User:Mz7/CVUA/Girth Summit

Hello Girth Summit, and welcome to your Counter Vandalism Unit Academy page! Every person I instruct will have their own page on which I will give them support and tasks for them to complete. Please make sure you have this page added to your watchlist. Your academy page has been specifically designed according to you and what you have requested instruction in - for that reason, please be as specific as possible in your answers, so that I know the best ways to help you (and do not be afraid to let me know if you think something isn't working). If you have any general queries about anti-vandalism (or anything else), you are more than welcome to raise them with me at my talk page.

Make sure you read through Wikipedia:Vandalism as that's the knowledge which most of the questions I ask you and tasks you do will revolve around.

How to use this page

This page will be built up over your time in the Academy, with new sections being added as you complete old ones. Each section will end with a task, written in bold type - this might just ask a question, or it might require you to go and do something. You can answer a question by typing the answer below the task; if you have to do something, you will need to provide diffs to demonstrate that you have completed the task. Some sections will have more than one task, sometimes additional tasks may be added to a section as you complete them. Please always sign your responses to tasks as you would on a talk page.

The start

Good faith and vandalism

When patrolling for vandalism, you may often come across edits which are unhelpful, but not vandalism - these are good faith edits. It is important to recognise the difference between a vandalism edit and a good faith edit, especially because Twinkle gives you the option of labelling edits you revert as such. Please read WP:AGF and WP:NOT VANDALISM before completing the following tasks.

Please explain below the difference between a good faith edit and a vandalism edit, and how you would tell them apart.

The difference is fundamentally about intent - a good faith edit is done by someone who is trying to improve the project (but may be misguided); vandalism is intentionally trying to frustrate the project's efforts.

Identifying someone's intent from what they have written/deleted is not always easy; you have to try and get a sense of the spirit in which the edit was made. As well as the edit itself, you may need to look at the editor's contributions - an unhelpful edit made by an IP or newbie with little history might be considered good faith, but an experienced user is likely to know more about how things work, and a pattern of subtly unhelpful contributions or repeated insertions of the same unhelpful content might raise concerns .

When I'm assessing whether an edit is vandalism or not, do my best to work out what the person's intent is.

• If content has been added that doesn't make sense, I try to work out what the author was trying to say from the content and the vocabulary - has the user attempted to add information, but perhaps their grasp of English isn't sufficient? Sentences that don't make perfect sense are likely to be good faith, whereas long strings of random characters are quite likely to be vandalism.
• If content has been removed, I look at what has been cut, and what has been retained - is someone making a good faith effort to remove information that they believe is wrong? If specific sentences or even whole sections are removed, it's likely to be good faith, particularly if they are sections that are likely to be contentious. If someone has deleted at random, cutting sentences into pieces, it seems more likely to be vandalism.
• If content has been changed selectively, again I'd try to work out the intent. An editor might have a preference for one name, phrase or spelling over another, which would justify a change in their eyes; or, even if they change reliably sourced content to some unsourced woo, they might still be doing this with the intention of improving the article, so it would not be vandalism. If, on the other hand, someone is changing dates of birth on a BLP to obviously impossible dates, or into the future, or replacing someone's name to 'Mickey Mouse' or whatever, it's obviously going to be vandalism.

Y @Girth Summit: Wow, these are really excellent, in-depth responses! Indeed, the key with good faith is intention. Even if an edit has completely messed up an article, or even many hundreds of articles, if it was done with the intent of improving the encyclopedia, then it is good faith and should be treated differently from vandalism. It is important not to mislabel good-faith edits as "vandalism", since this is unnecessarily hostility that can drive away potential newcomers and even experienced editors. If you are ever in any doubt as to whether an edit is good-faith or bad-faith, the guideline on Wikipedia is to assume good faith. Mz7 (talk) 08:26, 18 October 2018 (UTC)

Please find three examples of good faith but unhelpful edits, and three examples of vandalism. You don't need to revert the example you find, and I am happy for you to use previous undos in your edit history if you wish.

I've tried to present examples of each of the situations outlined above - addition of content, removal of content, and changing existing content. Hope that makes sense.

Good faith

• Addition of content: this edit was POV, ungrammatical and unencyclopedic; however, the IP only has two contributions, so it could easily be someone who doesn't know how Wikipedia works - probably a fan who wanted to add a tribute to his favourite player. Misguided, but not vandalism.

Y Mz7 (talk) 08:26, 18 October 2018 (UTC)

• Removal of content: This appeared to be a good faith attempt to improve the page. They removed an entire section that appeared to be well-sourced without any discussion, so I reverted the change and asked them to discuss it on the talk page; however, they left an edit summary explaining what they had done, and the section was removed without damaging anything else about the article, so I am convinced that the editor believed they were removing material that would not advance the reader's understanding of the subject.

Y Mz7 (talk) 08:26, 18 October 2018 (UTC)

• Change of content: this edit changed the 'CE/BCE' date format on a page to the 'AD/BC' format, which was used throughout the article and appeared to have consensus on the page. This is presumably the editor's preferred format, and WP:MOS doesn't say which we should use, so I wouldn't call this vandalism; however, WP:MOS does say that you shouldn't change from one convention to another without a good reason, and this change was undiscussed and did not have consensus, so I thought it was problematic.

Y Nice - great understanding of the MOS. The relevant section is MOS:ERA. Mz7 (talk) 08:26, 18 October 2018 (UTC)

Vandalism

• Addition of content: It's borderline, but I believe that this was vandalism. I guess it's conceivable that the editor thought that this was a useful thing to note in an Encyclopedia, and it was an IP with no other contributions; however, it contains profanity, and is clearly unencyclopedic in tone. I believe that the editor is a Scottish native English speaker (the subject is about a traditional Scottish drinking vessel, the way the profanity was spelled is a Scottish variant, and they are indeed often given as wedding gifts within Scottish culture), so this is someone who knows a word like that isn't appropriate - it would fall under the 'silly vandalism' category.

Y Given the profanity and the unecncylopedic nature of the addition, I'd say this is pretty clear vandalism. Nice find. Mz7 (talk) 08:26, 18 October 2018 (UTC)

• Removal of content: this appears to be vandalism - it removes part of the first sentence of the article, cutting a word in half and leaving the sentence devoid of meaning. The edit summary 'Good' gives no reason for the change, and I believe the editor was probably just messing about (a bored schoolkid working on an essay perhaps).

Y Mz7 (talk) 08:26, 18 October 2018 (UTC)

• Change of content: I believe that this is vandalism - the title of the article was 'Joe DiMaggio'; the lead gives the subject's name as 'Joseph Paul DiMaggio'; he's a very well-known figure. I don't believe that there's any likelihood that the editor actually thought his name was Cheeseman, they just thought that it would be funny to change the photograph's caption.

Y Mz7 (talk) 08:26, 18 October 2018 (UTC)

A note about Twinkle

Try to use either the green, blue, or red "rollback" links if you can, since they save a more relevant edit summary

I noticed that you've been using Twinkle's brown "restore this version" button to revert edits. You'll notice that if you examine the diff of the most recent change to a page, Twinkle will also provide you with green, blue, and red "rollback" links. See the screenshot. All three will revert all of the most recent consecutive edits made by a single user to a page.

Try to use these buttons where possible, since these links will save a more relevant edit summary (except the red one, which immediately reverts the edit, saving a generic edit summary – use that for obvious vandalism and other edits for which the reason for reverting is immediately clear). For example, if you use the green "AGF" rollback button, it will note in the edit summary that you thought the edit was good faith. All three will also identify the username/IP of the edit you are reverting, so it's clear to other patrollers. Use the brown "restore this version" button when you need to revert edits by multiple users. Mz7 (talk) 08:36, 18 October 2018 (UTC)

Perhaps I've misunderstood - can I use these functions without having rollback rights? I haven't experimented with them because I assumed they wouldn't work for me. GirthSummit _(blether) 08:55, 18 October 2018 (UTC)

Scratch previous comment - I've tried it now and it worked - probably should have read the instructions more carefully. I'll use that in future, thanks. GirthSummit _(blether) 09:09, 18 October 2018 (UTC)

Warning and reporting

When you use Twinkle to warn a user, you have a number of options to choose from: you can select the kind of warning (for different offences), and the level of warning (from 1 to 4, for increasing severity). Knowing which warning to issue and what level is very important. Further information can be found at WP:WARN and WP:UWUL.

Please answer the following questions

Why do we warn users?

