Guild Wars Wiki talk:Only revert once

Policy Vote

Yes

No:

1. No ty, this was obviously copied from Guildwiki.--Eloc 15:24, 29 March 2007 (EDT)
2. Strong "No" - 3RR is a far better policy, as outlined in discussions below, particularly, and eloquently, by Rezyk Fox 15:38, 29 March 2007 (EDT)

A vote? Since when do we vote on policy? What's this about, Eloc?--Dirigible 15:41, 29 March 2007 (EDT)

I oppose this vote. =P We're in the middle of some good discussion IMO, let's keep going with that... --Rezyk 16:22, 29 March 2007 (EDT)

I vote against this vote. LordBiro 16:52, 29 March 2007 (EDT)

I vote to abstain from voting about this vote about the vote Fox 19:55, 29 March 2007 (EDT)

I agree - down with the vote (and, for that matter, all votes on the wiki - can we please learn from the failure of voting seen at GuildWiki and not repeat those mistakes?) --- Barek (talk • contribs) - 21:48, 29 March 2007 (EDT)

I'm also against voting. Sockpuppet problem comes to mind. -- ab.er.rant 23:20, 29 March 2007 (EDT)

1. As Eloc said, stolen from guildwiki. -Cheese Slaya

1. Well, after certain considerations, I strongly agree with those against this. Policies like the 3RR are better for the fluent operation of a Wiki. MithranArkanere 19:06, 15 October 2007 (UTC)

Discussion

We don't have Zaishen Reject's permission to relicense, so I'm using an older revision I mostly penned. If you'd like to clarify, please do so. —Tanaric 21:33, 7 February 2007 (PST)

Just for reference, here's Wikipedia's 3RR. I never liked this policy much but clearly revert wars aren't helpful. --Fyren 21:36, 7 February 2007 (PST)

Can we define what a revert is? Are partial reverts considered reverts? Is removing a contribution several edits ago considered a revert? --ab.er.rant (talk) 17:33, 8 February 2007 (PST)

I am against this policy for various reasons. Two of the main ones are: I feel it is very prone to disputes and negative feelings when enforced. I also feel it is too strongly biased against the side who wishes to move away from the status quo. On the other hand, it does theoretically force a hard stop to most edit warring. --Rezyk 13:47, 15 February 2007 (PST)

IMO, revert policy is inherently problematic and difficult. I've posted a draft of an alternative revert policy at Guild Wars Wiki:Three-revert rule (pretty much equivalent to Wikipedia's), which I feel is also wildly imperfect and flawed, but also worthy of consideration. --Rezyk 13:47, 15 February 2007 (PST)

My only comment is that 1RV seemed to work fairly well at heading off conflict over at gwiki. At least it seemed that way to me... perhaps people who were more involved in controversy there had a different opinion. It seems it's strength is that it heads off revert wars quickly and moves things towards an admin before people get too worked up. I don't see any big difference between 1RV and 3RV policy though- constructive people are going to work through their edits, and other people will still need an admin to help out, so whichever we pick will be fine I think. Oblio 08:00, 16 February 2007 (PST)

Both 1RV and 3RR achieve the same goal - but I find 1RV to be much easier to explain to new users, as well as being simpler to monitor. I support porting some of the definitions from 3RR into 1RV, but for overall policy, I greatly prefer 1RV. --- Barek (talk • contribs) - 08:06, 16 February 2007 (PST)

I added a statement that an unopposed concensus of 1 is good enough as long as there's no objection to de-reverting after several days. This is intended to handle de-reversion for low-traffic articles, where the normal method has the potential to be quite slow. It shouldn't affect high-traffic articles. -- Gordon Ecker 21:30, 22 March 2007 (EDT)

Re-start discussion

Can we come back to this discussion? I would like to see something like this policy before we get too big of an influx of users. As said above, I prefer this policy over 3RR; but I would like to hear other views/opinions. --- Barek (talk • contribs) - 10:37, 27 March 2007 (EDT)

I'll discuss some of your points if you discuss some of mine. ;)

You've said that you find 1RV much easier to explain; for me, it's the other way around. For a new user, 3RR is essentially "an editor must not perform more than three reverts, in whole or in part, on a single page within a 24-hour period". What would 1RV be summed up as? (It seems more difficult to concisely capture its no-explanation-exception, corollaries, and whether it applies per editor, but maybe you can do it better than I.) 3RR also seems simpler to monitor to me (although I also question why it'd be important to monitor) -- a violation of 3RR means 4+ edits by one person within a day, which is much easier to notice in recent changes.

There was some mention of 1RV getting an admin involved sooner; I don't see what's so great about that. The only thing a sysop can specially do is block, and exercising authority can often exacerbate a conflict or leave people with negative feelings.

My basic opposing reasons given so far were:

1. I feel it is very prone to disputes and negative feelings when enforced.
2. I feel it is too strongly biased against the side who wishes to move away from the status quo.

To be honest, I consider these kinds of issues to be a much higher priority than how easy it is to explain/monitor. Any agreement/disagreement/thoughts on these? --Rezyk 17:15, 27 March 2007 (EDT)

So far I think we have seen very little vandalism here, but as soon as this wiki is announced properly there will be a huge influx of new users with their own opinions and whom have not read most of the policies of the wiki. As such it seems prudent that a 1RV rule would be much simpler to enforce during the chaotic beginnings which could then be reviewed and perhaps altered to a 3RR at a later stage. --Lemming64 17:22, 27 March 2007 (EDT)

I also prefer 1RV. "What would 1RV be summed up as?" You could sum it up in the same way as your summary, except replace "three" with "one". The idea is if user A makes a change, and user B reverts it, then you should take it to a talk page.

