors know how to game it, while good editors are capable of accidentally violating it precisely because they're not keenly aware of every edit they've made in the last 24 hours.
 * In the end, blocking without warning for WP:3RR is punitive, violates WP:BLOCK and is harmful to Wikipedia. Al  03:12, 2 July 2006 (UTC)


 * Um, does any of that personal stuff you included in your remarkably inaccurate diatribe regarding me have anything to do with 3RR policy? Jayjg <small style="color:darkgreen;">(talk) 03:52, 2 July 2006 (UTC)


 * I have no idea what you're talking about, but your characterization is over the top and hardly civil. I could give you a civility warning, of course, but such a thing would be toothless.
 * The current 3RR warnings are just as toothless, which is why we need a policy that gives them teeth, even if it displeases the "rouges" of the world. Al  04:08, 2 July 2006 (UTC)


 * "Well, if we're going to "get past the double-speak and be honest here", then we need to recognize that the only reason for this suggestion is so that edit-warriors can avoid getting blocked for their 3RR violations." that's hogwash. Right now people get blocked for 3rr sometimes up to 12 hrs after the edit war ended and an agreement was reached.  What is the benefit to wikipedia as a whole by blocking users that managed to reach a consensus? Sure there was an edit war, but if they can work it out amongst themselves there is no reason to block them. There is no "cool-down period" if the edit war is concluded already and a consensus reached.  It's simply a punishment, something which violates WP:BLOCK and the "3rr is not a punishment" line.  Seraphim  03:17, 2 July 2006 (UTC)
 * If that has ever happened, it's an extremely rare occurrence. There's no point in introducting huge loopholes in a policy based on one alleged situation where it was not applied in a reasonable way. Jayjg <small style="color:darkgreen;">(talk) 03:52, 2 July 2006 (UTC)
 * How is it a loop hole? If the talk page shows that a compromize was reached, it's reached. It's not like a user can unilaterally declare an editwar is over.  Seraphim  06:13, 2 July 2006 (UTC)


 * It makes sense to call it a loophole if and only if your actual goal is to block people, as opposed to ending edit wars. Al  06:35, 2 July 2006 (UTC)


 * You have yet to explain how the proposed policy offers any loopholes. Thank you for understanding.  Al  04:08, 2 July 2006 (UTC)

Speak for yourself. When I have been blocked, I have never gotten angry and the cool-down always helped. I know many users who respond this way. Slrubenstein  |  Talk 13:28, 1 July 2006 (UTC)


 * And such users can be safely disregarded, as this proposal doesn't affect them one way or the other. Al  14:50, 1 July 2006 (UTC)
 * Blocks are used to teach 3RR violators to stop violating 3RR. Usually it works. Some people cannot learn.  In either case, it's educational, not punitive. Jayjg <small style="color:darkgreen;">(talk) 03:04, 2 July 2006 (UTC)


 * That's like saying beatings are used to teach children and wives to obey. Al  03:12, 2 July 2006 (UTC)
 * Actually, it's nothing like that; prejudicial strawman analogies are not really useful. Jayjg <small style="color:darkgreen;">(talk) 03:52, 2 July 2006 (UTC)
 * Causing someone harm and aggrevation to "teach them a lesson" is exactly what we're talking about. Blocking without warning is punitive, not preventative.  If you disagree, offer me an argument instead of yet another unsupported conclusion.  Thank you for your time.  Al  04:08, 2 July 2006 (UTC)

If this is based on a personal experience of a couple of editors, they can make a complaint through the available channels, rather than attempt to change policy. The policy, as is, works remarkably well. ≈ jossi ≈ t &bull; @ 03:56, 2 July 2006 (UTC)


 * It's based on problems with policy that allow admin errors to harm editors. If it were just a few "rouges", then dealing with them would be the proper course. The policy, as is, works terribly, unless you think unnecessary harm is a virtue.  Al  04:08, 2 July 2006 (UTC)
 * Harming editors, you say? Can you please explain how a forced wikibreak can "harm" an POV warrior that engages in revert, after revert, rather than discuss and use the dispute resolution process of there is not consensus? ≈ jossi ≈ t &bull; @ 04:14, 2 July 2006 (UTC)


 * A forced break from one article might be justifiable when there is an intractable conflict occuring in that one article. A forced break from all articles is not justified at all by this, and is only harmful.  It's punitive, not preventative, and thus violated WP:BLOCK.  Al  04:17, 2 July 2006 (UTC)

Jossi, if edit-warring is, as you say, a bad thing, then you should lead by example. Start by walking away from your edit war to hide all mention of the view that Objectivism is a cult. Show us that admins are better than the editors they ban, please. Al 04:19, 2 July 2006 (UTC)
 * This talk page is to discuss policy, not to discuss content disputes. ≈ jossi ≈ t &bull; @ 04:25, 2 July 2006 (UTC)

Yes, and this policy is about averting edit wars, so the fact that you're an edit warrior is relevant. Al 04:32, 2 July 2006 (UTC)

Anyhow, my point remains. You're here arguing about how well the rules stop edit-warring, yet you're edit-warring yourself and no rules stop you. You're an admin, so you should be an exemplar of proper behavior. Instead, you're a role model for how to succeed as an edit warrior without getting caught. This is distressing. Al 04:58, 2 July 2006 (UTC)
 * If you have a complaint about my behavior, you are welcome to place a comment at WP:ANI. ≈ jossi ≈ t &bull; @ 05:45, 2 July 2006 (UTC)