It seems to me that there are three ways to answer this:

• Primarily, we warn them to let them know that they are not editing according to Wikipedia's policies, in case they don't realize that. Wikipedia invites everyone to edit, and there is no requirement for them to read through all the policies before they begin. The web is packed with open forums where people are actively encouraged to post opinions, rants, or anything they want - Wikipedia is a different sort of place, and we need to draw their attention to that.
• By issuing warnings of increasing levels of severity, we are creating a record of our having informed them. If admin attention is eventually required because of persistent disruption or vandalism, the admin can see that they have been warned multiple times, so it makes their decision about appropriate action easier.
• We use warning templates, rather than writing a more personal note to the user, because it's simply more practical - if you had to write an individual message to every user who makes problematic efforts, we would be swept away by the tide! Having said that, I like to use more personal messages at times, if I think it's a good faith user who might need more than an automated message to see what they're doing wrong. The templates that allow you to add a message manually at the same time as leaving the template are good for this.

Y @Girth Summit: Really in-depth answers! I agree with your view on template warnings; they're great for practical use, but when there is a situation that doesn't quite fit a template message, I wouldn't hesitate to write a custom message the old-fashioned way! The purpose of warnings is to educate users who might not know better – I feel that a good portion of Wikipedia vandalism is just people testing out the editing interface ("can I really edit?") or not realizing that their edits are really viewable around the world immediately. (Some newbies have confessed to thinking that their edits would only be saved on their own computer, not the whole world.) Warnings also inform vandals that their ability to edit can be revoked by administrators if they persistently vandalize – indeed, admins are much more likely to block a vandal who has been warned multiple times than a vandal who has never been warned. Mz7 (talk) 00:35, 25 October 2018 (UTC)

When would a 4im warning be appropriate?

According to WP:UWLEVELS, 4im assumes bad faith, and is for 'excessive or continuous disruption'. First you'd need to be confident that edits were vandalism rather than good faith. Then, you'd need to be sure that it was excessive or continuous.

• Excessive disruption might be outrageously offensive stuff, like inserting racist abuse into a page, or highly defamatory material into a BLP.
• Continuous disruption might be if I looked at a user's contribs and found a long history of adding obvious vandalism, but found no prior warnings on their talk page (or talk page history). Another example would be if I reverted a piece of vandalism, and the user reinstated it before I'd managed to issue a lower level of warning - they basically start edit warring to keep it on the page.

Y Right, I would jump to a level 4 immediately if the edits are particularly egregious or are spreading across many articles – use it against edits that need to stop as soon as possible, where it is clear that there are no good intentions. Mz7 (talk) 00:35, 25 October 2018 (UTC)

Should you substitute a template when you place it on a user talk page, and how do you do it?

I had to research a bit to find the answer to this! WP:UWUL and H:SUBST were instructive. Yes, you should substitute a template if you are adding it manually, so that the text displayed on the recipient's talkpage doesn't change if the template is changed at a later date. If you add a warning template using Twinkle (as I normally do since I started using it), this is done automatically. If I was manually adding a template, I would add subst: inside the braces before the template name. So, for example, {{uw-test1}} would become {{subst:uw-test1}}.

Y This is correct; you do need to substitute a templated warning message when you add it to user talk pages, and adding the subst: keyword in the curly braces is indeed how you do it. You also have the correct reason for why substitution is necessary: when we send a message, we don't want the message changing over time to something we didn't originally mean. Mz7 (talk) 00:35, 25 October 2018 (UTC)

What should you do if a user who has received a level 4 or 4im warning vandalises again?

I would seek assistance from an administrator, using the ARV tool in Twinkle, and select the AIV option, providing the page where the latest vandalism occurred, and ticking the box for 'Vandalism after a level 4 or 4im' box. GirthSummit _(blether) 09:44, 18 October 2018 (UTC)

• Here is an example of me doing this recently.GirthSummit _(blether) 06:44, 19 October 2018 (UTC)

Y WP:AIV, short for Wikipedia:Administrator intervention against vandalism, is the correct answer. Mz7 (talk) 00:35, 25 October 2018 (UTC)

@Mz7: This page is ready for you to review (apologies if you already knew that and have just been busy - I was just wondering if you were waiting for me to ping you correctly from within the form). GirthSummit _(blether) 07:30, 22 October 2018 (UTC)

@Girth Summit: Very sorry about the wait. I've been a little busy this past weekend, but I should be free to give feedback in the evening tomorrow. Mz7 (talk) 04:41, 23 October 2018 (UTC)

@Mz7: That's great - you did warn me at the outset that you might not be able to respond immediately, so no worries about that, just wanted to make sure that you weren't waiting for me to do something. Look forward to your feedback.GirthSummit _(blether) 06:15, 23 October 2018 (UTC)

Tools

Wikipedia:Recent changes patrol#Tools includes a list of tools and resources for those who want to fight vandalism with a more systematic and efficient approach. In addition to manually going through Special:RecentChanges, there are a large number of tool which assist users in the fight against vandalism. They range from tools which help filter and detect vandalism to tools which will revert, warn and report users. Here are a few.

Twinkle

Twinkle, as you know, is very useful. It provides three types of rollback functions (vandalism, normal and AGF) as well as an easy previous version restore function (for when there are a number of different editors vandalising in a row). Other functions include a full library of speedy deletion functions, and user warnings. It also has a function to propose and nominate pages for deletion, to request page protection to report users to WP:AIV, WP:UAA, WP:SPI, and other administrative noticeboards.

User creation log

In my early days of fighting vandalism on Wikipedia, one of the strategies I would use to find vandalism was to patrol the account creation log. This is located at Special:Log/newusers, and it logs every time a new user account is created on Wikipedia. You'll notice that new accounts with no contributions so far will have a red "contribs" links, whereas new accounts with some contributions will have blue "contribs" links. One great way not only to find vandalism, but welcome new users to Wikipedia is to check the blue contribs links that come in.

Edit filter log

Special:AbuseLog is the edit filter log. Every time you and everyone else submit an edit to Wikipedia, your edits are checked against hundreds of filters to make sure they aren't disruptive. These filters look for common vandalism patterns (e.g. adding profanity), and if a filter returns a hit, it gets logged in Special:AbuseLog.

In addition to logging, the edit filter can also trigger specific actions; namely, filters can show a warning to the user or even disallow an edit from going through entirely. Even if an edit is disallowed, it is sometimes useful to look at the editor's contributions to see if there was other vandalism not caught by the filter. The template warning series {{subst:uw-attempt2}}, {{subst:uw-attempt3}}, and {{subst:uw-attempt4}} are also useful against edits disallowed by the edit filter (there's no level 1 in this series, for some reason).

Question I'm having a look at this just now, and trying to figure out what everything means. Is it possible for a single edit to trigger the same filter multiple times? Looking at 06:22, 25 October 2018, I can see six entries for IP 141.226.9.80 - three triggering Filter 432, and three triggering Filter 135. Looking at the actions taken, they're all either 'Tag' or 'Warn' - but none of them are disallow. Looking at 141.226.9.80's contribs, I only see one edit making it through. Are you able to explain what is happening here?

STiki

STiki is an application that you download to your computer, and it provides you with diffs which either it or User:ClueBot NG have scored on their possibility of being uncontructive, and you are given the option to revert it as vandalism, revert it assuming good faith, mark it as innocent, or abstain from making a judgment on the diff. In order to use STiki, you need one of the following: (1) the rollback permission, (2) at least 1000 article edits (in the article namespace, not talk/user pages), or (3) special permission via Wikipedia talk:STiki.

Huggle

Huggle is also an application you download to your computer which presents you diffs (orders them on the likelihood of being unconstructive edits and on the editor's recent history) from users not on its whitelist. It allows you to revert vandalism, warn and reports users in one click. The rollback permission is required to use Huggle.

As you know, I've applied for rollback rights so I can try and use these (although I think I might be able to use STiki already based on my mainspace edit count). I'll download and have a look later on. GirthSummit _(blether) 06:38, 25 October 2018 (UTC)

Rollback

After reviewing your recent contributions, as well as considering the really in-depth nature of your responses, I noticed that you already seem to be pretty experienced at countering vandalism! I think that if you apply for the rollback permission at Wikipedia:Requests for permissions/Rollback, an administrator would be happy to enable it on your account. (Although I am an administrator, I prefer not to use the admin toolset on my students' accounts in accordance with the WP:INVOLVED policy).

The rollback user right allows trusted and experienced vandalism fighters to revert vandalism with the click of one button, not unlike the "rollback" button that you've already been using in Twinkle. This would give you a new rollback button in addition to the three you've been seeing in Twinkle. The new rollback button is faster than the Twinkle rollback button, but more importantly, having rollback gives you access to downloadable counter-vandalism software like Huggle.

If you're interested, take a look at our rollback guideline at WP:Rollback and feel free to answer the questions below. If you're not interested, feel free to skip this section. Mz7 (talk) 00:57, 25 October 2018 (UTC)

Describe when the rollback button may be used and when it may not be used.