I think the corollary at the bottom of the policy sums it up extremely well; "Do not re-revert, do not re-re-revert, do not re-re-re-revert." Words to live by! :P LordBiro 17:28, 27 March 2007 (EDT)

"An editor must not perform more than one revert, in whole or in part, on a single page within a 24-hour period"? To me, that is very fundamentally different than 1RV, and is something I would have much less issue with (although I still see it as a lot harsher on users than 3RR). Under this, if user A makes a change and user B reverts it, user A is allowed his one revert. --Rezyk 19:00, 27 March 2007 (EDT)

I think that, as Rezyk outlines, 3RR wins hands down over 1RV. It is also more commonly understood by the majority of wiki users. 1RV somehow seems totalitarian and I can see that it it will be abused greatly, particularly by those who want to enforce their opinions. Also, there is sa terrible siege mentality descending, and it often seems on some of the talk pages that people are taking ownership of articles and indeed the wiki itself. It ceases then to be a wiki and instead becomes a website. WPs battlecry to editors of "Be Bold!" should apply here also, and 3RR allows for this while still enabling admins to monitor and "step in" where necessary; 1RV is just padlocking content. This wiki is funded and hosted by ANet, and administered by volunteers, but it "belongs" to all of us - and none of us. 3RR allows for this, 1RV does not Fox 17:36, 27 March 2007 (EDT)

I disagree. To me, 3RR encourages the potential for revert wars. While 1RV, when followed, makes them virtually a non-issue. If user A makes a change, and user B reverts it, user A should take it to the discussion page. Let the community discuss and reach concensus. --- Barek (talk • contribs) - 20:40, 27 March 2007 (EDT)

You asked for a summary of 1RV. It's already in the policy - "If your edit is removed, and you feel it belongs, discuss it on the talk page." This does not prevent change - it encourages discussion to take place. It is no more prone to dispute than 3RR; but as it encourages discussion sooner in the process, it can defuse issues before they become heated.

As for bias - yes, it is biased towards the status quo - I cannot debate that. However, I feel that resolving content disagreements in discussion as soon as possible is the better course. --- Barek (talk • contribs) - 20:40, 27 March 2007 (EDT)

I hate to be nitpicky, but I'm going to have to be in order to figure out which other points to bring up. (I haven't mentioned all my reasons yet) "If your edit is removed, and you feel it belongs, discuss it on the talk page.":

• What if it's not my edit, but somebody else's that was reverted?
• What if I just revert partially? Is that okay?
• What if it was already discussed and there was no consensus either way?

The ideal of trying to resolve disagreements in discussion is generally something I agree with, but I'd say it's deceptively not straightforward which scheme does that better. A movement to discussion is not always going to happen more just because it's the only allowed next step; users will look down the line and weigh whether it's even worth the effort to try discussion. Sometimes the bias against them will make it not worthwhile, and the disagreement is left unresolved. The worst part is: we wouldn't know when it happened. 3RR is not strictly better than 1RV in this regard, but not strictly worse either. --Rezyk 16:56, 28 March 2007 (EDT)

The reason I wrote 1RV instead of a 3RR is because, unlike Wikipedia, many (possibly a majority) of our edits are opinion-based or subjective. The best way to beat a mission, the quickest way to farm for Ectoplasm, etc... these are subjective questions that require a good deal of expertise in an aspect of the game. On Wikipedia, realizing expertise is usually pretty easy -- "I've got a PhD in theoretical physics and can site a source backing up my edit, so I'm right." Unfortunately, things don't work that way here for at least three reasons.

1. People who play video games generally assume that the way they do something is the best way until shown incontrovertibly otherwise. If the guide here differs with their experience, they will generally edit to reflect what they do, even if their way is worse.
2. There exist no neutral sources of information to verify edits here. You can't, in general, cite a book that says your edit is correct. This is why we have no analogue to Wikipedia's "no original research" policy -- the wiki here is essentially all original research.
3. People who play Guild Wars are, generally, somewhere between 14-19. People who edit the Wikipedia are significantly older and more mature.

While I think 1RV would fail on Wikipedia (or, at least, fail more than 3RR would), I think it's appropriate for us. It allows bad advice to be quickly removed and it allows revert wars based on adding that bad advice back in to be quelled via sysop-intervention -- usually an incredibly short-term block -- immediately upon its re-addition. In practice, this has resulted in extremely quick conflict resolution while maintaining high-quality articles.

—Tanaric 12:49, 28 March 2007 (EDT)

Just in case, I'd like to point out that I have never pushed "Wikipedia does this" as a reason we should do 3RR; I always, always, always have had a much deeper reason than that behind my support/opposition for anything. Just saying, because this tends to come up sometimes. ;)

In the case of "blatantly bad advice removal" (and where the original author does not become upset/disillusioned by the bias in effect), I agree that 1RV handles it better/easier. I also agree that it's more of a problem here than, say, Wikipedia, for some of the reasons you've mentioned (although I don't think your characterization of Wikipedia would fly over there =). I see this as one of the definite advantages of 1RV over 3RR.