Thank you for the suggestion. I'll file it under "tilting at windmills". Al 06:35, 2 July 2006 (UTC)


 * If we don't want wikipedia to go the way of the Stanford Prison Experiment then we need these blocks to only be used in clear cases. This rule is so lax I have also seen blocks imposed many hours after the events, and the current multiple possible definitions of 4 reverts, as shown in the above examples, makes it open to creative interpretation - never good for a fair and open system. Protesting your block is only likely to get it lengthened and you accused of a PA attack for effectively calling the blocking admin incompetent - get out of that one. Clever POV warriors don't get caught and I still say this rule encourages full page reverts to save on your "allowance". Sophia  08:51, 2 July 2006 (UTC)

Concrete Example of a Problem
This is a RFC report that was filed on the ANI 3rr page

User:Matyldalondyn and User:87.227.28.6 reported by User:Batman2005
Matyldalondyn as well as his (probable) IP address User:87.227.28.6 have violated the 3RR rule on the Lukas Podolski page.

Edits for Matyldalondyn
 * 1.
 * 2.
 * 3.
 * 4.
 * 5.

Edits for 87.227.28.6
 * 1.
 * 2.
 * 3.
 * 4.

Please do something about this as well as the sock puppetry being employed by this user to input trivia into an article that has been reverted by 3 or 4 users as dubious and unsourced. Time Report Made 22:09, 2 July 2006 (UTC) Batman2005 22:09, 2 July 2006 (UTC)

After toying with sprotect I've blocked both for 12h William M. Connolley 22:34, 2 July 2006 (UTC)

- The problem in this situation is simple and obvious to anyone that takes even 20 seconds to view the page history. Why is the person who reported the 3rr violation given a free pass? He was just as guilty of violating the 3rr, infact he violated the 6rr! (7 reverts) It's blatantly obvious to anyone who looks at the page's history that there is an edit war going on between atleast 2 parties, the fact that you are the first to report your enemy does not give you a free pass. In this case the user who reported the problem basically "won" the edit war due to the action of the blocking admin This is a PROBLEM. Cases like this are simply unacceptable, and are a great example of why the 3rr policy and the administrator enforcement of it is severely broken. Seraphim 23:19, 2 July 2006 (UTC)


 * Since it appears Seraphim  has not read the 3RR policies, I'll explain here.

3RR clearly states the the removal of vandalism is not a violation of 3RR. Furthermore, i had posted the repeated vandalism on and no action had been taken thus far. The removal of vandalism is allowed, and I don't particularly appreciate this user trying to make me her example for not liking the way a policy works. Batman2005 23:40, 2 July 2006 (UTC)
 * 1) continuously inserting dubious, unsourced information into an article is vandalism.
 * 2) sockpuppetry to do the above is also vandalism.


 * I'm not using you as an example, i'm showing how the admin enforcement of the rule is inconsistent and that a problem does exist. I'm using WMC as an example, not you. Seraphim  23:43, 2 July 2006 (UTC)


 * Per Batman2005, "continuously inserting dubious, unsourced information into an article is vandalism." Umm.... no. No it isn't. Read WP:VAND. --CBD 09:59, 3 July 2006 (UTC)


 * Umm.... yes. Yes it is. See "misinformation."  See the warning template that clearly says "Please do not add nonsense to Wikipedia. It is considered vandalism."  Said user was inserting nonsense onto the pages.  It was determined to be nonsense by another user who verified it himself. Call it what you want, but its vandalism at its finest and my block was bullshit. Makes me not want to do the right thing anymore, from now on i'm just gonna leave said nonsense on pages and let someone else get punished for removing it, so much for wikipedia being about posting the verifiable truth.  What a joke this is. Batman2005 02:52, 4 July 2006 (UTC)


 * The accounts you were edit warring with were trying to insert info about a soccer player not singing the national anthem of Germany when playing against his native Poland. That was presumably based on news stories like this one, and thus an entirely reasonable change. I see no discussion of the issue on the article talk page. Just reverting, false accusations of vandalism, and general incivility on your part. The block was warranted. Please learn to follow the dispute resolution procedures rather than edit warring and swearing when people do things you disagree with. --CBD 11:00, 4 July 2006 (UTC)


 * Please stop talking to me as if you're my father. We disagree, that news report is wrong, as was shown by another editor who clearly stated that the player NEVER sings the national anthem, you call it a false accusation of vandalism, I call it vandalism....and it was. I'm guessing you've not watched the world cup games to see for yourself that the player never sings the national anthem? Batman2005 15:33, 4 July 2006 (UTC)

Everyone's kinda missing the point of my posting of this example. It's not a dispute over if Batman should have been blocked or not, he was in clear violation, the issue was that the blocking admin did NOT take the time to actually look at the situation and do his job. This is a concrete example of why there needs to be some sort of checklist/proceedure that admins must go through when blocking under the 3rr. It's simply unfair to the 1 side that gets blocked for 3rr that he got blocked, but the other warrior didn't because the other warrior knew the 3rr policy enough to post a violation. The way I look at it the person that posts the violation on the ANI page is demonstating that they completly understand what 3rr is and if they break it also they should not be given a free pass, especially when alot of the time the person they are reporting is unaware that the policy even exists. My point is simple, there must be a checklist or proceedure created for 3rr blocks, and admins must be required to follow them. If an admin cannot take the time to follow a simple proceedure they should not be touching the 3rr page. I'd be very interested to hear any objections to creating a simple form that admins must fill out each time they block. Remember, not being able to spend 5/10 minutes investigating a block is an unfair answer, if the admin cannot dedicate atleast 10 minutes per block they shouldn't be viewing the 3rr page. Seraphim 02:39, 6 July 2006 (UTC)


 * Just to reiterate my point, my block was bullshit, and the shit the other guy was doing is vandalism. Batman2005 02:13, 7 July 2006 (UTC)

An altogether immodest observation
Scanning the current state of this talk page, I see that the admins who enforce the 3RR are uniformly, without exception, against all changes which might soften the rule to allow for ordinary human imperfection, in, for example, forgetting how many reverts one has made in the past day.