The main advantage of rollback is its speed, and its main purpose is to revert obvious vandalism as quickly as possible. The vandalism has to be so blatant however that nobody would need an edit summary to understand why you made the change.

As well as reverting vandalism, it is permissible to use rollback to revert edits on your own user pages, to revert your own edits, to revert edits of blocked or banned users (but you should be ready to explain your actions if you do this), or to revert multiple unhelpful edits from a malfunctioning bot or misguided user - but if you do this, you need to leave a note explaining why you're doing it at a suitable location, e.g. the article's talk page (which would probably take longer than just doing a regular revert with an edit summary...).

The above are the only circumstances that rollback may be used in. It may not, for example, be used to revert changes that violate policy but are not vandalism - unsourced additions, OR, SYNTH, good faith removal of content, etc. In these situations, a regular revert with edit summary is required.

Y Mz7 (talk) 20:59, 14 November 2018 (UTC)

Hopefully this will never happen, but it does occasionally. If you accidentally use rollback, what should you do?

There are two options.

• If you didn't mean to do a revert, you can self-revert manually, with a summary along the lines of 'revert accidental use of rollback'
• If you wanted to revert, but accidentally used rollback rather than a manual revert, you can follow the rollback with a dummy edit and insert an explanatory edit summary.

Y Mz7 (talk) 20:59, 14 November 2018 (UTC)

Should you use rollback if you want to leave an edit summary?

By default, the answer would be no; however, according to WP:RBK, it is possible to use certain tools to modify your rollback functionality and have it insert edit summaries when you use rollback. I haven't played around with any of this yet, so can't describe how this works; to be honest, I think that if I wanted to leave an edit summary, I would probably just use a regular revert or Twinkle blue rollback.

Y Mz7 (talk) 20:59, 14 November 2018 (UTC)

Protection and speedy deletion

Protecting and deleting pages are two additional measures that can be used to prevent and deal with vandalism. Only an administrator can protect or delete pages; however, anyone can nominate a page for deletion or request protection. If you have Twinkle installed, you can use the Twinkle menu to request page protection or speedy deletion (the RPP or CSD options).

Protection

Please read the protection policy.

In what circumstances should a page be semi-protected?

Semi-protected prevents unregistered users (IPs) and very new accounts from editing a page, but allows confirmed/auto-confirmed users to edit it normally. Since the majority of vandalism comes from unregistered users (or new accounts created purely to cause disruption), it is useful for pages that constantly attract a large amount of vandalism (e.g. Barack Obama) or content policy violations (e.g. Vaccine) to be permanently semi-protected.

Semi-protection can also be applied temporarily when a page is getting a lot of traffic because its subject is currently in the news, if an edit war is causing disruption, or if a particular vandal is targeting it and blocks aren't working because they are IP-hopping.

Semi-protection should not be applied pre-emptively - wait for the problems to start happening before dealing with them. (This holds true for 'pending changes' and 'full protection' too.)

Y Right, semi-protection is the go-to level when a subject is getting an unusually high amount of attention and vandalism, usually because the subject is in the news. Sometimes, we get vandals that hop on different IPs, targeting a specific article; in some cases, we protect articles due to these vandals as well. Mz7 (talk) 20:59, 14 November 2018 (UTC)

In what circumstances should a page be pending changes protected?

This is similar to semi-protection: anyone can edit, but edits from IPs/new users are only visible to confirmed users until they are approved. The advantage over semi-protection is that it allows positive contributions from people who don't have accounts; the disadvantage is that it creates work for editors to check and approve the changes. It should be used in similar situations as semi-protection, but for pages that receive low volumes of traffic (and so don't generate a mountain of approval work). HIV/AIDS denialism is an example - it gets a certain level of POV-pushing edits, but doesn't attract enough attention to require semi-protection.

Y Exactly, low-volume of traffic is the key here. Also keep in mind that it doesn't really stop the vandalism from happening; the main benefit is in preventing the vandalism from being seen by readers, which is particularly useful on BLPs that aren't closely watched. Mz7 (talk) 20:59, 14 November 2018 (UTC)

In what circumstances should a page be fully protected?

This is the highest level of protection - only admins can edit. Again, it's for articles that receive a lot of vandalism and/or disruptive content disputes, but only when the disruption is regularly caused by confirmed/auto-confirmed accounts (and so couldn't be stopped with semi-protection) - Quran is one example. It's also used for critical templates, like Template:Citation needed, which would screw up thousands of articles if someone tinkered with it. Deceased Wikipedian's user pages (but not talk pages) are should also be fully protected.

Y Content disputes and high visibility templates are the most common reasons why pages are fully protected. It is rarer to see an article fully protected for vandalism, particularly with the introduction of extended confirmed protection in 2016, but it has happened before, e.g. on Donald Trump during his election year. Mz7 (talk) 20:59, 14 November 2018 (UTC)

In what circumstances should a page be creation protected ("salted")?

Salting would take place when an article which has been deleted is repeatedly recreated. These might be for someone who is constantly trying to create a promotional page for a non-notable person or company, or for attack pages, or for plain old offensive page names.

Unlike other forms of protection, this type of protection is used pre-emptively, via the title blacklist, to stop people creating obviously unnecessary pages.

Y You're right that there's some preemptive title blacklist entries, but I would say most salting occurs after there is evidence that a title is being repeatedly recreated. Mz7 (talk) 20:59, 14 November 2018 (UTC)

In what circumstances should a talk page be semi-protected?

Article talk pages can be used to request changes to protected pages; they are also less frequently the target of vandalism than mainspace articles. For these reasons, they are generally not protected, but in the most serious cases of vandalism they may be semi-protected for a limited duration, allowing only confirmed accounts to request changes to the page.

Y Mz7 (talk) 20:59, 14 November 2018 (UTC)

Correctly request the protection of one page (pending, semi or full); post the diff of your request (from WP:RPP) below.

Here is an example of me requesting temporary page protection for People's Vote, which is in the UK news currently following a protest march in London. The article had been subject to unsourced POV-pushing editing, so I had added a couple of sourced sentences myself about the march in an attempt to achieve a level of balance that would keep both 'sides' happy, and the article was on my watchlist. I noticed a couple of IPs repeatedly inserting the 'goatse' image to the page, after being reverted by several editors including myself. The IPs were quickly blocked, but I was concerned that they would come back under a new IP, so requested page protection.

Y Request was carried out by Ymblanter. Mz7 (talk) 20:59, 14 November 2018 (UTC)

Speedy deletion

Please read WP:CSD.

In what circumstances should a page be speedy deleted?

Deleting articles is usually done by community consensus following discussion at WP:AFD. In certain cases, where it is obvious that a page needs to be deleted, administrators are allowed to use their discretion and delete a page without discussion - this should be done when the article meets one or more of the speedy deletion criteria.

This diff is me nominating a page (an obviously promotional userpage) for speedy deletion via Twinkle.

Just to evidence the fact that I've read through them, I'll summarise my understanding of the criteria below - hope this isn't excessively wordy, but I find the process of summarising helpful in making sure that I've understood something properly.

• G1 - patent nonsense. Meaningless text, random characters etc. (This is explicitly not just stuff that is obviously wrong, poorly written, or foreign language - it's only stuff that doesn't make any sense linguistically.)
• G2 - test pages. Pages that have been created to test whether something works, and are no longer needed.
• G3 - pure vandalism and blatant hoaxes - obviously intended purely to offend or misinform readers
• G4 - recreation of a page deleted after a deletion discussion. Does not apply if the content is substantially different from the previous version, or to pages that were deleted previously via speedy deletion or a soft deletion.
• G5 - pages created by a blocked or banned user, during their block or ban, and not substantially changed since creation by other users.
• G6 - technical deletions. Tidying up basically - deleting orphaned templates, empty categories, disambiguation pages with only one entry, etc.
• G7 - author requests deletion. Only permissible if no other editors have put substantial work into the article.
• G8 - pages dependent on a non existent page. Eg: talk pages without an article, redirects to invalid targets.
• G9 - office actions. WMF reserves the right to speedy delete pages if it decides this is necessary.
• G10 - attack pages - pages consisting of unsourced negative material, libel, legal threats or that are intended purely to distress or harass someone.
• G11 - unambiguous advertising or promotion - pages which are only there to push a particular product or company, and which would need to be fundamentally rewritten to achieve a neutral POV. I've seen quite a few of these, often in user space.
• G12 - copyright violations, e.g. stuff copied directly from another website. If it's only part of the page, it might be better to remove the copyvio and fix the rest of the page.
• A1 - no context. Very short articles, without enough information to let you work out what the article is about. Not to be applied to new pages, allow sufficient time for more content to be added that might establish context.
• A2 - foreign language articles that are essentially identical to existing articles that exist elsewhere on Wiki (e.g. duplicates from French Wikipedia). Not to be used if the content is substantially different - in that case, flag for translation into English.
• A3 - no content - articles with no actual text about the subject, just external links or categories.
• A5 - transwikied content - stuff that's already been moved to Wiktionary or elsewhere on the project
• A7 - no indication of importance (people, animals, company, website, events) - lower threshold than WP:NOTABILITY, content has no indication whatsoever of why they would be worth having an article about. E.g. if I created an article about my dog.
• A9 - no indication of importance (musical recordings) - if the artist/s have no article, and there is no indication of why the song is important enough to have an article. E.g. if I made a list of all the songs I played when I was in a (very bad) student band 20 years ago.
• A10 - recently created article, duplicate existing topic - if someone creates a new article for a subject that already has a page, and the article has no useful new content
• A11 - obviously invented. Subject has obviously been invented by the article's author, and contains no credible indication of why the invention is important. (Different from hoax, which is intended to mislead)

I haven't been through the redirect and file criteria, since your question was about pages - let me know if you would like me to comment on these as well.