A problem, though, is that 1RV's scope is not nearly limited to this. It applies to almost ANY content dispute. It could be good advice, a question of getting facts straight, differences in style/layout preferences, and can even block the removal of bad advice. And in practice, it has been used on all of those we haven't tried it here yet so we can't know for sure what will happen. The inherent-status-quo-bias of 1RV can have a profoundly negative impact for those cases, and that's what bothers me the most. To be fair, 3RR also carries a different bias but its effect is much more mitigated (I can go into why if anyone wants). I also wonder if the full ramifications of 3RR are really understood well here -- for example, it also allows bad advice to be removed quickly. --Rezyk 16:56, 28 March 2007 (EDT)

Generally most admins are pretty active (*cough!*), the speedy removal of bad advise isn't normally a problem. 3RR in my mind just encourages edit wars while 1RV more strongly encourages users to discuss the issue at hand, it brings things to a head quicker. Both policies are designed primarily to stop edit wars, to me 1RV does this better. Perhaps the proof's in the pudding, there's a backlash against 3RR at Wikipedia whereas 1RV seems to work well on the GuildWiki. --Xasxas256 18:41, 28 March 2007 (EDT)

A 2 year old essay at WP hardly constitutes a "backlash"! 3RR remains their standard. You have to remember also that many people do not go to the talk page of an article - if they consult an article, they will invariably assume that the content is accurate and up-to-date, perhaps even more so as this is the "official" wiki. If due to 1RV the article has been left containing inaccurate information pending discussion, why would the casual reader think to look at the talk page? Fox 22:18, 28 March 2007 (EDT)

I agree, I followed that link and even though I missed it being two years old, I didn't see anything resembling a backlash. But, on your other argument, I would use it as reasoning to not use 3RR. Under 3RR, the article could be repeatedly reverted - 3RR reached, next day it starts again - a casual user wouldn't know to check the history tab to see that a series of reverts exist between two versions that could be giving different or even conflicting advice or strategies - which version he sees becomes luck of the draw. This does not happen under 1RV. I've never seen an article wait long for a discussion to resolve an issue in GuildWiki - so I fail to see why it would be any different here. The sooner discussion starts, the sooner concensus can be reached, and the sooner the article can be agreed to have or updated to have the best data available at the time. --- Barek (talk • contribs) - 23:45, 28 March 2007 (EDT)

I'd love to hear why 3RR's bias is more mitigated than 1RV's. I'm genuinely interested in putting together the best policy for reversions on this wiki, regardless of which one it might be. Just as you aren't pushing 3RR because Wikipedia does it, I'm not pushing 1RV because GuildWiki does it. :) —Tanaric 22:58, 28 March 2007 (EDT)

I would also like to here more of the reasoning on this. I understand your concerns, but I don't see the majority of the issues brought up thus far being significantly worse under either proposal. --- Barek (talk • contribs) - 23:45, 28 March 2007 (EDT)

I'm not a wikipedian, so I too am quite interested! The thing that has always seemed strange to me about 3RR is that if, for example, 3 users wanted to have their own text on an article, User A might alter the article, User B might alter it again, User A reverts, User B reverts, User A reverts, User B reverts, User A reverts, User B reverts (they have used up all of their reverts for this article), User C comes along and adds what he wants. Can User A and User B revert that? I don't know, I haven't used 3RR enough to know this, but it just seems odd. LordBiro 05:54, 29 March 2007 (EDT)

(1)"An editor must not perform more than three reverts, in whole or in part, on a single page within a 24-hour period. A revert means undoing the actions of another editor, whether involving the same or different material each time." In the scenario outlined above, User A and User B have effectively edit-duelled and their editing privileges for that article are now curbed for 24 hours. Benefit - introspection and detachment from the subject often serve to cool tempers. (2)"An editor does not have to perform the same revert on a page more than three times to breach this rule; all reverts made by an editor on a particular page within a 24 hour period are counted." No, they cannot revert User C's edit. This may be beneficial because User C may be User:Mr.Objective, stepping in and attempting to curb egregious edits.(3)"Nobody ever died in an edit war." Fox 06:20, 29 March 2007 (EDT)

Yes, but equally User C may be User:Mr.Incorrect. Perhaps the chance of another objective user coming and sorting it all out is high, but with 1RV the idea is that it is not the number of reverts per user, but per article. An "edit-duel" could not take place with 1RV. LordBiro 09:24, 29 March 2007 (EDT)

(reset indent)

I consider the baseline (i.e. unbiased) to be roughly "whatever consensus would be if we had a nice, thorough discussion between all interested parties". The further a policy can keep things different than this, the more biased it is.

Basic situation: An article starts with some text in it: blah blah original blah. User:A edits the "original" part to say "revised". User:B reverts it back to "original". No vandalism is involved, as both editors are genuinely trying to improve the article. With 1RV, the onus is now completely on User:A to get a consensus if he still wants revised. If consensus is for original, or there's no discussion and consensus, we stick with original. I guess we're all in agreement that this is a strong bias towards original.

I'll first discuss a hypothetical reduced version of 3RR to keep things simple (1RR): "An editor must not perform more than one revert, in whole or in part, on a single page within a 24-hour period". After the basic situation, under 1RR, User:A can then revert to revised, and User:B is stopped (for 24 hours). However, as you guys have brought up, XRR is different than 1RV in being "per user" and User:C can come in and revert to original. The key is not whether User:C or original is "correct" or "incorrect", but that it reflects the opinion of 2 users over 1. Maybe User:B will then gather some users who agree with him, and get it reverted to revised. What we really end up with is an initial bias toward the majority opinion. Majority opinion isn't the same as consensus opinion (we're not a democracy, etc), but they match quite often and so I consider this bias to be pretty small already.

Also, 1RR is different than 1RV in that it is "per 24 hours". This changes the equation in a few ways: the bias is kept centered on current opinions rather than first day opinions, and it shifts things toward "who cares more strongly about the issue". (Maybe tomorrow, those whose opinion was changed by discussion or wasn't that strong to begin with, are no longer willing to revert.) This is not perfect, but not so far from consensus either (stronger opinions counting for more than weak opinions). It also doesn't allow as much good reason to be upset over "losing a revert-duel". (Under 1RV, one could easily say "the only reason I lost was because the other guy was here first". What could you say under 1RR? "I gave up because the majority of everyone with strong enough opinions was against me"?)