"Follow 1RR and you won't have anything to worry about," they say.

"Don't allow temptation to cross your mind and you won't have anything to worry about," my pastor says.

To what extent is 3RR enforcement a psychological release for these admins?

After a hard day of fighting vandals and newbies, and POV-pushers, who can blame them for wanting to pop a few blocks? NoNameIsMe 10:16, 3 July 2006 (UTC)
 * Not I! Edit warriors are indeed easy targets, and as pernicious, if not more so, than outright vandals. If you're concerned with losing count of the number of reverts you've made, well, you're edit warring and trying to keep under the radar. The proper solution is "don't edit war, period" -- otherwise you'll catch the attention of one of us who need a psychological release from thumping morons. --jpgordon&#8711;&#8710;&#8711;&#8710; 13:41, 3 July 2006 (UTC)


 * Why is the word "war" used to describe a couple of people who both want to contribute to an accurate encyclopedia, but hold different points of view? (One of whom, I might add, stands a pretty good chance of not really wanting to contribute that much as soon as some frustrated admin lowers the boom.) 209.11.184.1 09:18, 4 July 2006 (UTC)

Use of 3RR to destroy neutrality on debate
I'm afraid of violating the 3RR rule on the article Asian fetish, but a particular user keeps on removing the one side of the debate which makes the article represent a neutral point of view. The Asian fetish debate revolves around users who believe that it exists and users who do not believe it exists. It is similar to gay rights and abortion in that users want the article to express their beliefs. A user keeps on removing the views of people who do not believe it exists, creating a single point of view on a two-sided issue. I don't want to keep on reverting his removals of his opposition's viewpoint, because the 3RR rule applies even if the revert is not the same section. Could someone recommend fair arbitration that would rule both sides of the debate need to be included in the article for the sake of neutrality.?--Dark Tichondrias 05:28, 4 July 2006 (UTC)


 * Actually, said user's edits happen to make the article more NPOV. Hong Qi Gong 06:05, 4 July 2006 (UTC)


 * I feel that a netral point of view on a subject whose existance is heavily debated would include the views of those who do not believe it to exist alongside the views of those who do believe it to exist.--Dark Tichondrias 06:20, 4 July 2006 (UTC)


 * I absolutely agree. But that's not what his edits do.  His edits remove POV statements with no sources, which are also arguably irrelevant.  Hong Qi Gong 06:37, 4 July 2006 (UTC)


 * Similar things happen in many articles.
 * "Over time, contentious articles will grow from edit-war inspiring to eventually reach a compromise that is agreed upon by all involved editors. This equilibrium will inevitably be disturbed by new users who accuse the article of being absurdly one sided and who attempt to rewrite the entire article. This is the cyclical nature of controversial articles."
 * -- Raul's Fifth Law of Wikipedia
 * Imagine what a great place this would be if admins spent more time insuring that both sides of a debate were included than they spent enforcing 3RR.209.11.184.1 09:22, 4 July 2006 (UTC)
 * Admin's aren't allowed to do that, instead of having a group of people that can make binding content decisions wikipedia has the horribly flawed "Assume Good Faith" ideal that everyone is trying to improve the encyclopedia, when the reality is some people do have other objectives.  Seraphim  02:41, 6 July 2006 (UTC)
 * But the assumption of good faith is, very often, dropped with no evidence. 3RR enforcement is a perfect example of this:  If two people with different points of view are rapidly reverting, the official policy is to assume that they are in a "war" and acting maliciously.  Who are the victims?  Where is the evidence of malice?  What are the reasons that letting them duke it out and exaust each other would result in a worse encyclopedia than blocking one or both of them?  This policy, while popular, is not well thought out, and based on the infirmest of foundations. Ptcru 00:03, 11 July 2006 (UTC)

Have editors forgotten Dispute resolution? If you are engaged in a revert war, ask for assistance via an RfC, and if that does not work, proceed with the other steps. Revert wars never achieve anything other than increasing wikistress. The WP:3RR is designed to provide a respite and for editors to cool off. Attempting to resolve a dispute via reverts simply does not work. ≈ jossi ≈ t &bull; @
 * I find that blocking people has the opposite effect of cooling them off. --AaronS 20:17, 10 August 2006 (UTC)

bots
I took out the recent addition re bots. I don't think a blaket exemption like this is right. People should use their judgement, and be able to block bots if they think appropriate. At the very least this should be discussed before being added William M. Connolley 22:56, 5 July 2006 (UTC)