@Mz7: I've responded to the latest round of questions. (I'm on half-term holiday at the moment, and my partner's away at a conference, so got a bit of time on my hands...). Just in case you miss it, I asked a question in the section about the Edit filter log. Looking forward to your feedback when you have time; I know I've put quite a bit of text here so appreciate that it might take you a while to get round to it. Cheers! GirthSummit _(blether) 09:32, 25 October 2018 (UTC)

Today, I was for the first time the subject of vandalism on Wikipedia, which I can only assume came from a frustrated vandal whose fun I had interfered with. I assume that this is something that editors working on counter vandalism have to get used to, and it hasn't deterred me - to be honest, I was rather tickled, nobody's ever described me as a giant, terrifying beast before! I celebrated this rite of initiation with a userbox on my user page - I hope you don't think that this is excessively flippant; I'd be happy to remove it if you think it might be an incentive to further disruption. GirthSummit _(blether) 18:31, 25 October 2018 (UTC)

Hah! It's nice to see you're taking this in stride. There are a few editors who have felt threatened by more personal vandalism. I don't think there is a big problem with your userbox, since the vandalism wasn't that serious, but for the future keep in mind that more chronic vandals love getting attention – see WP:DENY. Sometimes it's in the best interest of the project to just revert and ignore. Mz7 (talk) 22:58, 26 October 2018 (UTC)

Questions about reporting vandalism for oversighting/revision deletion

@Mz7: I hope it's OK for me to ask you some questions as part of this process! Following the comment from the admin approving my rollback rights, I've done a bit of reading about oversight and revision deletion. I'm not 100% clear yet on when I should request them, and how best to go about doing it. I don't want to fail to report anything that needs reporting, but at the same time I realise admins are busy, and you probably don't want hundreds of e-mails or talk page messages asking for unnecessary revdels - I want to strike the right balance. Perhaps this is something that we will cover anyway as part of the CVUA training, but would you mind giving me your thoughts on the following? Thanks in advance GirthSummit _(blether) 16:43, 26 October 2018 (UTC)

How would you describe the threshold for a piece of vandalism to trigger a report to admins for revision deletion?

Are there specific triggers that would always trigger it, or is it always more of a judgment call?

@Girth Summit: These are really good questions. Officially, criteria for revision deletion can be found at WP:REVDEL#Criteria for redaction. There are three typical cases:

1. RD1: Blatant copyright infringement
2. RD2: Grossly insulting, degrading, or offensive material
3. RD3: Purely disruptive material

You can read the criteria for a fuller description of each criterion. Against vandalism, RD2 and RD3 are most often used (the distinction between RD2 and RD3 is subtle; I think in a lot of cases, they are fairly interchangeable). For me, the threshold is twofold. Firstly, is there a specific, identifiable individual being targeted by the vandalism? Secondly, is the vandalism more serious than just "common vandalism"? For example, "X is gay" or "X is an asshole" is fairly common vandalism that doesn't really need revision deletion, but unsourced edits that accuse a specific individual of sexual crimes is more egregious that would warrant revision deletion and likely also oversight. Use your best judgement here. Mz7 (talk) 22:49, 26 October 2018 (UTC)

Thank you for that - that's clarified for me the distinction between a general (but vandalstic) edit, and one which invokes a specific charge needing admin attention.

Similarly, how would I judge whether to request oversight rather than revision deletion?

The formal criteria for oversight can be found at Wikipedia:Oversight#Policy. If an edit contains any kind of personal information that isn't publicly available, such as a phone number or address, then oversight is the first thing that should pop into your mind. If you spot sensitive information about a child (e.g. a nine-year-old editor posting their age and full name on their user page), the oversight team has historically been willing to suppress that information as well.

Oversight is a step above revision deletion in that while administrators can still view revision-deleted material, not even administrators can view oversighted material. For other kinds of vandalism, consider whether the allegation in the vandalism might constitute libel or slander – potentially defamatory information should be forwarded to the oversight team for review. It's probably better to refer an edit to the oversight team and have the request declined than to not refer a potentially defamatory edit and have Wikipedia be the subject of media attention or a lawsuit. For example, as I mentioned above, if an edit accuses a specific individual of a sexual or particularly heinous crime, that strikes me as something that should go to oversighters for review. Mz7 (talk) 22:49, 26 October 2018 (UTC)

What is your preferred method for editors to request revision deletion?

The advice on WP:OVERSIGHT is clear - you should e-mail the oversight team. The advice at WP:Revision deletion is a bit vague however - it says you can drop a note on the admin's talk page, but then it also says that it might best you might want to e-mail it to them. I've had a look at a few of the admins' pages at CAT:REVDEL, and I'm not seeing many 'e-mail me' links. Does it mean e-mail the oversight team?

It's probably best to keep stay off user talk pages expect as a last resort. We want to avoid the Streisand effect, where our attempts to hide something inadvertently cause it to receive more attention. My preferred method is to use the IRC channel #wikipedia-en-revdel ^connect. Be sure the include the phrase !admin when you make your request, as a lot of admins in that channel have that word highlighted so that it makes an audible ping on our computers when it's used. When you look at the user pages of the admins at CAT:REVDEL, the "Email this user" link can be found in the left sidebar (underneath the Wikipedia logo, with the links to the Main page, etc., under the heading "Tools"). Alternatively, you can use Special:EmailUser and type the username of the Wikipedia user you want to email in the field provided. Almost all of the admins at CAT:REVDEL should have email enabled.

Yep - as you predicted, email links are indeed presented - my bad, I just didn't know where to look for them. GirthSummit _(blether) 00:04, 27 October 2018 (UTC)

To request oversight, the most reliable way is to use Special:EmailUser/Oversight or email oversight-en-wpwikipedia.org directly. You can also join #wikipedia-en-revdel ^connect and use the phrase !oversight – this is really quick if an oversighter is around (typically one is during the day in North America). Mz7 (talk) 22:49, 26 October 2018 (UTC)

@Girth Summit: Just to let you know – regarding the edits you reported in #wikipedia-en-revdel today (re. sexual crime allegations) – I spoke to the administrator that handled your report, as well as an oversighter, and we determined that the material did deserve to be oversighted. The diffs are now suppressed. Mz7 (talk) 03:18, 29 October 2018 (UTC)

@Mz7: Thanks for letting me know - yes, I thought it was a pretty severe case, a named individual and specific crimes etc. Regarding my 'Level 1 warning' for the account that made the edit, I'd left that on the talk page before I'd really appreciated how serious the edits were. I assume a case like that would warrant a 4im normally? (Having already put the level 1 on there, I thought that going back and changing it might provoke them, and that it would be best leaving it to an admin to deal). GirthSummit _(blether) 07:22, 29 October 2018 (UTC)

Hi @Mz7:, I wonder whether it would be possible for you to give me the next round of questions in the CVUA training? I appreciate that it might take you some time to get around to giving me feedback on my responses so far (I know tend to be a bit verbose thorough), but I believe I've got a handle on the areas we've covered so far. I had a look at Huggle, but to be honest I couldn't really get my head around the interface - I haven't decided yet whether or not I want to spend the time reading the Help pages and figuring out how it all works. I also couldn't get the STicki installer to run on my PC; I think I must be missing a Java script component or something, the PC didn't know how to open the file. Again, I may get around to looking at this in the future, but for now if there's anything more I can be learning about our policies, and things I could be doing using Twinkle, I'd be happy to proceed with that. Thanks again for all the support so far. GirthSummit _(blether) 08:24, 1 November 2018 (UTC)

@Girth Summit: Yes! I apologize for the delay. Regarding STiki, I believe you need to install Java (which is different from JavaScript): try installing the first one here. Mz7 (talk) 22:04, 2 November 2018 (UTC)

Speedy deletion examples

In past iterations of this course, I have had mentees go out and actually tag pages for deletion, but with the introduction of WP:ACPERM, the amount of straight vandalism that gets created directly in mainspace has reduced dramatically. As such, I'm now having mentees answer a set of hypothetical scenarios. What would you do if you saw the page listed in each scenario? Not all scenarios may warrant speedy deletion.

