Wikipedia talk:Edit warring/Archives/2012/August

The 3RR rule should be considered lenient.

Having gone over much of the talk page and reviewed in full the edit warring article, I am under the impression that the 3RR rule is far more lenient than many users seem to treat it. My belief is that whether or not someone is banned comes down to the context of the supposed edit war. If someone is merely trying to correct a mistake, both parties believe they are correct, and the 3RR rule is mostly obeyed, often neither user is banned. However, if someone is arguing for the sake of arguing with no evidence to support their case (3RR violation or not) they are banned. The 3RR guideline is merely that: a guideline, and should be interpreted as such. —Preceding unsigned comment added by 74.198.165.124 (talk) 15:06, 7 April 2011 (UTC)

Ban? Do you mean a block? There is a difference between a ban and a block. Bejinhan talks 11:14, 12 April 2011 (UTC)

Blocks and bans are to protect the encyclopedia. As such there is no primary need to block a user who is not believed likely to cause disruption in the future - either because they now understand the 3RR, or because they have engaged in dialogue, or because they have "gone away" - to "encourage the others" has been put forward as a reason many times, but I have seen little widespread support for this as an explicit concept, being only a hairsbreadth away from punishment, which we claim not to be dispensers of. Rich Farmbrough, 18:41, 29 April 2011 (UTC).

If they have "gone away"; they should be blocked immediately when they return, if they continue reverting. As an aside, is there a central reposatory (sp?) for "hanging" blocks; users correctly warned that, if they continue what they are doing, they will be blocked, no matter when they return to continue? — Arthur Rubin (talk) 20:30, 13 August 2012 (UTC)

Change to definition of "consecutive edits"

A user (you know who are, but I'm bringing this on your behalf, and we both seem to think the other is harmful to Wikipedia) has been blocked for a 3RR violation when the 1st and 2nd reverts were separated only by a bot edit. I think a reasonable change would be that "reverts" by a user count as consecutive (and hence, as a single revert) if there are no intervening edits to the area and/or topic covered by the the 2nd edit, or possibly by either edit. (The reason for "either edit" is the possibility that the reverting user wants either some offending text removed or the section tagged as WP:NPOV or off-topic, and, if the text is restored by another editor, the reverter might edit war to restore the tag, instead.)

Any suggestions? — Arthur Rubin (talk) 20:54, 13 August 2012 (UTC)

I'm having a little trouble following, but I think such a change would be too broad. At the moment, putting bot edits aside, it doesn't matter whether an editor's reverts are related to a particular area of the article (although there are editors who think all the reverts have to be related). What I think you're suggesting would change that, no? Perhaps for clarity, you could set forth the proposed text of the new definition here for us to look at.--Bbb23 (talk) 21:14, 13 August 2012 (UTC)

I'm having trouble describing the precise text, especially since the blocked party doesn't understand existing policies and guidelines, and my explanations thereof. The idea is that two "reverts" by the same person would be counted as "consecutive" if there are no edits between them related to either edit, instead of there being no edits between them. Perhaps replace:

A series of consecutive saved revert edits by one user with no intervening edits by another user counts as one revert.

by

A series of consecutive saved revert edits by one user with no related intervening edits counts as one revert,

but that may be more capable of misinterpretation. — Arthur Rubin (talk) 01:42, 15 August 2012 (UTC)

In other words, the reverts don't need to be related; but if the intervening edits are unrelated, then the reverts count as one. — Arthur Rubin (talk) 01:43, 15 August 2012 (UTC)

That's what I thought you meant. Thanks for putting it in concrete terms. As I suspected, the replacement language is too broad - or, at least, as you say, "capable of misinterpretation". For example, Editor X makes a change to the article (revert #1). Editor Y reverts X. Editor X reverts Y (Revert #2). Editor Z makes a change to the article that has nothing to do with the material affected by X and Y. Editor X reverts Z. This last edit by X should, in my view, count as Revert #3, but would seem to be exempted by the new language.--Bbb23 (talk) 02:04, 15 August 2012 (UTC)

OK, it needs more work. Edit Z is related to revert 3, though, so 2 and 3 shouldn't be combined. — Arthur Rubin (talk) 06:52, 16 August 2012 (UTC)

How about:

A series of consecutive saved revert edits by one user, with no intervening edits related to any of the reverts, counts as one revert.

In your example, Edit Z is related to Revert #3, (or, to be precise, Revert #3 is related to edit Z), so Reverts 2 and 3 cannot be combined. — Arthur Rubin (talk) 07:05, 16 August 2012 (UTC)

I'd still like to get this change made. It may depend on the definition of "related", but if an edit reverts another, they are certainly "related", so Bbb23's example should still leave all three reverts by X counted separately. Basically, what I'd like to see is that, if editor X is editing one section of an article, making multiple reverts, there are intervening edits, but to different sections of the document, then the "multiple" reverts count as one. — Arthur Rubin (talk) 20:49, 27 August 2012 (UTC)

I wasn't exactly ignoring your response from August 16, so much as I couldn't follow it, and I wasn't sure if it was my "fault" or yours, so I did nothing. At this point, absent input from anyone else, I would be against the change unless I see language that is clear to me and I agree with.--Bbb23 (talk) 00:40, 28 August 2012 (UTC)

Requirement of warning

This page currently states: "A warning is not required" The hidden text here seems to imply that it is required: "Warn the user if you have not already done so" Which is it? I think this page should be rewritten to include something like: "A warning is required, warnings can be given after three reverts or when the behaviour is obvious edit warring" 85.167.111.129 (talk) 13:59, 15 August 2012 (UTC)

Edit war?

Is it Edit war or Edit-war? Should it be capitalised? Jdaniels15 (talk) 15:13, 20 August 2012 (UTC)