Interpretation again
I constantly encounter 'differing interpretations' of this policy. So, for those who haunt this page... opinions please. In a recent situation it is being argued that adding a new 'tag template' to a page four times in under 24 hours is not four reverts/3RR violation, but rather "one edit and three reverts". The first addition of the tag 'doesn't count' unless that same tag had been placed on the page at some time in the past. Yet, I have seen this ruled the other way previously and all agree that removing the tag four times is a violation. IMO the sequence: 'Add', 'Rmv', 'Add', 'Rmv', 'Add', 'Rmv', 'Add', 'Rmv' means that both the person adding the info/tag and the person removing it have violated 3RR... not just the person removing. Both made the same edit four times. The counter-argument is that since the first edit never existed before it isn't a 'revert'. For example, if I added, "Blocks for 3RR violation are often very subjective and based on wildly differing interpretations of this policy", to the project page and another user reverted me I could re-add it three more times without violating 3RR, but the other person would be in violation if they reverted me each of those four times. Is that the intent? --CBD 12:06, 11 July 2006 (UTC)


 * The counter-argument is that since the first edit never existed before it isn't a 'revert'. Exactly. Its the old "adding junk" problem. If I add (different each time) POV stuff to a page, I make no reverts; the person removing it each time makes 1 revert. Its asymmetrical; this appears to be Just Life. Yet, I have seen this ruled the other way previously - really? Where? William M. Connolley 12:15, 11 July 2006 (UTC)

The case to which CBD is presumably referring is that of Timothy Usher adding a Giovanni33 sockpuppet tag to User:Deuteronomy2000, Giovanni removing it, and Str1977 (who was a witness each time Giovanni33 was caught in sockpuppetry, and who saw the linguistic and behavioural similarities between Giovanni and several new users who were reverting to Giovanni's version) readding it. Deuteronomy is not a puppet of Giovanni, though the others undoubtedly are. CBD left a message for Timothy telling him he had violated 3RR. Timothy denied it. CBD said, "You inserted the same material to the page four times in under 24 hours. That's a 3RR violation... and a fairly straightforward one at that." I said that it isn't because the first edit was not a revert. So here we are. A revert means undoing the work of another editor. In no circumstances can Timothy be considered to be "undoing the work of another editor" when he created a page which did not exist prior to that by adding a sockpuppet template. I'm surprised that anyone could even think that was four reverts. I've never seen an administrator block in such a case. If it did happen, it was clearly an error &mdash; unless, as can happen, the administrator decided to block after only three reverts because the user in question had a history of excessive reverting. The policy says that you can be blocked on fewer than four reverts. AnnH <b style="font-size:medium;">♫</b> 16:38, 11 July 2006 (UTC)
 * I agree. Applying this interpretation would make it a three-edit rule. If it's disruptive edit-warring, block for disruptive edit-warring. Tom Harrison Talk 17:38, 11 July 2006 (UTC)
 * I agree as well. Disruptive edit-warring warrants a block as per WP:BLOCK ≈ jossi ≈ t &bull; @ 18:07, 11 July 2006 (UTC)
 * Which presumably was the basis for blocks I have seen in such cases in the past (though that wasn't the impression I had gotten). So not 3RR, but still clearly edit warring and disruptive activity by all three (Giovanni33, Timothy Usher, and Str1977). --CBD 22:10, 11 July 2006 (UTC)
 * Actually, it was 3RR in the case of Giovanni. Regarding the other two editors, making three reverts is not "clearly edit warring and disruptive activity". People can be blocked for making three reverts, if they have a history of 3RR violations are are considered to be "gaming the system". However, most people who make three reverts within a 24-hour period are not blocked. Str1977 reverted only once. I asked him not to, in the case of Deuteronomy2000, and he immediately agreed. He has been a witness to all the puppetry going on for months, and was online the night that Giovanni posted a denial that he had any connection to the new users who were supporting him &mdash; forgetting that he was logged on as one of them. Regarding Timothy, I'd hesitate to state that it was disruptive activity. Those puppet templates exist for a reason. We don't have templates saying things like "this user is a total moron"; if we did, they would get deleted. So if we legitimately have templates saying that user X is a suspected sockpuppet of user Y, where should those templates go, if not on the user pages of suspected puppets? And are users who follow a user who's known (checkuser and signature blunder) to have used puppets in the past, and revert to his version, support him on talk pages, and vote for what he wants, who post from the same area (at least in the case of those whose IPs are public knowledge), and show similarities in writing style really considered not to be suspected sockpuppets? AnnH <b style="font-size:medium;">♫</b> 22:51, 11 July 2006 (UTC)
 * Regardless of past history or whether it is reasonable to place a 'suspected sockpuppet' tag on a page once... surely it is clear that these edit wars were 'disruptive'. It looks like Timothy and Str1977 just missed 3RR on each page (either stopping after three or making the fourth (and fifth, sixth, et cetera...) after 25+ hours had passed), but if that isn't "edit warring and disruptive activity" then nothing is. Regardless of whether someone is 'right' about a suspected sockpuppet tag or not that isn't the way to go about it. Sometimes we place too much emphasis on the minutiae over the intent... in my view if an edit is likely/apparently intended to provoke an edit war it is just as bad as a 'revert' in terms of edit warring, even if it was the first edit made. Likewise, making your 4th revert just over 24 hours after the first, the 5th over 24 hours after the second, et cetera to stay just outside the '3RR penalty box' is, if anything, worse than just making 4 reverts in 24 hours. Nobody who does such things should be allowed to think they 'have not done anything wrong' and should continue doing so in the future. --CBD 12:00, 12 July 2006 (UTC)