Scenario 1

A user with the username "BobSucks" creates an article called "John Smith" that contains solely the following text:

John Smith is the worst elementary school teacher on the planet.

• I'd use the Twinkle CSD tool, nominate for G3 (pure vandalism) and G10 (attack page). I'd also look at the user's other contributions - the username could be seen as disruptive, but probably not bad enough to report to UAA, but depending on there behaviour elsewhere I may report to AIV as a vandalism-only account.

Y Either G3 or G10 work, and I'd say that most administrators would look very favorably upon a vandalism-only block, given both the nature of the vandalism and the username. Mz7 (talk) 20:09, 7 November 2018 (UTC)

Scenario 2

A user with the username "GoodTimesLLC" creates a user page with the following text:

'''Good Times LLC''' is an organization dedicated to helping your children get the highest quality education at an affordable price. Visit our website at goodtimes.info and contact us at 123-456-7890.

• Twinkle - CSD - G11 (promotional userpage under a promotional user name). When I've done this in the past, the admin dealing with the nomination usually blocks the username as well, but if they didn't I'd report the username to UAA as promotional and implying shared use.

Y Mz7 (talk) 20:09, 7 November 2018 (UTC)

Scenario 3

A user creates an article titled "Edward Gordon" with the following text:

'''Edward Gordon''' (born July 1998) is an aspiring American actor and songwriter. So far, he has starred in many school plays and has published two albums on SoundCloud. He has over 5,000 subscribers on YouTube.

• This touches on a few possible infractions - COI editing, notability, etc. I'd start by checking whether the information is true - Google Edward Gordon, check YouTube and SoundCloud - is the information true? If I can't find any evidence that it's true, I might consider nominating for G3 as a hoax. Assuming it's true, I probably wouldn't nominate for CSD. If I could find some mentions in secondary sources, I might insert a couple of references and leave it as is (maybe remove the word 'aspiring'); if I couldn't find any mentions anywhere, I might send to AfD as non-notable, but I don't think it unambiguously meets A7. I would put a note on the user's talk page about COI editing, and suggest that they declare their COI on their userpage etc.

Y OK. A7 is probably the most subjective of the speedy deletion criteria. In general, I don't think that starring in high school plays and self-publishing albums on the Internet count as a credible claim of significance, nor does having 5,000 subscribers on YouTube. My opinion is that this would probably meet A7. However, it depends on the available sources – my answer would certainly change if there were more secondary sources, and you're definitely in the right to do a search for sources beforehand. Mz7 (talk) 20:09, 7 November 2018 (UTC)

Scenario 4

A user creates an article titled "Bazz Ward" with the following content:

Bazz Ward was a great roadie and I wish he was as well known as Lemmy. Cheers Bazz.

(Attribution: Ritchie333 came up with this scenario as a question to an old RfA candidate. Since my creativity sucks, I've borrowed his example here. Hint: Try Google searching a few key terms from this short article.)

• Thanks for the hint - I'd heard of Lemmy from Motorhead, but not Bazz Ward. I googled him, and found a few mentions to him (including our article on The Nice). The text as it stands is entirely unencyclopedic, and so would need either to be deleted or entirely re-written; I'm not very well-versed on music sources though, so I'm not sure whether the stuff I found like this or this would be reliable enough to build an article around. I might nominate to AfD as non-notable, and note that I wouldn't be averse to someone building a better article about him if they think it's worthwhile.

Y Nice find. Another alternative to deletion, if a topic is related to another one but doesn't quite deserve an article, is to redirect the article – this can be even done per WP:BOLD without discussion. Mz7 (talk) 20:09, 7 November 2018 (UTC)

Scenario 5

A user creates an article that was clearly copied and pasted directly from another website, which states "All Rights Reserved" at the bottom of it. Would your answer be the same if it didn't state "All Rights Reserved" at the bottom?

• CSD G12 in either case. I'm not very well-versed in copyright law, but I would have thought that the copyright owner would have to give us explicit consent to use their material, rather than our assuming consent when there is no text saying otherwise.

Y Mz7 (talk) 20:09, 7 November 2018 (UTC)

Scenario 6

A user creates an article, but you can't understand any of it because it's in a foreign language.

• I'd start by grabbing a chunk of text and put it into Google Translate. If it was impermissible (vandalism or promotional for example) I'd act accordingly - I once did this with a promotional user page in Arabic script (advertising a car dealership IIRC). Otherwise, once I'd identified the language, I'd compare with the relevant foreign-language Wikipedia - if the article was identical (or almost identical) to that, I'd nominate per A2. If the text was substantially different, I'd flag for translation.

Y Excellent. The key here was realizing that A2 only covers articles that are available on the relevant foreign-language Wikipedia. Mz7 (talk) 20:09, 7 November 2018 (UTC)

Scenario 7

A user creates an article, but shortly after creating it, the same user blanks the article by removing all of its content.

• I'd probably leave a note on the article's talk page, pinging the user, asking them what their intentions were. If they wanted to delete the article, I'd nominate for G7; if they were still working on it, I'd leave it a while and give them chance to add new text. If they didn't get back to me after a few days, I'd have a look at the text and see if it was worth keeping - if it's a useful article, I'd revert the blanking, and maybe add to it a bit myself (so that it's no longer one person's work); if it wasn't good enough to bother with, I'd take the author's blanking as a request for deletion, and nominate per G7.

Y  Note: I suppose it's never wrong to be conservative and ask the user beforehand; however, if a user blanks an article that they themselves created, then WP:G7 allows you to assume that this is a deletion request right away (i.e. without waiting a few days) as long as there are no other major contributors to the article. Mz7 (talk) 20:09, 7 November 2018 (UTC)

Scenario 8

A new user creates a user page with nothing but the following content:

Jlakjrelekajroi3j192809jowejfldjoifu328ur3pieisgreat

How would this scenario be different if the page was created in a different namespace?

• On a new user's user page, I wouldn't do anything - I'd assume that they were just testing to see whether they could edit it. Even if they left it there, we allow quite a lot of freedom in what users can put on their user pages, and if this user wants to have gobbledygook they're not harming anyone.
• It would be different if they created an article saying this. That would be CSD G1 - patent nonsense.

Y Good, we are very lenient about what we can have on user pages – this includes random gibberish. An article with this content, however, would be speedily deleted. Mz7 (talk) 20:09, 7 November 2018 (UTC)

Emergencies

I hope this never happens, but as you participate in counter-vandalism on Wikipedia, it is possible that you may come across a threat of physical harm. In the past, we have had vandals submit death threats in Wikipedia articles, as well as possible suicide notes. The problem is, Wikipedia editors don't have the proper training to evaluate whether these threats are credible in most cases.

Fortunately, there's a guideline for cases like this. Please read Wikipedia:Responding to threats of harm carefully and respond to the questions below.

Who should you contact when you encounter a threat of harm on Wikipedia? What details should you include in your message?

• I'd e-mail emergencywikimedia.org, with details of the threat and a link to the diff. I'd then contact admins via the IRC channel, again giving them the diff, and leave them to deal with it appropriately. If I personally felt threatened, I'd call the police and tell them what was happening.

Y Mz7 (talk) 21:01, 14 November 2018 (UTC)

What should you do if an edit looks like a threat of harm, but you suspect it may just be an empty threat (i.e. someone joking around)?

• I acknowledge that the link says that we are not trained to distinguish between credible and non-credible threats, and so in theory I should report any such threat as described above, regardless of my assessment about whether or not it's empty. In practice, I'd use a little discretion. If it was a threat of self-harm, I would always follow the procedures outlined above; if it was a random vandal saying 'I'm gonna fuck you up if you delete this', I'd probably just ignore it unless I had reason to suspect they had some means of knowing who I was. I'd always err on the side of caution, of course.