Switching from 1RR to 3RR, the majority and current-opinion effects still hold, and more reverts early on increases the "who cares more strongly about the issue" factor. I'm not sure if I consider this increase good or bad. (Personally I prefer 3RR over 1RR for other reasons)

When it comes down to it, all our proposals would eventually hit that unbiased state if/when discussion reached a consensus. Here, I am just really concerned with what can happen before that point, as that can profoundly impact the will to contribute or try for that consensus. Even though revert-duels would not actually go as far as I've outlined in practice, it's important that the potential is still there so that everyone is driven toward building consensus. Think about how one would "game" XRR (because ultimately, as much as we want to discourage it, we cannot completely prevent gaming the system) -- the way to win the theoretical "revert-duel" is get more users to strongly agree with you and add their revert-power to your side. At that point, you're really just building a consensus and might as well do it in discussion! With 1RV, you also get that on one side, but it's much much easier to game or luck into the status-quo side and simply defend against incursions and seek no consensus.

The main concern with 3RR so far is more potential for edit-warring than 1RV. This is also a strong concern of mine (wasn't I the first to point out 1RV's advantage here? =). But completely stopping edit wars is not the highest priority of the wiki, and not worth the bias that 1RV puts on the table. Edit wars are as much a symptom of disagreements as they are a cause, and we should avoid cutting too close to the bone, lest we suppress the symptom without having addressed the root cause. --Rezyk 15:13, 29 March 2007 (EDT)

I think the reason why I'd still be more comfortable with 1RV rather than 3RR comes from a disagreement with this statement you make in your last paragraph, "But completely stopping edit wars is not the highest priority of the wiki, and not worth the bias that 1RV puts on the table". For this wiki I think the priority actually should be the other way around. Here's why I think that's so:

• First, demographics. As Tanaric said above, the GWW and Guild Wiki average editor is much younger in age than the average editor at Wikipedia. Revert wars seem to me a much more serious issue than you're giving credit them for. Even on GuildWiki with 1RV, there were still people trying to weasle through another edit to put them ahead of their "opponent". It too often becomes about getting one's version of an article to remain, rather than the best version regardless of who proposed it. As you said, Rezyk, 1RV has an advantage over 3RR as far as revert wars are concerned.
• Second, I myself would be quite comfortable with a bias towards the original on the GWW. From my experience at GuildWiki, when someone gets his "revised version" of an article reverted, 9 out of 10 times it's done so with a very good reason, (either adding something completely subjective, inane or plain out nonsense). Wikipedia has reason to avoid a bias towards the original because it also has to deal with many issues that we don't have to face. Concepts like neutral point of view and notability would be alien to this wiki, and are the major ones that are affected by that bias towards the original. This wiki is merely documenting a game, there's no issues such as needing to keep a Ranger article politically correct so that it doesn't offend Monks and Warriors. Skill articles, creatures, locations, guild pages, templates, policy and formatting guidelines, etc etc, these are all elements that greatly benefit by that bias towards the original, as far as reverts go. The only potential areas of trouble would be the guides (as builds were on GuildWiki), and I hope that the policy that will govern those pages will be strict enough to be able to deal with subjective and dubious edits without the need to fall back on 1RV/3RR.

Because I see bias towards the original as a good thing on this wiki, and revert wars as highly likely and very unproductive, I personally would have to favour 1RV over 3RR. --Dirigible 17:10, 29 March 2007 (EDT)

I still think we have to concern ourselves with neutral point of views, particularly on subjective comments regarding NPCs, skills, locations, armor, etc. But those are minor stuff compared to Wikipedia. As for 3RR, 1RR, or 1RV, personally, I'd stick with 1RV for its simplicity, and mainly because the majority of our articles would hardly fall into edit wars, given the factual and brief nature of our articles. -- ab.er.rant 23:20, 29 March 2007 (EDT)

Tidying up this Policy

Personally, I favor this policy over the 3RR policy, however I feel that it needs to be tidied up before becoming accepted policy. In particular, I'd like to see the 'What's a revert?', 'Exceptions' and 'Enforcement' sections from 3RR be integrated in some form as I feel they are helpful in explaining situations clearly. I would also like the reference to assume good faith to be removed, as that policy has been rejected. I would also like to see the corollary be presented more directly, in that it may be easier for new users to understand than the paragraphs above. The only other thought I have on this policy is to emphasize more clearly the idea of 'one revert, then take it to the discussion page', although I am unsure how to accomplish this. --Indecision 03:46, 4 April 2007 (EDT)

I did a mayor rewrite, introducing parts of 3RR (but not fully). Main goal was to shorten the policy from "essay format" to "easy to read" format to make it more acessable to users. --Xeeron 04:49, 20 April 2007 (EDT)

I still think that the "unopposed concensus of one" exception is necessery for low-traffic pages, but otherwise the policy is fine. -- Gordon Ecker 05:07, 20 April 2007 (EDT)

I'm mostly alright with the policy, with the exception of this particular sentence here, which I find somewhat confusing and out of place, "Note that in the case of vandalism, blocking editors who have engaged in vandalism will often be better than reverting." Does this need to be here? First, I think I disagree with the message (I'd rather just revert the vandalism, reserve blockings only for when really needed, i.e. repeated vandalism), and I think that it shouldn't even be part of this policy, leave it to Guild Wars Wiki:Blocking policy to determine whether immediately blocking a vandal is a good idea or not. Secondly, I'm not sure what exactly it's suggesting in the context of this policy... it makes it sound as if reverting a vandal edit is an inferior and less-than-ideal approach to take. *confused*.