 * As my name has been mentioned in this context, I want to explain myself, if I may.
 * Gio, coming back from a block, entered into a campaign of removing "puppet tags" from different accounts' user pages. Some of these are clearly proven and confirmed puppets (through user-check and/or other evidence as per administrators), while others are suspected ones (also through evidence). He has no business removing the "confirmed" tags from confirmed puppets. He has no business removing the "suspected" tags from suspected puppets. He has no business removing anything at all from another user's page - unless he is that user, but that would confirm the suspicion and validate the tag.
 * Mika and Kecik also entered into such a campaign on their own and others' pages (which is another piece in the puppet puzzle) but because they are only suspected puppets, I hesitated when they edited their own page (in the case of other pages, they have no business either, especially since they are not even mentioned in the tag).
 * The case of Deuteronomy is different alltogether: he is clearly someone's puppet, created solely for informing on an alleged 3RR violation. Now, it was not unreasonable to suspect Gio, especially since he immediately jumped into the malaise. OTOH, there are also points indicating that Gio is not behind Deuteronomy. In the end, there is not conclusive evidence enough for such a suspicion (and it looks like D2000 has had his day in the sun anyway). I reverted the tag removal on his page only once, in the heat of Gio's (or was it Mika or Kecik) tag removal campaign. With such great numbers, one page slipping through youzr fingers is quite easy. Hence my revert.
 * Now, I agree that 3RR should be adhered to by all. I also agree, that a persistent adding of the tag in the case of D2000 can be seen as disruptive. However, in all the other cases, I deem the puppet tag not some normal content on a page (which still would raise the question why Gio removing it on another account's talk page is not vandalism) but as a sort of official WP document or rather a street sign saying "Beware! Here be snakes!" Has the snake the right to remove that sign with impunity, so that it will get more prey? (And before anyone shouts, that's speaking in metaphors, I am not calling Gio a reptile!) Str1977 12:37, 12 July 2006 (UTC)

3RR more than once
Hi, I am sure this has been answered somewhere before. I just could not find it. What happens if the same user is blocked for 3RR on the same article more than once. Is there a crossover limit for something more drastic? -- Lost 07:47, 13 July 2006 (UTC)


 * Nothing is mandated by policy in that area.Geni 07:51, 13 July 2006 (UTC)
 * So I guess it depends on the blocking admin then? Thanks -- Lost 07:55, 13 July 2006 (UTC)
 * Pretty much so -- we have a lot of leeway when it comes to stopping edit wars. Some admins double the timeout for the second offense, and so on -- "if putting you in the corner for 24 hours didn't work to make you stop edit warring, maybe 48 will." --jpgordon&#8711;&#8710;&#8711;&#8710; 13:53, 13 July 2006 (UTC)
 * Thank you -- Lost 15:31, 13 July 2006 (UTC)

Grey area?
In a case where there are only 2-3 constructive editors under a barrage of unconstructive IPs, albeit non-edit-warring ones (4 IPs at 3 edits a piece), 3 reversions per editor can quickly run out, and that is without any content disputes. It seems that this is would create an application of 3RR which diverges from its original purpose, and could be taken advantage of to block editors not engaged in edit-warring. Is this a real, if unfortunate side-effect, or have I just misunderstood 3RR's application? Thanks for your assistance,  Tewfik <sup style="color:#888888;">Talk 22:43, 28 July 2006 (UTC)
 * If someone does eventually reply to this, I would appreciate if they could notify me on my Talk. Thanks,  Tewfik <sup style="color:#888888;">Talk 02:49, 3 August 2006 (UTC)
 * If the non-constructive edits are blatant vandalism, as opposed to content disputes, then revert away. The 3RR does not apply if you are reverting blatant vandalism.  Warning:  this only applies to blatant simple vandalism.  If it is a matter of content disputes, then what one person may consider unconstructive could be what others honestly see as proper (assume good faith).  In that case, the 3RR applies and is meant to prevent edit wars.  If you think someone is changing IPs to circumvent the 3RR, call it to an administrator's attention.  If the non-constructive edits aren't simple vandalism and you don't suspect sockpuppetry, then list the article at Requests for comment.  This will call attention to the problem and hopefully attract non-partisan third parties to help with the issue.  As long as you follow appropriate dispute resolution procedures, it is actually quite difficult for people to use the 3RR policy as a club with which to beat you over the head.  When you take the dispute resolution process into account, I don't see this as a grey area so much the way the system is intended to work.  SWAdair 03:42, 3 August 2006 (UTC)

Discussion of punitive blocking
The current policy says "Blocking is always preventative, not punitive." I don't think this is a good statement to make. I have seen cases at WP:AN/3RR where it takes quite some time to determine if a block was or was not in order. At that point, if it is decided that a block was justified, the edit war may have abated. I don't think the person who violated 3RR should get off easily just because it takes us a while to get the block done. I support removing this sentence. Johntex\talk 17:53, 3 August 2006 (UTC)
 * It being a week since this suggestion was made, and with no objections having been raised, I am making the edit ot the section. Johntex\talk 16:35, 10 August 2006 (UTC)

Two on one
What should be done when one editor is in a revert war with two others. All three are using reversion as an editing method but only the solo editor technically violates 3RR. Should all three be blocked?