Y Fair enough. I've definitely ignored things that could have been threats, but were just random vandals. Use your best judgment and if you're not sure, err on the side of reporting, is the advice I would give. Mz7 (talk) 21:01, 14 November 2018 (UTC)

Usernames

Wikipedia has a policy which details the types of usernames which users are permitted to have. Some users (including me) patrol the User creation log to check for new users with inappropriate usernames. There are four kinds of usernames that are specifically disallowed:

• Misleading usernames imply relevant, misleading things about the contributor. The types of names which can be misleading are too numerous to list, but definitely include usernames that imply you are in a position of authority over Wikipedia, usernames that impersonate other people, or usernames which can be confusing within the Wikipedia signature format, such as usernames which resemble IP addresses or timestamps.
• Promotional usernames are used to promote an existing company, organization, group (including non-profit organizations), website, or product on Wikipedia.
• Offensive usernames are those that offend other contributors, making harmonious editing difficult or impossible.
• Disruptive usernames include outright trolling or personal attacks, include profanities or otherwise show a clear intent to disrupt Wikipedia.

Please read WP:USERNAME, and pay particluar attention to dealing with inappropriate usernames.

Describe the what you would about the following usernames of logged in users (including which of the above it breaches and why). If you need more information before deciding what to do, explain what more you need.

DJohnson

• There an existing user called Djohnson. No reason to suspect that this was intentional however, so rather than report to UAA as a misleading username, I'd leave a note on their talk page, mentioning the problem and giving them the link to WP:UNC.

Y I didn't even realize that. Good catch – the key here was realizing that real names are allowed. Sometimes, editors mistakenly report these as violations, which they are not. Keep in mind, if the editor claims to be a famous person, e.g. Dwayne Johnson, then we may have to ask them to verify their identity to prevent impersonation – see WP:REALNAME. Mz7 (talk) 21:07, 14 November 2018 (UTC)

LMedicalCentre

• Report to UAA - promotional, implies shared use.

Y But make sure you look at their edits first. If they haven't edited, then there's no need to report. If their edits have nothing to do with any medical centre, perhaps a question should be asked before reporting. Mz7 (talk) 21:07, 14 November 2018 (UTC)

G1rth Summ1t

• Report to UAA - misleading, possible disruptive (impersonating me - cheeky devil)

Y heh heh Mz7 (talk) 21:07, 14 November 2018 (UTC)

JoeAtBurgerKing

• Nothing wrong with this - if it was just 'BurgerKing', it would imply shared use, but 'Joeat' identifies the individual. I might take a look at editing patterns, and see whether they are engaged in WP:PAID and/or WP:COI editing, but any action I took would be based on that behaviour rather than the name itself.

Y Mz7 (talk) 21:07, 14 November 2018 (UTC)

Mz7IsALoser

• Report to UAA - disruptive.

Y Mz7 (talk) 21:07, 14 November 2018 (UTC)

JoeTheSysop

• Report to UAA - misleading.

Y Mz7 (talk) 21:07, 14 November 2018 (UTC)

😜

• This violates WP:NOEMOJI, but no reason to suspect bad faith - I'd drop them a note with a link to the policy and to WP:UNC.

Y Right, the key here is realizing that WP:NOEMOJI requires discussion at at WP:RFCN prior to blocking. Mz7 (talk) 21:07, 14 November 2018 (UTC)

@Mz7: - OK, I think I'm done with these answers - ready for feedback, or next set of exercises, when you are. Thanks! GirthSummit _(blether) 14:06, 4 November 2018 (UTC)

Watchlists

@Mz7: Hi, thanks for the feedback you've given me so far - I take the point you made about the high school kid's notability, I was probably being a bit conservative there - I didn't realise that SoundCloud was a self-publishing service, I should have done a bit of research.

Thanks also for the tip on installing Java - I've now got Stiki working, and I've been using it a bit; I've also been trying to get to grips with Huggle, although I managed to accidentally revert an experienced editor's good edit yesterday, and template them at the same time (I have apologised profusely), so I'm going to go slow with it and read the instructions a bit more carefully.

I've got a question about watchlists that I'd like your take on if possible. Since I started doing anti-vandalism work, I've been watchlisting articles when I do a revert. This is useful obviously, since vandals often reinstate their edits (or make new ones) on a page, but it means that my watchlist now has thousands of pages on it, and is practically useless for keeping an eye on articles about subjects I'm actually interested in. I can see that this is something that has been discussed many times before at the village pump, so I had a look at the watchlist customisation instructions, and at using the 'Related Changes' to give a sort of watchlist view, but obviously that's public, and would take me quite a while to set up (I had about 150 pages watchlisted previously).

It seems to me that the option to have a second watchlist would be very useful; or, alternatively, the option to have Stiki/Twinkle/Huggle add a page to your watchlist for a fixed period of time (e.g. 3 days), and then have it automatically removed unless you manually added it permanently. I can see many discussions about similar functions over the years, so I assume that there is a workable solution already that regular anti-vandalism editors use. Can you tell me how you manage this?GirthSummit _(blether) 12:21, 11 November 2018 (UTC)

@Girth Summit: Just want to give you a status update – a recent wave of real-life work has finally finished, and I should have enough time to dedicate to this course more thoroughly in the next few weeks. I'll leave responses to the current set of questions tonight. Mz7 (talk) 00:49, 14 November 2018 (UTC)

@Mz7: Thanks for all the feedback - much appreciated. Looking forward to the next section, and any tips about watchlists, when you have time. Cheers GirthSummit _(blether) 07:19, 15 November 2018 (UTC)

@Girth Summit: Hmm, I've mulled over your watchlist question for a few days, and I'm afraid I don't have a very good answer. To my knowledge, there is in fact no easy way to have multiple watchlists on Wikipedia, and there isn't really a tool I regularly use to monitor pages I'm interested in. I don't think STiki, Twinkle, or Huggle have the ability to temporarily add watchlist entries – I will note that this is one of the features on this year's Community Wishlist Survey, which determines what software the Wikimedia Foundation's Community Tech team will work on in 2019. If you are interested in such a watchlist improvement, I encourage you to support that proposal. I suppose you could create a second non-editing account if you want a second watchlist, but that also seems a bit tedious. Personally, I have more than 5,000 pages on my watchlist, and it can indeed get a bit tedious to sift through all of the entries. I apologize for this non-answer, but I'm really not sure. The watchlist is not the greatest feature of Wikipedia at the moment; there are definitely ways it could be improved. Mz7 (talk) 00:27, 18 November 2018 (UTC)

Thanks for your thoughts on this - I quite understand, it's just a problem that people have to deal with. I've added my support to the wishlist discussion, I do think that would be a good solution. I had considered creating a second account in order to get an extra watchlist, but I think that would just be too much of a headache, logging in and out all the time - I'll cross my fingers and hope that something gets done about it in the future, and if I get fed up I'll create a 'articles I'm interested in' page and try using related changes.

Dealing with difficult users

Occasionally, some vandals will not appreciate your good work and try to harass or troll you. In these situations, you must remain calm and ignore them. If they engage in harassment or personal attacks, you should not engage with them and leave a note at WP:ANI. If they vandalise your user page or user talk page, simply remove the vandalism without interacting with them. Please read WP:DENY.

Why do we deny recognition to trolls and vandals?

Because many vandals are seeking a way to get recognition; if we engage with them, we are giving them what they crave, and actually encouraging them to create more disruption and get what they are looking for. Instead, we want to make it clear that what they are doing is not a big deal, it's easily undone, and they are not going to get a personal response from us for their efforts.

How can you tell between a good faith user asking why you reverted their edit, and a troll trying to harass you?

That's a good question - I didn't see any specific guidance on this in WP:DENY, so I'll base this response on my own thoughts - if there is guidance that I should have read, please let me know.

First of all, I'd think about what their edit was that I reverted. If it was obviously and definitely bad faith vandalism, then it's probably not worth getting involved in a discussion with them - although I've come across exceptions to this.

Next, if they haven't reinstated their edit, but have come to my talk page or the article talk page to discuss it, I'd take that as a sign of good faith.

I'd also look for civility - even if someone reinstates their edit, but asks a questions civilly or makes a reasoned argument for why they have made their edit, I'd take that as a sign of good faith and be willing to engage with them.

If they are abusive, or come out with 'I can do this all day' or whatever, then that's a pretty obvious sign of bad faith, and I simply wouldn't engage with it - report to AIV, and revert edits if necessary.

There are quite a few sections (eg: here, here, here, here, here, here, here and here) on my talk page at the moment where I've had conversations with people I've reverted. There have also been a couple of instances of vandals trolling me on the talk page, with threats and with insertions of pornographic images (if you want an example, see the contributions made by this user, although I guess I should warn you that the video they were posting is explicitly sexual. I was going to ask for a revdel on it, but since it's just on my talk page history I haven't bothered - I'd be happy for you to do so if it's not too much trouble). In situations like this, I haven't engaged with the vandals, I've just reported and deleted them.