As for the low-traffic pages, Gordon, a request for comments can be made, so it should be ok, I think. --Dirigible 05:34, 20 April 2007 (EDT)

Copied that part over from 3RR, but you are right: Blocking of vandals is better dealt with at Guild Wars Wiki:Blocking policy instead of here. Removed the sentence. --Xeeron 09:24, 20 April 2007 (EDT)

Questions

Would reverting be allowed once discussion is underway, even if a consensus hasn't been reached? --Rezyk 01:39, 21 April 2007 (EDT)

I would say yes, as long as the preliminary discussion is in favor of reversion. -- Gordon Ecker 01:52, 21 April 2007 (EDT)

Would any removal of text be considered a (possibly partial) revert, or only if none of the removed text was part of the very first article edit/creation? --Rezyk 03:39, 28 April 2007 (EDT)

Suppose a delete tag is added to an article, but this is reverted. From that point on, would all editors (besides the reverter) have to go to discussion before marking the article for deletion? --Rezyk 03:39, 28 April 2007 (EDT)

I'd prefer to add something like a 1 week time limit to avoid questions like this. I don't think a rule with a penalty for violation should have this kind of vagueness. -- Gordon Ecker 04:39, 28 April 2007 (EDT)

I think that removing a delete tag is not allowed as the delete notice should not be removed without discussion as stated on the template itself. Not sure how this policy should work with that situation though. -- (gem / talk) 08:44, 10 May 2007 (EDT)

Historically, delete tags on the GuildWiki had a dual purpose. One could place one on an article to indicate that one thought the article should be deleted. Anybody who disagreed could remove the article and take the discussion to talk. Thus, if the delete tag remained on the article for a few days, it implied consensus for the deletion. Simply speaking, I disagree with Gem – delete tags ought to be subject to 1RV just like everything else... even if 1RV isn't selected as policy. —Tanaric 11:52, 10 May 2007 (EDT)

Revive

I'm thinking that either 1RV or 3RR be made official, depending on which is more complete. We can continue to argue which is better later on. I think something should be in place for the influx of new users. -- ab.er.rant 02:58, 25 May 2007 (UTC)

It looks to me that 3RR is more complete, and 1RV has more favour. I'd run with the one that more people like - and just officialise this one - then worry about the decision afterwards. Ale_Jrb (talk) 07:43, 25 May 2007 (UTC)

Agreed -- but I'm biased. —Tanaric 18:49, 25 May 2007 (UTC)

I am one of those favoring 1RV, but tell us what needs work and maybe we can update the article right away. --Xeeron 21:04, 25 May 2007 (UTC)

Not that I favor 1RV, but please, please, at the very least try to figure out how to add some sort of hard time limit to it (see Gorden Ecker's comment in the Questions section above) to mitigate the locking-your-opinion-out effect. --Rezyk 21:39, 25 May 2007 (UTC)

And before you ask me to try adding it myself...I would, but my attempt opens up a hole that I feel obliged to patch up, which opens another hole to patch, etc, and ultimately I always end up converting it into a 1RR variant. =P --Rezyk 22:20, 25 May 2007 (UTC)

While I personally prefer 3RR to this, I think that either will do while the Wiki adapts to the pressure of being the 'in-game' help file. Afterwards, if the consensus is that the wrong one was chosen now, it can easily be changed. As it stands, it seems to me to be complete enough to accept it. I say just go for it :). Ale_Jrb (talk) 23:41, 25 May 2007 (UTC)

Applying either one as they are now is a pretty solid commitment, as it will likely be difficult to reach any consensus for change later. I propose ratifying one of them with a built-in expiration date ("This policy only applies until September 2007"). And I would prefer seeing 3RR for this because (1) I'm biased towards that one and (2) it seems to have less strong objections against it (although also less popular). --Rezyk 16:06, 26 May 2007 (UTC)

I would prefer seeing 1RV for this purpose, as our sysop and bureaucrat teams in general are more familiar with it, and it has more support. I'd also argue that this policy only has one strong objector. While your many objections are all quite strong, nobody else has taken up the banner to support or defend those objections, while plenty of 1RV supporters have rallied to their cause. That said, I'm quite fond of 1RV and I have watched it do quite well on the GuildWiki since I wrote it, so perhaps my argument is worthless. —Tanaric 01:16, 9 June 2007 (UTC)

Fox is clearly another strong objector, and multiple other users have expressed a dislike of 1RV. I admit to having a strong distaste of it myself after giving up on objecting to it on GuildWiki and watching it continue to perform predictably, horribly bad there. --Rezyk 17:03, 9 June 2007 (UTC)

I don't understand why you feel 1RV has done poorly on the GuildWiki. It's my opinion (and appears to be the opinion of others above as well) that is has done its job well. —Tanaric 17:51, 9 June 2007 (UTC)

I'm not sure that it's worth going into. The common sentiment seems to be that examining other wikis is incomparable in principle (at least, depending on who does it). --Rezyk 18:40, 9 June 2007 (UTC)

Frankly, if we really want to get revert policy moving, then I suggest that it would also be fair to mark this policy as rejected (others have been rejected with much less objection). Then we should start discussing the pros/cons of 3RR -- but generally compared to our current state to see if it's an acceptable policy in itself, not compared to 1RV. (Maybe it will get rejected too, and then we should see if we can come up with something better than both, or perhaps revive 1RV as a proposal.)