To further spice the pie, suppose it is clear that editors A and B share the same POV and editing interests, and revert against editor C at multiple articles. Sockpuppetry is alleged but there is no RFCU to back it up (for whatever reason). Thatcher131 (talk) 20:28, 9 August 2006 (UTC)


 * If only one editor violates 3RR, only one should be blocked, unless there are exceptional circumstances. For example, editors with a history of edit warring can be blocked on fewer than four reverts, especially if they make exactly the same revert just after a block expires; in such cases they're blocked for disruption rather than for 3RR.


 * As regards the sockpuppetry, we know that people can edit from different IPs. We also know that people can be genuine editors who share an IP. If lots of new editors show up at the same article with the same IP, I'd be suspicious. If two editors share a POV on some controversial topic such as Abortion and related articles, or Islam and related articles, but one of them makes a lot of edits about Michael Jackson, and 18-century poetry, while the other makes a lot of edits to articles about computing, and deafness, it is probably that they're not puppets. That's one reason why editors who come to Wikipedia with a strong POV on some religious or political topic should be encouraged to edit unrelated topics (butterflies, Vitamin C, etc.) just for fun. If they're puppets, they only want to push a POV; if they're here because they like Wikipedia, their contributions will show that.


 * But to get back to your question, if one person is reverting against two, then he's trying to force his version against consensus. Part of the rationale behind the three-revert rule is that if you've made three reverts and something other than vandalism really needs to be reverted, someone else will probably do it. If nobody does, then the edit probably isn't as bad as you thought. AnnH <b style="font-size:medium;">♫</b> 21:29, 9 August 2006 (UTC)


 * I'm not entirely sure it's fair to describe two on one as "against consensus" when A has a long history of reverting C on any article he edits related to a particular topic and B is a brand new editor who acts just like A (although A=B is not proven at the moment). Still, interesting comments. More thoughts from anyone else? Thatcher131 (talk) 21:39, 9 August 2006 (UTC)


 * Well, as a party in a similar situation where I'm in the majority (yet not necessarily agreeing with the 'consensus', just enforcing it) and another editor and their possible multitude of sockpuppets are in the minority (the difference being a RFCU and suspected sockpuppet have been filed), I'm going to have to agree with Musical Linguist and add.. Until there is a tie between A and B wikipedia treats them as separate editors. If C suspects that B is a sockpuppet of A then they should file a CheckUser request and while that is going through, either consider avoiding the edits in question for the time being, discuss it on the article's talk page and try to build consensus against A and B, and/or stick within 3RR. --Bobblehead 21:53, 9 August 2006 (UTC)


 * Thatcher131 (talk), I agree with AnnH. Also you can add a welcome template to new user's talk page if not done already. This lets them know that someone is watching them. Make sure that all the editors are aware of 3RR policy. This helps put every editor on notice that Wikipedia has rules that must be followed. If appropriate, you can let editor C that made a 4 rv revert themselves if they promise to go to the talk page and discuss. If the other two editor are good guys they will go to talk page and discuss. You can encourage the new user (Editor B) to engage in discussion. It help to spot sockpuppets, since certain patterns show up in discussions as well as their rv patterns and edit summaries. -- FloNight  talk  19:59, 10 August 2006 (UTC)

Suggestion
This suggestion comes from WP:AN/I. Two notorious trolls, Thewolfstar, RJII, and their numerous sock minions, often disrupt the anarchism-related articles (see Template:Anarchism sidebar for the articles in question) by causing edit wars or by pushing their distinct POVs. Sometimes, good-faith editors in good standing get caught in the crossfire as administrators, not fully understanding the situation, can't distinguish sock puppets or trolls from actual editors, and either block everybody or block the wrong person. The tricky point is that, although reversions can cause disruption, it is necessary to continually revert these socks, because their edits completely derail the progress being made by drawing all attention away from it. It might be helpful to tack a note somewhere on WP:3RR pointing out the special attention that is required when dealing with this part of the encyclopaedia. --AaronS 16:27, 10 August 2006 (UTC)
 * Yes, reverting a sock of a banned user is undoubtedly a defence - but there is no deadline. Post at WP:ANI or WP:AIV, and wait a short while.  There's usually someone about who will wield the cluebat. Just zis <span style="border: 1px; border-style:solid; padding:0px 2px 2px 2px; color:white; background-color:darkblue; font-weight:bold">Guy you know? 22:12, 10 August 2006 (UTC)
 * Oftentimes, actually, nothing happens. That course has definitely been taken before. --AaronS 15:11, 28 August 2006 (UTC)

3RR nullifies NPOV
The pathetic thing about 3RR is that it makes it impossible to uphold the NPOV rule. If one or two biased editors decide that they want to enforce their POV in an article with only one neutral editor on duty (which happens often in my experience), all they have to do is keep on changing the article until the neutral editor can not revert any more. If he reverts once more, he gets banned instead of the policy violators. What happens is that articles become POV for 24 hours, and any vandal or biased editor gets to have a field day on that article in the meantime. And please don't say that an admin can intervene; in my experience they either never do, or don't take action until long after the 24 hour waiting period is up.

