User:K6ka/Testing Wikipedia's defenses
Wikipedia prides itself on its newbie-friendly policies, and the policy on assuming good faith and being bold. Wikipedia ensures that as much information in its articles are accurate, with its strict policy on citations and reliable sources. Wikipedia maintains a high quality of writing — at least in its better articles — through its neutral point of view policy.
Editors that try to build the encyclopedia are always welcome, as long as they are genuinely trying to help the encyclopedia out. But while there are editors that try to help, there are some that are intrigued by the very fact that you can edit its pages. Some shy editors try this out, making extremely minor edits, save their edit, and are almost scared to see what happens to it next. Braver editors edit a bit more, and with enough frequency for them to be noticed and then later welcomed. These editors may find the welcome message to be the "ignition" for the editor inside them, and they start involving themselves in community or article discussions, or just continue to build, fix, or copy-edit articles, or start taking up different tasks. Some people are curious to see what would happen if they use the edit button to insert profanity, nonsense, or their stupid commentary. These editors are usually just curious to see what happens, and usually leave after the first or second warning on their talk page. Rarely, said editors may decide to "convert" and build articles, or take up the task on the person that reverted their edit and help catch other vandals. There are a large number of editors, though, that deliberately try to damage the encyclopedia. Many of them know that anyone can edit Wikipedia and use that as their argument as to why Wikipedia is not reliable. Others have vandalized Wikipedia in the past and enjoyed the attention they got, maybe not from readers, but from the "soldiers" of Wikipedia that look for and revert vandalism. These people all have one thing in common — they test the encyclopedia's defenses.
What are Wikipedia's defenses?
Before we begin, it's important to know that virtually all actions on a MediaWiki website can be reversed. Almost all actions are logged, and all revisions to a page are stored in its page history. This trait exists in all MediaWiki websites, on all wikis. This is probably the biggest reason why wikis like Wikipedia are able to survive. In short - it's a defense, and possibly the biggest defense.
Edit filters (formerly "Abuse Filter", and the extension is called Abuse Filter) are another defense of Wikipedia. These filters monitor every single change made to the encyclopedia and compares it with specific conditions. If the conditions match, the filter can either tag the edit for human review, block the edit, warn the editor, or block the editor. All filter actions are logged in Special:AbuseLog. The filters catch a massive amount of vandalism and spam before they even make it to the encyclopedia.
ClueBot NG is probably the most documented and the most famous of Wikipedia's defensive squad. The recent changes patrolling bot catches over 40% of vandalism that slips past the edit filters with a 0.1% false positive rate. At times where the bot is down, the rate at which vandalism slipped past undetected were reported to increase significantly, making this bot an extremely important asset for Wikipedia.
The most numerous of Wikipedia's defenses are, predictably, its human editors. Specifically, it's human recent changes patrollers, new pages patrollers, and random page patrollers. While humans may not examine every edit like CBNG or the edit filters do, they are also more flexible than the automated systems, can determine what is made in good faith and what isn't, can catch vandalism that were missed by the automated defenses, and can talk to vandals. While human defenses number in the thousands, not all of them are active.
All of these defenses can, will, and are being tested on a daily basis.
The issues with testing Wikipedia's defenses
Soldiers in the military are tested, or drilled, constantly. If you are a sailor on a navy submarine, you probably know that a drill can be called at any time, ready to drag you out of your bunk at 3am to see how well you can manage with a toxic gas drill. Drills in the military are important, as they keep the soldiers alert and on their feet, so they will be ready if a real enemy submarine shows up and threatens to attack.
Wikipedia does not work that way.
There are no formal "drills" on Wikipedia — skilled or experienced editors do not deliberately vandalize in order to whip the recent changes patrollers back into shape. For one thing, recent changes patrollers have enough work of their own to do. Deliberately trying to crank it up is of no use and completely pointless. Secondly, this is against Wikipedia's policies. It doesn't matter if Jimbo Wales himself decided to see if recent changes patrollers would catch his own vandalism. He would be reverted and blocked, and the whole world will know that Wikipedia's founder was blocked from his own website in a matter of minutes.
Recent changes patrollers are thrown with over 100 edits a minute, and RCP has a high editor burn-out rate if one is not careful. No matter how fast a RC patroller works, he or she will miss lots and lots of edits. Tools like Huggle, Twinkle, and STiki make the job a bit easier, but it is no doubt stressful.
So, by adding onto a RC patroller's work by deliberately vandalizing Wikipedia is of no good to them, to the website, and to the vandal. To the vandal, because they are clearly wasting their life pissing other people off.
Many teachers and educators who are against Wikipedia often vandalize Wikipedia to prove a point. This falls under Do not disrupt Wikipedia to illustrate a point. You will certainly illustrate your point, but you will be seen more as a troublesome troublemaker that needs to be fired ASAP instead of a crown researcher that brought knowledge into the world.
Most other people who vandalize Wikipedia deliberately are usually seeing if the edit button actually works, to gain attention and fame, to test the defenses of Wikipedia, or a mixture of all three. Neither of these reasons are valid reasons to vandalize. And it's just more work for a RC patroller to clean up after such people, when everyone could've been writing a featured article. Wasted manpower, wasted time, wasted energy.
But what if I'm doing this for research purposes?
Really? Well then you must be the worst researcher ever! Remember that any research that is legally and professionally done causes no harm to both the subjects and the researchers. If your research operations involve interviewing editors (usually via their user talk page), or a group of editors in a WikiProject, or watching editors do their work, then you're more than welcome to learn more about our website. If your research operations involve going to random articles and vandalizing them to see how fast ClueBot NG can revert them, or to build spambots not only to test how quickly administrators can block the bot, but to test your spambot-programming skills, then you can forget about any "research" operations and burn your notebook and your laptop in a fire. And I don't say this lightly. Really, you should not disrupt Wikipedia to illustrate a point, nor should you climb the Reichstag dressed as Spiderman. You will make a fool of yourself, and piss everyone else off.
At the end of the day...
Treat Wikipedia's defenses as an emergency telephone number, like 9-1-1, 9-9-9, or 1-1-2. They should only be used for emergency purposes. It is illegal in many countries to abuse the emergency services, such as calling them when there is no emergency or prank calling them. The same goes here. It is unlawful to vandalize Wikipedia, and so please do not waste other people's time trying to revert your stupid edits, and please do not waste your own time by doing something so pointless and cheesy and stupid. The time you waste vandalizing Wikipedia could be better spent on your academic studies, or maybe you should go and pump some iron. Or even go and upload pictures of your breakfast to Instagram. You clearly have better things to do. So do we.
- User:ClueBot_NG#Statistics: False positive rate used to be 0.25% which caught 55% of all vandalism. Previous anti-vandal bots only managed around 10% vandalism.