If we keep approaching this from the standpoint of "we decided we want to restrict reverts, now pick one of the policies that does it", then we'll naturally never get a strong consensus because people have good reasons to have differing preferences. It's similar to saying "we decided we want to allow subjective builds, now pick one of the policies that does it" -- opinion will naturally be heavily divided among favorites, rather than letting each proposal survive or fall on its own merits. --Rezyk 17:54, 9 June 2007 (UTC)

If you mark this one as rejected, then I feel that 3RR should be as well. There are at least as many comments above rejecting Guild Wars Wiki:Three-revert rule as there are for 1RV - if not more rejections. In my opinion, the only reason the talk page for 3RR is so short is because the comments were taking place here - I think the two are irreversibly linked in the discussions. --- Barek (talk • contribs) - 18:52, 9 June 2007 (UTC)

Perhaps I have missed it but I also fail to understand the problems that have arisen on GuildWiki thanks to 1RV. Regardless of the decision made on this wiki regarding reverts I would be interested to know more. Rezyk, you say that discussing the failures of 1RV at GuildWiki are not productive here. In that case could you please elaborate on my user talk page or another location of your preference? LordBiro 21:22, 9 June 2007 (UTC)

What about a vote or opinion poll with options like this?

• In favor of 1RV.
• In favor of 3RV.
• In favor of either.
• Opposed to both.

Or something more complicated like this?

• In favor of 1RV, opposed to 3RV.
• 1RV preferred, 3RV acceptable.
• 1RV and 3RV are both acceptable, no preference.
• 3RV preferred, 1RV acceptable.
• In favor of 3RV, opposed to 1RV.
• Opposed to both 1RV and 3RV.

The way I see it, the problem is that we have two good, popular policy proposals which are mutually exclusive. -- Gordon Ecker 02:13, 10 June 2007 (UTC)

Votes = bad. Especially on policy decisions.

For what it's worth, I'm against 1RV, and I would be in favor of either a 1RR or 3RR. MisterPepe talk 02:34, 10 June 2007 (UTC)

Agreed with MisterPepe - a wiki is not a democracy, contrary to sometimes-popular belief, and policy decisions are some of the worst to leave to votes. Better no policy either way while the two are still in contention.

I'm actually starting to wonder if a modified version of 1RV might not be better - namely, each editor should only revert a change once without prior discussion - i.e. Editor 1 edits a page, Editor 2 reverts it: Editor 1 shouldn't revert that revert without discussing with Editor 2, but if Editor 3 agrees with Editor 1, they should be able to revert Editor 2's revert, and then if Editor 2 thinks both 1 and 3 are wrong, they can take it up in discussion, et cetera. Essentially, it provides, to some extent, a fallback for both bad edits and bad reverts, while at the same time at least somewhat stifling revert wars. Potential downside would be 'cliques' of editors banding together to multiple-revert a page to make a change go through, but there's only so large that can get before you'd effectively have a consensus anyways - and I'd think that'd be much less likely than any individual disagreements. (Aiiane - talk - contribs) 02:45, 10 June 2007 (UTC)

Aiiane: That's pretty much a 1RR. MisterPepe talk 02:48, 10 June 2007 (UTC)

"Pretty much" is not the same as equality. I don't support the 24-hour period concept, as I think it's both artificial and doesn't serve the overall purpose well. I also don't support only 1 revert per page, I support 1 revert per edit. If someone adds one bad element to a page, and it's reverted, the editor that did the revert should not be prevented from reverting another bad element added to the page. Hence my elaboration, rather than referring to a rather flimsily-defined acronym. (Aiiane - talk - contribs) 02:55, 10 June 2007 (UTC)

I suggested a vote or opinion poll because the discussion isn't going anywhere. If we only had one proposal, or we had one obviously good proposal and one obviously bad proposal, we would have already resolved the discussion and had a policy, but the system doesn't seem to work for two good, mutually exclusive proposals. -- Gordon Ecker 06:14, 10 June 2007 (UTC)

LordBiro, you didn't miss anything; I generally prefer to refrain from bringing up contentious GuildWiki history if it's not warranted. Why not ask those who have voiced their disfavor on GuildWiki and are still there? --Rezyk 18:01, 11 June 2007 (UTC)

(RI) I like Aiiane's suggestion a lot. Basically, because it addresses all features I hate about 3RR: Why 24 hours? Totally arbitrary and leads to protracted revert wars that go on for months. Why 3? I fail to see in what way the allowed 3rd revert could be any better than the forbidden 4th one. And relaxing 1RV to allow one revert per editor at least stops one single editor from locking the article in a very bad state for a period of time if most editors feel otherwise. --Xeeron 10:11, 10 June 2007 (UTC)

1RV already prevents this. Two article editors in talk who agreed a revert was harmful have the right to re-revert under 1RV. Perhaps this isn't as clear as it could be. Since 3RR isn't going where and isn't particularly supported, could somebody draft 1RR and we can compare side by side? —Tanaric 16:05, 10 June 2007 (UTC)

1RR still has the issue of the 24-hour timer, which is something I strongly oppose, for reasons already mentioned. (Aiiane - talk - contribs) 16:17, 10 June 2007 (UTC)

Barek, my general point was: All the reasons given here so far against 3RR don't really oppose it, but instead just explain why that user wants 1RV (although yes, the two are inexorably tied together politically). I would be interested in finding out if anyone has anything against 3RR outside of that. (A straw poll makes sense to me to gauge consensus, as long as it's not considered a vote.) If not, we could easily and immediately have a policy that limits the most turbulent revert wars, and that nobody objects to having over our current situation (although it wouldn't go as far in restrictiveness as many want). I wonder if some of you see 3RR's limitations as too arbitrary/soft and would simply rather wait until reverts become an issue before making it one. Is that it?

I'm not against 1RV because I want 3RR. I'm against it because its strong bias (in my view) greatly threatens the ideals of consensus, equality, and bold editing whereas our other policies are relatively decent in that respect. And it appears that I'm not the only one who feels this way. If we find/fix some policy to not break these, then I won't have such strong objections -- it doesn't have to be 3RR.