If you need a good example of this, check out Ayn Rand. This article, for whatever reason, is avoided almost entirely by admins. It took a year to get one vandal banned from destroying this article, and he wasn't even the worst one (the worst one is still editing it, despite multiple bans for unrelated infractions). Now, a new biased editor is messing with it, and, as usual, no admin is willing to get involved. It's sad that Wikipedia is so willing to coddle vandals at the expense of terminal POV articles. -- LGagnon 19:59, 10 August 2006 (UTC)
 * The same thing happened with anarchism and related articles. I suspect it's because it's a tricky situation -- but that's all the more reason to get interested, not avoid it. It took well over a year to get the main antagonists of those articles banned. --AaronS 20:15, 10 August 2006 (UTC)
 * Funny you should mention anarchism; the guy who took a year to be banned at the Rand article vandalized anarchy-related articles so much there was an article about him in Indymedia pointing out the failings of Wikipedia. -- LGagnon 20:18, 10 August 2006 (UTC)
 * I remember reading that. <b style="color:black;">Ungovernable Force</b><sup style="color:green;">The Wiki Kitchen! 22:05, 10 August 2006 (UTC)
 * You must be speaking of RJII. Randites and the like are probably the hardest to deal with, considering that Objectivism allows for very little compromise. --AaronS 20:21, 10 August 2006 (UTC)
 * That's exactly who I mean. He and User:LaszloWalrus (the worst vandal of the Wikipedian Randist cult) have made a disasterous mess of articles related to Rand, and have made it impossible to make any of them neutral. For all intents and purposes, the NPOV rule doesn't apply to those articles, as the admins often refuse to ban Randists for policy violations on them., they give bizzare excuses for not taking action, such as "it's too confrontational in there" or "I'm not really in the mood to deal with this". I can't help but sometimes wonder if Jimbo is trying to protect follow members of the cult. -- LGagnon 20:55, 10 August 2006 (UTC)


 * No, 3RR does not nullify WP:NPOV. There is no deadline, if there is a POV-pusher then use WP:DR, or go to WP:RFPP. The only exception is WP:LIVING, and that's already exempt. Every POV pusher asserts that their POV is neutral, this is not about content it's abotu edit wars. Just zis <span style="border: 1px; border-style:solid; padding:0px 2px 2px 2px; color:white; background-color:darkblue; font-weight:bold">Guy you know? 22:10, 10 August 2006 (UTC)


 * I've made this argument over and over again, and I'll say it again: Dispute resolution does nothing. It's this very article that proves my point. You have no clue how many of the routes I've taken on the suggestion of admins and came back with nothing. I've tried every method the system provides, and none of them worked. The closest thing to a working solution was to send an admin there to fix the problem, and even then the problem just came back once he left because no vandals and/or biased editors were dealt with (the admin simply fixed up parts of the article). The problem isn't that I haven't tried everything; it's that the admins have tried nothing. -- LGagnon 22:54, 10 August 2006 (UTC)
 * Yeah, it took forever to block RJII even though it was obvious to probably dozens of editors that he was up to no good. He had numerous RFCs and one or two arbitrations and nothing ever came of it. And protecting pages punishes all the editors when there is usually only one or two that are causing problems. Too much collateral damage. <b style="color:black;">Ungovernable Force</b><sup style="color:green;">The Wiki Kitchen! 23:00, 10 August 2006 (UTC)
 * 3RR also seems to facilitate gaming the system. --AaronS 15:47, 14 August 2006 (UTC)

If you need further proof that the admins take forever to get anything done, check Personal attack intervention noticeboard. Another user (one of the guys involved in the Ayn Rand dispute) has not only done 3 post-final warning personal attacks on me, but he has gone so far as to literally declare war on me while the admins stand back and do nothing. -- LGagnon 23:16, 10 August 2006 (UTC)


 * Complaint on WP:Pain actioned--A Y Arktos\talk 00:17, 11 August 2006 (UTC)

Again, the pro-Rand faction has nullified the NPOV policy through the use of 3RR. Several of them have ganged up on the Rand articles, and now have totally altered the POV of them in their favor. And again, admins stand back and allow them to get their way. I wish Jimbo would get his fellow cult members under control so that, at the very least, Wikipedia's articles about Rand wouldn't look like a front for his ideologic leanings. -- LGagnon 04:43, 17 August 2006 (UTC)

Request for guidance
I am reverting edits made by an anonymous user on McComb, Ohio. The anonymous user keeps adding weblinks to the Public Library and Methodist church there. Neither of these is mentioned in the very stubby article. I have explained my rationale for removing these in my edit summaries, in the user's talk page, and at the article's talk page. What else can I do to avoid getting into 3RR trouble here while still keeping these links off the article (or is this spam reversion and OK)(or are these great links to have there)? Any guidance would be appreciated. Thanks in advance, Ruhrfisch 02:39, 14 August 2006 (UTC)


 * Stay within 3RR. What you are doing is not immune to 3RR. Its also not important enough to get worked up about William M. Connolley 07:41, 14 August 2006 (UTC)


 * Thanks very much. I will stay within 3RR (and have to date). The anonymous user has responded to my messages so I think we can work something out. I just didn't want to get into an edit war or other trouble. Thanks again, Ruhrfisch 11:11, 14 August 2006 (UTC)

Removing unreferenced content
There have been some occasions in which I've come close to violating the three-revert rule when removing unreferenced content from articles. In many cases editors simply restore their contributions without citing their sources, and because they weren't the first to revert they have the "last revert" (so to speak). Verifiability supports the removal of such content ("Editors adding new material to an article should cite a reputable source, or it may be removed by any editor"), so I'm wondering if there should be a change to the current 3RR policy. Extraordinary Machine 18:27, 17 August 2006 (UTC)

