User:Heimstern/Editwarring

Caveat lector: This page is likely a tad out of date.

On administrative action against edit warring
Edit warring can be a complex issue. It is clear that it is destructive to Wikipedia, and therefore it is essential that administrators use their tools to prevent it. But it is often difficult to determine how to do this.

Our options
There are two administrative tools we can use against edit wars: blocking and full protection. Alternatively, we can defer use of the tools and try simply warning the user(s) in question. When should the various options be used? What factors should we take into account?

Full protection
Fully protecting an article is a surefire way to end the edit warring there for however long the article is protected; however, this solution runs counter the basic tenet of Wikipedia of being open to editing by all. Once an article is protected, no one but admins can edit it, and even they are only supposed to edit it under limited circumstances. Thus, it affects the good editors who want to improve the articles just as much it affects the edit warriors. For that reason, this solution should be used sparingly. Before protecting an article, ask yourself: "Could this edit war be prevented effectively with a block or two?" If the answer is yes, you should probably not protect the article. In general, protection is most appropriate for an edit war with multiple users on each "side". By contrast, in an edit war between only two users, blocks can stop the edit war effectively, leaving non-edit-warring users free to edit the article. In general, an article should only be protected to stop an edit war once for each dispute. If the same users edit war over it again, it would seem that the first protection was ineffective in getting users to resolve their dispute and there is little reason to think a second protection will be more effective.

Blocking
Blocking is an unfortunate solution to the problem, but is often necessary. Remember that edit warring is fundamentally at odds with the idea of Wikipedia, and thus it's necessary to use what tools we need to use to stop it, including blocks. Blocking is preferable to protection in cases where the edit war does not consist of several users on each side, as it locks only the offending editors out of the article, while leaving it free for users who are not edit warring to edit.

The three-revert rule is the most common yardstick used to measure edit warring, and much of the time, it's a good one. But we need to guard against mechanically enforcing the letter of this rule. One example of a mistake we can make is blocking one editor who reverts four times while his or her "opponent" reverts only three times. Remember: the prohibited behaviour is edit warring, and it does not help our community to block only one of two edit warring users simply because the one has crossed the arbitrary threshold of three reverts in one day. (Of course, this is suspect to occasional exceptions; for example, if one editor is clearly being uncivil or has a long history of edit warring, it's not unfair to block only that editor.) Perhaps more important is to realise that there is a great deal of edit warring that 3RR does not cover. It's even possible to edit war while following 1RR if, for example, one consistently reverts to one's own revision once per day, every day. We mustn't be afraid to block for edit warring, even if the edit warring doesn't cross the threshold of three reverts in one day.

If an admin feels action needs to be taken against an edit war, but is hesitant to block, here's a solution I've found: Watchlist the article and warn the participants that if they continue the edit war, they will be blocked. This prevents the edit war from continuing (at least among the current participants), assuming that you make good on it and block any user who continues the edit war.