Xeeron, the third revert is a rough balance between being draconian and permissive. It's not fundamentally different from the fourth revert, but mainly is more oriented toward tolerance for new users (so that they get fair warning more often) and the freedom to make some reverts in any productive back-and-forth-revert process. Regarding protracted revert wars, the 24 hours seems strictly better than the 0 second equivalent we have now. For the arbitrariness of the exact numbers 3 and 24, I don't understand the problem. Would you oppose election policy for arbitrarily picking periods of 7 days instead of say, 8?

--Rezyk 18:01, 11 June 2007 (UTC)

What I dislike about 3RR is the numbers. Now keep in mind I am quite in favor of less restrictive policies, but with 3RR a group of 4 users (2 argueing each side), not such a great number, could add up to 12 reverts on one article per day. After a week, they'd be at 84(!) reverts, without ever having broken 3RR. To me that is plain silly. At the latest when the first re-revert happens, it should be clear to everyone that there is some serious disagreement, which should be taken to the talk page right away, instead of further reverts. --Xeeron 20:18, 11 June 2007 (UTC)

I dislike the 24-hour timer not because of any arbitrary dislike for the number of hours, but because the fact that such a set timer exists at all. Why would a revert that was contentious one day be allowed to go forward the next, even if no consensus has been found? Under 3RR, two editors who happened to disagree over a subject and were unable to reach consensus could carry on a protracted revert war for days, weeks, even months without ever breaking the policy, as long as they just logged in once a day to revert the page to their favored version. Yes, at that point it'd probably be time for arbitration, but there's nothing in the policy that forces either editor to pursue that option, if both decide to adopt a passive-aggressive attitude. Under 1RV and 1RR, there is at least a point at which one editor is forced to move for arbitration to continue to have a say in the matter. (Aiiane - talk - contribs) 20:25, 11 June 2007 (UTC)

If the 84 reverts or month long revert war situations are possible under 3RR, then they're allowed right now -- 3RR does not convey an entitlement to revert three times each day. It just provides one "electric fence" that deals with even sillier cases like someone making the same revert 100 times in a week -- other policies can still impose more restrictions. Is there anything particularly bad about 3RR compared to no revert policy?

Now, switching gears and talking about a completely different proposal instead: =P 1RV generally prevents these situations automatically (rather than relying on a drive towards consensus which can be initiated by anyone, to deal with them); I just wish it didn't carry so many other issues with it. One thing I particularly hate is how it encourages one side (who was here first, or gamed it right, or is simply willing to revert more) to adopt a passive-aggressive attitude without compromise, by putting the entire burden on the other party. Suppose someone makes an edit that is disliked. With 1RV, it's so much convenient and tempting to just revert it without a second thought, rather than say, seek a middle ground solution that would appease both parties. --Rezyk 21:14, 21 June 2007 (UTC)

Obviously, what we have "right now" isn't an ideal solution, either - so not sure why you're referencing that in your first paragraph. The point is that I don't like timers in that regard, because they essentially create more problems than they solve.

As for 1RV, I dislike it for the same reason you do, hence why I prefer the 1RR version Xeeron wrote up. (Aiiane - talk - contribs) 02:44, 22 June 2007 (UTC)

1RR

Wrote something up based on how I interpret Aiiane's comments above. Note that this does differ from simply replacing 3 with a 1 in 3RR: Guild Wars Wiki:One-revert rule --Xeeron 16:53, 10 June 2007 (UTC)

Awesome, that's exactly what I intended with my comment. Glad we're on the same page. I'll look it over tonight when I have more time. —Tanaric 18:47, 10 June 2007 (UTC)

Continuing to move onward

Has anyone taken a detailed look at Guild Wars Wiki:One-revert rule yet and can share a relative opinion? (Aiiane - talk - contribs) 11:44, 19 June 2007 (UTC)

Put it on GWW:RFC. --Jamie 11:45, 19 June 2007 (UTC)

I will when I get a chance, but tbh the real intent of adding this section was to ping the people who already had it on watchlist. (Aiiane - talk - contribs) 03:27, 20 June 2007 (UTC)

Sorry for the long delay in giving my critique of it. (I should also actually respond to your comment and Xeeron's above.) I'll try to get to both soon. --Rezyk 03:44, 20 June 2007 (UTC)

I'd like to throw my $.02 in on this. 1RV can be abused in a very straightforward way, say someone makes an article change that I disapprove of (I find it inaccurate for example). I revert. In this case I have the final say until the matter is settled on the talk page. That's fine if I'm reasonable, but if I'm unreasonable and eloquent (or I have an agenda to push on some particular page) then I can abuse 1RV to keep my favored material visible during any dispute. So to me 3RR seems like a less abusable policy. --Tankity Tank 02:21, 6 August 2007 (UTC)

1RV gives each editor one revert, so if you reverted something and I disagreed I could revert it to before your revert. Then if you still disagreed it would be taken to the talk page. - BeX 02:39, 6 August 2007 (UTC)

That's incorrect, Bex. You're thinking of 1RR. 1RV is one revert, by anyone, period. (Aiiane - talk - contribs) 03:42, 6 August 2007 (UTC)

Sorry I assumed he was talking about the one you were linking to. Otherwise this deserves a new section. - BeX 06:36, 6 August 2007 (UTC)

Users must remember that this is not Guild Wiki. In some cases, a mistake of one user may bring slowness in edition. Let's say that, for example, one user adds info about pet evolutions, then another that thinks that different pets have different traits (more HP and so) adds that false info. Now, the page should be frozen and wait for both users to agree, or should the first users remove the false data as fast as possible? This case happened to me, with the Critical Agility. It was not stated as an Enchantment I edited it to denote it's an enchantment, and then another user reverted it, based on the info found in Guild Wiki. What should users do then. Wait one day? Ask a friend to fix the undoing? Nope. When there is absolutely no reasonable doubt about a piece of information (that is, when something is bolean, yes/no, true/false, to be/not to be) then the mistaken users should be warned with a message in his or the article talk pages explaining why were they wrong, and the page should be fixed as soon as possible to keep the information true.