I should also note that there have been a couple of times when users have correctly pointed out mistakes I have made in my reverts: once it was a total accident when I started using Huggle and clicked on the wrong thing; and another time, when I had not properly read a sentence, and mistook an actual improvement for a grammatical error. I know that I'm fallible, and will always read what they have to say and engage with reasonable comments on my talk page, unless they are obviously just trying to start a fight. GirthSummit _(blether) 17:22, 25 November 2018 (UTC)

@Mz7: Thanks for the advice above; latest answers ready for review. GirthSummit _(blether) 09:04, 24 November 2018 (UTC)

Hi @Mz7:, just a note to say that I've thought a bit more about the last round of questions, and have changed/added to my answers a bit; I think I'm ready for the next round of questions, whenever you are. Cheers GirthSummit _(blether) 20:08, 29 November 2018 (UTC)

@Girth Summit: Thanks, I'll take a look tonight. Mz7 (talk) 21:22, 29 November 2018 (UTC)

@Mz7: Hi, just a quick reminder about the latest set of answers/next set of questions. No problem if you're too busy at the moment, just wondered if it might have slipped off your radar. Cheers GirthSummit _(blether) 17:57, 4 December 2018 (UTC)

Girth, I can only apologize. I know I told you I'd take a look "tonight", but it's clear I had a rather loose definition of the term "tonight".

I agree with the majority of your answer! Certainly it can sometimes be difficult to tell whether someone is good-faith or a troll just trying to pull your leg and waste your time. Definitely look at the types of edits that the user was trying to make – pornography is quite indefensible, so I would raise an eyebrow anytime a user tries to justify something like that without owning up to it as a mistake.

Although civility is important, I would keep in mind that it can be easy to get frustrated when one's good faith changes are reverted. I've interacted with many good-faith users who were livid that they had spent an hour trying to figure out how Wikipedia's syntax works, only to have their submission reversed. Just because someone is attacking you rudely does not necessarily mean they're a troll. All in all, I think more emphasis should be put on the quality and intention of the underlying edits, rather than just the tone and civility of the editor.

Regarding this course, I think we've covered a pretty wide breadth of what counter vandalism consists of on Wikipedia. We've touched recent changes, semi-automated tools, protection, deletion, oversight, revision deletion. I think what I will do next is give you a "final exam" of sorts, after which I think I'm ready to graduate you from this program. Please let me know if you have any questions. I'll probably post the exam this weekend. Mz7 (talk) 02:08, 7 December 2018 (UTC)

Thanks - I take the point about annoyed good-faith users potentially being less civil than a troll trying to wind you up - I'll make sure to keep that in mind. I'll look out for the exam questions, cheers. GirthSummit _(blether) 07:42, 7 December 2018 (UTC)

@Girth Summit: I posted the questions below. I think it should be a breeze for you! Mz7 (talk) 05:15, 10 December 2018 (UTC)

Final Exam

When responding to numbered questions please start your response with "#:" (except where shown otherwise - with **). You don't need to worry about signing your answers.

Part 1

For each of these examples, please state whether you would call the edit(s) described as vandalism or good faith edit, a reason for that, and how you would deal with the situation (ensuring you answer the questions where applicable).

1. A user inserts 'ektgbi0hjndf98' into an article. What would you do if it was their first warning? What about after that.

   I'd do an AGF revert, and give a level 1 'Test Edit' warning; if they did it again, I'd continue adding 'test edit' warnings, with increasing levels. If they continued despite multiple warnings, I'd report to AIV as disruptive.

   Y Great! Random gibberish is often a user trying to test the editing function, especially if it's the first time the user has done this. Mz7 (talk) 04:39, 18 December 2018 (UTC)

2. A user adds their signature to an article after one being given a {{Uw-articlesig}} warning. What would you the next time they did it? What about if they kept doing it after that?

   If they were doing it without adding useful content, I'd treat it as above with test edits. If they were adding useful content with sourcing, I wouldn't revert, but would edit manually to remove the signature but retain the content, and rather than using a warning template I'd attempt to engage with them on their talk page, explaining that I appreciate their content creation, but that signatures aren't used in article space. If they kept on doing it, and wouldn't engage on the talk page, I'd probably ask a friendly admin to review - I wouldn't want to stop them from adding useful content, and it's possible that they might not know how talk pages work, but ultimately WP:CIR, and the signatures would be disruptive.

   Y Mz7 (talk) 04:39, 18 December 2018 (UTC)

3. A user adds 'John Smith is the best!' into an article. What would you do the first time? What about if they kept doing it after that?

   Depends on what the article is about - if it's about John Smith, I'd do an AGF revert, with an edit summary along the lines of 'Unsourced subjective opinion', and give them a level 1 warning (escalating as necessary) for NPOV. If the article had nothing to do with any John Smith, it would be hard to see a good faith reason for adding it, and so I'd probably treat as vandalism - vandalism revert, level 1 vandalism warning (or higher if already warned).

   Y Excellent Mz7 (talk) 04:39, 18 December 2018 (UTC)

4. A user adds 'I can edit this' into an article. The first time, and times after that?

   As above with the 'ektgbi0hjndf98' - good faith revert and test edit warnings.

   Y Mz7 (talk) 04:39, 18 December 2018 (UTC)

5. A user removes sourced information from an article, with the summary 'this is wrong'. First time, and after that? What would be different if the user has a history of positive contributions compared with a history of disruptive contributions?

   Having recently been involved in this issue, I'd be careful about this. I would check the sources, to see whether they were reliable, and whether they supported the assertions that had been removed - paying particular care in a BLP article. If the content and sources were good, I'd AGF revert the user, and give them a 'Removal of content/blanking' warning, escalating the warning levels if it continued and ultimately taking to AIV if necessary. If the content/sourcing was problematic, I would do a dummy edit, and leave an edit summary explaining in more detail why this was a good removal of content, in case another change patoller came along and didn't check carefully. I'd then try to keep an eye on the page to see if anyone comes along to edit war the content back onto it. It shouldn't really make a difference whether it was an experienced editor or a random IP/new account that did the removal - as I've just learned, even a new account removing content without proper edit summaries and making legal threats may actually have legitimate concerns about the content.

   Y Yeah absolutely. Back in my early days of editing I had a similar experience. "Unexplained removal of content" can be good-faith, and we should be careful to examine the edits closely before making a vandalism accusation. Mz7 (talk) 04:39, 18 December 2018 (UTC)

Part 2

Which templates warning would give an editor in the following scenarios. If you don't believe a template warning is appropriate outline the steps (for example what you would say) you would take instead.

1. A user blanks Cheesecake.

   subst:uw-blank1

   Y Mz7 (talk) 04:39, 18 December 2018 (UTC)

2. A user trips edit filter for trying to put curse words on Derek Jeter.

   subst:uw-attempt2 (I'm slightly confused by this one - why don't we have a level 1 warning for this? If they had successfully vandalised, they'd have got a level 1 vandalism warning, but a failed attempt gets a level 2?)

   Y Not sure why this is. Mz7 (talk) 04:39, 18 December 2018 (UTC)

3. A user trips edit summary filter for repeating characters on Denis Menchov.

   subst:uw-efsummary

   Y Mz7 (talk) 04:39, 18 December 2018 (UTC)

4. A user puts "CHRIS IS GAY!" on Atlanta Airport.

   subst:uw-vandalism1

   Y Mz7 (talk) 04:39, 18 December 2018 (UTC)

5. A user section blanks without a reason on David Newhan.

   subst:uw-blank1

   Y Mz7 (talk) 04:39, 18 December 2018 (UTC)

6. A user adds random characters to Megan Fox.

   subst:uw-test1

   Y Mz7 (talk) 04:39, 18 December 2018 (UTC)

7. A user adds 'Tim is really great' to Great Britain.

   subst:uw-vandalism1

   Y Mz7 (talk) 04:39, 18 December 2018 (UTC)

8. A user adds 'and he has been arrested' to Tim Henman.

   subst:uw-biog1

   Y Mz7 (talk) 04:39, 18 December 2018 (UTC)

9. A user blanks Personal computer, for the fifth time, they have had no warnings or messages from other users.

   Straight to 4im - uw-delete4im

   Y Mz7 (talk) 04:39, 18 December 2018 (UTC)

10. A user blanks Personal computer, for the fifth time, they have had four warnings including a level 4 warning.

    Report to AIV using Twinkle for vandalism after a level4 warning

    Y Mz7 (talk) 04:39, 18 December 2018 (UTC)

11. A user blanks your userpage and replaced it with 'I hate this user' (you have had a number of problems with this user in the past).

    Depending on the number of problems I'd had in the past, and whether they'd been warned already, I might either subst:uw-harass4im (or other level if I thought it was appropriate), or report directly to AIV with a note about the harassment in the report.