No, it wouldn't be clear cut enough. Excuses such as reference books at the end of the article and so forth would have to be discussed. Addhoc 18:44, 17 August 2006 (UTC)


 * I didn't make myself clear, sorry. I meant cases in which the editor restoring the material makes no indication whatsoever that the article already contains references supporting it. Obviously, if the editor does say it is already sourced, then that's different. Extraordinary Machine 19:50, 17 August 2006 (UTC)


 * My response was going to be "ok, that would be a bad faith revert, which is considered vandalism", however the WP:Vandalism has been recently been redrafted and bad faith reverts are no longer included. Personally, I would be happier with reverts that don't include adequate explanation being considered vandalism and therefore reversible, than modifying WP:3RR. Overall, however I consider that WP:BRD is a better strategy than multiple reverts in 24 hours. Addhoc 20:08, 17 August 2006 (UTC)


 * I would say that removing unsourced stuff isn't immune from 3RR. Only blatant vandalism is. If its sufficiently obvious, it should be possible to find someone else to help you William M. Connolley 20:38, 17 August 2006 (UTC)


 * There is an exception already for possible libel in biographies of living persons. Jkelly 20:39, 17 August 2006 (UTC)


 * OK, granted. And its caused problems at least once, since he threshold is unclear. But apart from that... William M. Connolley 20:44, 17 August 2006 (UTC)


 * Even Vandalism as an exemption can be a fuzzy line. Trying to come up with a list of objective improvements that are immune to it is probably not a good idea, regardless; the system is already too prone to gaming.  Actually, I'd be willing to consider an experiment of getting rid of WP:3RR entirely if someone could come up with some system that allowed for ending edit wars which didn't encourage sockpuppetry.  Jkelly 23:34, 17 August 2006 (UTC)


 * I'm not comfortable with "parachuting in" users completely unassociated with the dispute and having them revert the other party just so that I avoid breaking the 3RR. That said, the "BOLD, revert, discuss" cycle doesn't always apply, particularly when the other party refuses to discuss the issue. Extraordinary Machine 00:23, 25 August 2006 (UTC)
 * Bold revert discuss is your last best hope. If it doesn't apply, the wiki you're using it on is broken. Big trouble! Please be sure to report such situations! Kim Bruning 08:54, 25 August 2006 (UTC)
 * Wait? When dealing with *un*reasonable people? Hmm, then the people are broken and probably should be blocked. Call in an admin and show what's up. It's very interesting that the 3RR and Verifiability collide, despite claims by some that the latter would be non-negotiable. Could you provide diffs and a log entry? It's time we start providing references for guidelines! :-) Kim Bruning 08:57, 25 August 2006 (UTC)
 * As far as I know, there's no policy or guideline that allows the blocking of users who repeatedly insert unreferenced content without discussion or explanation. Repeatedly changing referenced content can be interpreted as "sneaky vandalism", but that's pretty much it, unless it's considered unreasonable and disruptive to restore uncited content repeatedly. Hmm, maybe verifiability isn't taken seriously enough (IMO). As for diffs, I had a run-in with an editor who kept reinserting unsourced material into and  (see,  and , among others). Extraordinary Machine 15:01, 28 August 2006 (UTC)

socks
''Exceptions to this rule-of-thumb can occur in the case of editors identified as sockpuppets by administrators, the arbitration committee, or developers, where the sockpuppet tag is continually removed from the user page by the user. The page might be protected in those situations.''

The best procedure found so far is to block all socks, and have the socketeer select one single account to use. Hence I've removed this for now. Can someone show in which circumstances specifically this would be the optimal procedure?

Kim Bruning 08:54, 25 August 2006 (UTC)


 * Kim Bruning, this refers to editors that object to a sockpuppet label on their multiple accounts. Lately there has been a flood of users making multiple sockpuppets to complain on AN/I about a blocking admin or and spam users pages for help. Then these editors try to argue that their multiple accounts are appropriate and should not have a sockpuppet template. Of course admins can block the editor and be done with it if they remove the sock template. But editors get into tussles over the tag. If explanation is need in a policy, I think it should be done in sockpuppet policy rather than here. Trying to be too specific in policies can backfire by encouraging wikilawyering for everything not specifically mentioned.  FloNight   talk  10:40, 25 August 2006 (UTC)


 * I'm aware of that. Don't label, just immediately block any unlabeled sock with no warning, then ask the socketeer to choose what their primary account is.


 * Anytime someone has more than one account, and doesn't declare this fact, block all undeclared accounts, and have them designate a primary, and declare or abandon all others.
 * Either be social or give a PowerAnswer, but don't be wishy washy in between. It doesn't work. Be firm and clear. Kim Bruning 11:36, 25 August 2006 (UTC)


 * Not a problem for admins because they can block. It is a problem for editors. The addition to policy is trying to assist helpful editors. Keep them from getting a block because an user keeps removing a tag and they keep putting it back on. Remember that many socks are made out of frustration. The sock user does not agree with some policy or the way a conflict involving them is being handled. Often they are not in an agreeable mood and do not want to accept the decisions made by admins. They take out their frustration on helpful editors that happens to retag the blocked sock account. FloNight   talk  16:36, 25 August 2006 (UTC)


 * WP:AN/I ? :-) Kim Bruning 19:59, 25 August 2006 (UTC)

Moving/renaming
"This policy does apply to repeatedly moving, renaming, deleting..."

Is there any difference between moving and renaming? googl t 16:59, 29 August 2006 (UTC)