What I mean is, in the same way that users should not revert when there is doubt, users MUST revert without limitations when there is ABSOLUTELY no doubt. A piece of information that IS wrong is much worst than a piece of information that MAY be wrong. Of course, users must prove beyond any reasonable doubt they are correct to bypass revert limitations, and users that cannot prove they are right should refrain form reverting. When it comes to the 'shape' of the information (formatting, colors, appearance, expressions, presentation) policies should stay strict and enforce players 'talking' before editing. But when it comes to raw data (creature families proved by Andrew McLeod, in-game skill descriptions, armor costs, etc) veracity it's the only thing that matters. MithranArkanere 06:07, 13 October 2007 (UTC)

A revert war is worse than either correct or incorrect information, hence the reason for a hard-limit "electric fence" policy. (Aiiane - talk - contribs) 07:39, 13 October 2007 (UTC)

That's why a 'stop reverting' message must be sent to the wrong info editor. It's not, just 'revert and revert'. It's 'if the info is wrong, revert without a talk'. If your information is completely certain (like an ingame text you are actually reading in front of you, with the game running versus a text made up or copied from elsewhere out of the game), the opinion of the 'wrong reverted' doesn't mater. Just a message like 'Please stop reverting current information. Open the game and read the description. If the wrong reverted keeps reverting, the fault would be his, not of the currect reverter, that just keep true information; you could say that a warned wrong reverter that keeps editing an article t show wrong information it's making vandalism. But this would be only in those cases were there is no place for doubts, like I previously said. There are many things that have place for doubts (plot, lore, unconfirmed family, army and world of creatures, trivia, creature descriptions, ...) but If a skill is an hex, it's an hex. If someone keeps editing it to be an enchantment another user should not wait days to revert the wrong information. MithranArkanere 15:38, 13 October 2007 (UTC)

The problem is that if someone is willing to add in an edit multiple times, there's a decent chance that they believe it is correct - and they may or may not "see the light" in such "evidence". It's better to simply stop the war outright and sort everything out, rather than to continue to revert something over and over just because someone feels like adding it. It's far too difficult to draw the line exactly where "incontrovertible" stops and the gray area begins, without making it useless anyways.

That said, I dislike 1RV as well, because it does stall out reverts too quickly for my taste - hence my support of 1RR (linked above and from GWW:POLICY. (Aiiane - talk - contribs) 04:00, 14 October 2007 (UTC)

The problem isn't two editors that are doubtful. Those needs no policy, they'd talk about it anyway. Reverwars happen just because two editors are convinced that they are obviously right. Any section OKing reverts when you are sure will cause just what we want to avoid. Backsword 08:41, 14 October 2007 (UTC)

Yeah, but I'm not saying that revers should contine and contine, I'm saying that if user A reverts something that is correct, and user B can probe it's not, (like a screen capture of the description of a skill in English language) then, instead of opening the talk, and waiting for consensum, B can fix the wrong information, and start the talk, without waiting for A response. Then user A must check the information if he pretends to re-revert. If he thinks he can probe B is wrong, then we have doubts! Both can't be right! So A would not be able to revert unless they find out who's right. For example. Article N says a skill deals 10...20 damage. Then an update changes it to 15..25. B edits it to the new properties. A check Guild Gars Guru skill descriptions an revert it back to 10..20. So, now, B, reverts the artcile again, and send a message to A to check the info. Now, A must go to the game and check the info, if he goes back to Guild Wars Guru [skill] tags to check the data, that won't do. He would have to go ingame, find a monster that uses the skill (or ask a friend and go skirmish), and screen capture the x..y damage from the skill description. If he do that and find incongruencies, then no more revrts would be allowed without a talk.

Long story short, if you have proof (namely a screen capture), you can revert without an answer, bt you have to warn the other user and send him proof. If you don't have or something is wrong (two captures with different data), you have to wait for answers in the talk before reverting anything else. That's what I mean. MithranArkanere 18:51, 14 October 2007 (UTC)

It's just not that simple, Mithran. It's a lot simpler to just stop the revert war and sort things out, regardless of what's on the page - it will work out in the end, and in the meantime a note can be added that something is contested (we have a template or two for that sitting around somewhere). It's better not to give someone a potential reason to continue a revert war, regardless of the proof. (Aiiane - talk - contribs) 20:28, 14 October 2007 (UTC)

No, no. It's not 'continue a revert war', it's allow a second revert. This policy talks about ONE revert. The third would require the talk. That is. The exception I talk about is two allow a second one for a quick fix. MithranArkanere 13:35, 15 October 2007 (UTC)

"Allowing a second revert" is exactly what this policy does not. That exception would make no sense, since the reverters probably both think they are correct. This opens up for "justified" revert wars. It is, as Aiiane says, a lot better to not revert, even if you know you are right, and discuss it. Some inaccurate info (with a disputed tag?) for a short while does not harm anyone. Either we have one revert, or we have two reverts. Not one revert with a possible second if you know you are right. - anja 15:00, 15 October 2007 (UTC)

The problem I have with this proposal is that it's far too strict for its' intended purpose of preventing revert wars. I think it's unreasonable to expect every editor to sift through an article's revision history to make sure that they aren't accidentally undoing an unlabelled revert of someone else's edit. I think that users should be able to undo unlabelled reverts, as well as reverts more than 50 edits back. -- Gordon Ecker 09:35, 13 October 2007 (UTC)