    Y WP:ANI is a suitable alternative, particularly if this user has made good-faith edits. Mz7 (talk) 04:39, 18 December 2018 (UTC)

12. A user adds File:Example.jpg to Taoism.

    I'm in two minds about this. From the nature of the image in question, I'd interpret this as a test edit rather than deliberate vandalism, so I might use subst:uw-test1. I see that there are also 'image vandalism' warnings (subst:uw-image1 etc.), but I'm not so sure since this doesn't appear to be deliberate vandalism. The level 1 one linked above is pretty polite, so that's probably the best one to use in the first instance. (Ideally, an 'image test edit' warning might be better than image vandalism, but there doesn't seem to be one of them).

    Y Yeah, I would also see this as a test edit more than "image vandalism". I think both warnings are okay for the reasons you described. Mz7 (talk) 04:39, 18 December 2018 (UTC)

Part 3

What CSD tag you would put on the following articles (The content below is the article's content).

1. Check out my Twitter page (link to Twitter page)!

   G11 - db-spam - blatant promotion

   Y Mz7 (talk) 04:42, 18 December 2018 (UTC)

2. Josh Marcus is the coolest kid in London.

   A7 - db-person- no indication of notability of this person.

   Y Mz7 (talk) 04:42, 18 December 2018 (UTC)

3. Joe goes to [[England]] and comes home !

   Might depend what the article title was - it might be an attempt to describe the plot of a film, novel or whatever; if so, it would need expansion/improvement rather than speedy deletion. Otherwise, A7 db-event- no indication of notability of this event.

   Y Another possibility is WP:A1 – insufficient context to identify the subject of the article. Who is Joe? Mz7 (talk) 04:42, 18 December 2018 (UTC)

4. A Smadoodle is an animal that changes colors with its temper.

   A11 - db-invented - obviously invented - I Googled smadoodle, and there are some people using it as a twitter/instagram handle, but I don't see any use of it with anything like this meaning.

   Y Mz7 (talk) 04:42, 18 December 2018 (UTC)

5. Fuck Wiki!

   G3 - db-vandalism - pure vandalism.

   Y Mz7 (talk) 04:42, 18 December 2018 (UTC)

Part 4

Are the following new (logged in) usernames violations of the username policy? Describe why or why not and what you would do about it (if they are a breach).

1. TheMainStreetBand

   Promotional, per ORGNAME, and implies shared use, per ISU. I'd look at contributions first - if they had been editing about their band (or indeed about other bands - they might be dissing the competition!) I'd report to UAA; if not, I might just drop them a welcome-COI, or a personal message about username policy.

   Y Excellent. It depends on whether they've made edits—we don't typically block disruptive usernames unless there are edits (with the exception of obvious hate speech or harassment). Mz7 (talk) 04:47, 18 December 2018 (UTC)

2. Poopbubbles

   Kind of borderline. It might be seen as offensive, but it's fairly innocuous - I'd probably look at the edits and see what they were doing. If they were vandalising, I'd report them for that, and point out that the username doesn't inspire faith that they are WP:HERE; if they were doing good work however I'd just ignore it.

   Y Mz7 (talk) 04:47, 18 December 2018 (UTC)

3. Brian's Bot

   Misleading, per MISLEADNAME - implies that it's a bot account. Report to UAA.

   Y Mz7 (talk) 04:47, 18 December 2018 (UTC)

4. sdadfsgadgadjhm,hj,jh,jhlhjlkfjkghkfuhlkhj

   This is long and difficult to read, so probably violates UNCONF. I'd look at the editing history, and see what they were up to. If they were vandalising, I'd report them for that; if they were doing good work, I'd drop them a friendly note with a link to the policy, and suggest that they consider changing their username, or at least create a signature that is more readable, then if they didn't respond I might raise it at WP:RFCN.

   Y Mz7 (talk) 04:47, 18 December 2018 (UTC)

5. Bobsysop

   Misleading - implies that it's an admin. Report to UAA.

   Y Mz7 (talk) 04:47, 18 December 2018 (UTC)

6. 12:12, 23 June 2012

   This would be confusing because it obviously looks like the timestamp that gets appended to your signature when you sign a comment - that would make it very confusing trying to read a discussion they were involved in. Clearly violates UNCONF. I think this case is worse than the random string of characters above - in the previous case, it could be someone who just mashed the keyboard rather than make up a practical name, but copying our timestamp format so precisely looks like an intentional attempt to confuse - I'd probably take it directly to WP:RFCN.

   Y Mz7 (talk) 04:47, 18 December 2018 (UTC)

7. PMiller

   Nothing wrong with that, leave them to it.

   Y Mz7 (talk) 04:47, 18 December 2018 (UTC)

8. OfficialJustinBieber

   This is probably misleading per IMPERSONATE, and so I would report to UAA. If it turned out to be the actual account of the Canadian songster, he would be able to get unblocked by emailing info-en@wikimedia.org; in that case I would be happy to apologise to him for the inconvenience.

   Y Mz7 (talk) 04:47, 18 December 2018 (UTC)

Part 5

Answer the following questions based on your theory knowledge gained during your instruction.

1. Can you get in an edit war while reverting vandalism (which may or may not be obvious)?

   Not if you're doing it properly; or perhaps I should say yes, if you're not careful to avoid doing so. Reverting obvious vandalism is exempt from 3RR, and there are further exemptions for BLP violations, and so stopping someone adding obvious, unambiguous vandalism or unsourced defamation would not be considered edit warring. However, with more subtle stuff, you would need to be more careful - take time to check what they're doing and any sources they're bringing, and be very sure that it is deliberate vandalism before performing multiple reverts. If in doubt, get more eyes on the page, perhaps by posting on an associated Wikiprojects page, or even in an extreme case going to ANI. It's quite easy to get caught up in the heat of the moment if someone is adding content that you are sure is wrong or unnecessary, and you should never forget that someone misinterpreting our policies, or even willfully ignoring them, is not necessarily the same as someone deliberately vandalising an article. The world isn't going to end if something you suspect might be vandalism, but are unsure about, stays on the page a bit longer until you can get others to agree that it should be removed by consensus.

   Y Great thoughts. It is true that vandalism is one of the specific exemptions to the edit warring policy, but it is important to realize that more subtle vandalism might not always be thought of as exempted. Restrict yourself to applying to exemption on obvious vandalism, and escalate more subtle cases where appropriate. Mz7 (talk) 04:52, 18 December 2018 (UTC)

2. Where and how should vandalism-only accounts be reported?

   I would generally use Twinkle to report them to AIV.

   Y Mz7 (talk) 04:52, 18 December 2018 (UTC)

3. Where and how should complex abuse be reported?

   ANI is for 'chronic, intractable' problems, so that might be an alternative to AIV. I would notify the user that I was reporting them there if it came to it.

   Y Mz7 (talk) 04:52, 18 December 2018 (UTC)

4. Where and how should blatant username violations be reported?

   I use Twinkle to report them to UAA.

   Y Mz7 (talk) 04:52, 18 December 2018 (UTC)

5. Where and how should personal attacks against other editors be reported?

   According to NPA, for first offenses or isolated incidents, it might be best to ignore personal attacks, but if it is a recurrent problem you can report it at AN (or, presumably, ANI for urgent cases).

   Y I would say WP:ANI is the better venue for this kind of report. WP:AN is for more general discussions, like announcements and bans, but in practice, I think both venues would result in the same response. Mz7 (talk) 04:52, 18 December 2018 (UTC)

6. Where and how should an edit war be reported?

   Assuming I had already warned the warring users, and attempted to reach resolution at the article talk page, I would use Twinkle to submit a report to AN3.

   Y Mz7 (talk) 04:52, 18 December 2018 (UTC)

7. Where and how should ambiguous violations of WP:BLP be reported?

   You could start a discussion at WP:BLPN, and put the {{BLP noticeboard}} tag on the article talk page. Alternatively, if it was an urgent situation, you could presumably get eyes on it quickly at WP:ANI.

   Y Mz7 (talk) 04:52, 18 December 2018 (UTC)

@Mz7: Hi, I think I'm done with my exam questions - look forward to your feedback when you're ready. Cheers GirthSummit _(blether) 17:48, 10 December 2018 (UTC)

Completion

Congratulations from both myself and all of the instructors at the Counter Vandalism Unit Academy, on your successful completion of my CVUA instruction and graduation from the Counter Vandalism Unit Academy. You completed your final exam with 100%. Well done!

As a graduate you are entitled to display the following userbox (make sure you replace your enrollee userbox) as well as the graduation message posted on your talk page (this can be treated the same as a barnstar).
{{User CVUA|graduate}}:

This user is a Counter-Vandalism Unit Academy graduate.