User talk:Wanderer57/Feedback re warning vandals

This is copied from AIV talk page, Feb 4, 08

Sick and tired of "soft" administrators
We editors fight the vandalism ongoingly. When a user (an IP user at that!) logs onto Wikipedia ONLY to vandalise, has been given level 4 warnings ongoingly, has already been blocked for vandalism, and an admin takes a "soft" approach to new vandalism by this editor, what are we to do? The editor has already established himself as a persistent vandal, but some admin brushes off the report and blanks it from the report page. What are we to do? This site would be taken over by vandals if not for the tireless, ongoing work of dedicated editors and like minded admins. Seems like some of the admins want to mire themselves in deep, never ending discussions of etiquette and proper procedure. Well, the proper procedure is spelled out for you! Anyone who vandalises after many warnings and a block should summarily be blocked automatically when he vandalises again. Right?

If we can't get support from the admins, why should we edit any more? We should just give up. Let the vandals take over the site.

I would like to see an "anti-peer" review of admins by all Wikipedia editors, and "soft" admins should be demoted. I bet that would get them on their job. -- Elaich   talk 00:31, 3 February 2008 (UTC)
 * I consider myself a "soft" admins, and no threats will make me block recurrent vandals without reviewing their cases one at the time. Feel free to create this "anti-peer" review and get me demoted. -- ReyBrujo (talk) 01:56, 3 February 2008 (UTC)
 * This is a nice change from complaints about block-happy admins, at least. My advice to you is to find one of those complaints (its not hard) and direct your request to the subject of that complaint in future. Rockpock  e  t  02:00, 3 February 2008 (UTC)
 * Well, this is a new change. I consider myself a hard admin, but I think an anti-peer review would be detrimental. After all, if we purge our admins, how many more people do we get to help fight vandalism? We only lose more admins.  bibliomaniac 1  5  02:04, 3 February 2008 (UTC)
 * Didn't you hear? If all softie leave, the only admins that will stay are the ones who will block on sight. Vandalism will be handled faster that way!
 * Sarcasm aside, the only way he will understand is becoming an admin and having to deal with vandals himself. Getting pissed because you reported a vandal and was not blocked is somewhat childish, but that is my opinion. -- ReyBrujo (talk) 02:12, 3 February 2008 (UTC)
 * I consider myself to be pretty quick to hand down blocks, but I'm reticent to give out long blocks to IPs. All too often, there's collateral damage, and we end up blocking some sweet grandmother (which we then hear about on the unblock mailing list). Call me soft if you want, but I'm a stickler for assuming innocence until proven otherwise and I will restrict myself from significant time blocks against IPs unless I know it's static. - Philippe &#124; Talk 02:41, 3 February 2008 (UTC)


 * (e/c; I type slowly) Hi Elaich, something in the phrasing of your comments leads me to think you believe one IP = one person, or at least believe it's usually that way. These days, I imagine it's the opposite; many people are editing from home (almost always a dynamic IP), the library (shared IP), wifi networks (shared), school (shared), etc.  No idea what the proportion is, but I imagine it is a significant majority.  The reason we start from ground zero after a few days or weeks of the last warning is that it could very easily be a new person.  If you notice something about the vandalism that leads you to think it's the same person over and over, like vandalizing the same articles or the same "type" of vandalism, make a note of that in your report to AIV, and I think you'll be pleasantly surprised how often "soft" admins will agree with you and block.  And if you can't detect such a pattern, then consider the likelihood that the "soft" admins are right, and it really is a newbie that can be turned with just a warning. --barneca (talk) 02:51, 3 February 2008 (UTC)

I'm guessing the original message was in response to this block refusal. Frankly, I agree with the admin in that case. Since a three-month block on October 3, 2007 expired (which I'm not sure I would have done in the first place - it was only the fourth block), the IP had sat dormant for more than two weeks before a test edit on January 21, 2008, a possibly legitimate edit two days later, and two vandalism/test edits six days after that. Then, you didn't report the IP until two days after the last edits. Four questionable edits in three weeks are hardly worth the enormous six-month block you're campaigning for. Folks need to remember that vandalism is not going to be stopped no matter how quickly or how hard we block. It can only be slowed down from time to time. We need to tread a fine line between stopping silliness and losing useful edits because we've blocked an IP unnecessarily. We should be erring on the side of blocking too softly because nothing is worse than losing good contributions - or, God forbid, good editors entirely - because someone sits at a public computer and are unable to edit. If you are so frustrated by fighting vandalism, I recommend ignoring the recent changes screens and your watchlist for a while. Just add your own contributions and ignore everything else. I've done it numerous times and it works wonders. —Wknight94 (talk) 03:25, 3 February 2008 (UTC)
 * Endorse and agree with wknight's summary above. - Philippe &#124; Talk 03:43, 3 February 2008 (UTC)
 * If vandalism comes from an IP and warnings are issued, and then a few days later another incident of vandalism comes from the same IP but on different articles, I am going to (with good faith) assume that the second editor is not the same person as the first editor. If the first incident got to a level 3 or 4, I might start at level 2 for the 2nd editor. Sure, we may learn that the 2nd editor is the same as the 1st editor, but I must assume good faith. I don't think that is soft. Kingturtle (talk) 03:48, 3 February 2008 (UTC)

"We should be erring on the side of blocking too softly because nothing is worse than losing good contributions - or, God forbid, good editors entirely - because someone sits at a public computer and are unable to edit."

Even someone who sits at a public computer is able to register and use an account. Correct?

From IP editors, I see maybe 1% useful edits and 99% outright vandalism.

Anyone is able to register for an account. Those who don't or won't usually have many reasons not to.

My argument is that IPs should be admined HARD. There is simply no reason not to. IPs are almost always vandals, in my experience. I don't sit around waiting for them to make some valuable contribution. There is no reason they cannot create an account unless...

In the case I submitted, the IP did not edit for months after a 3 month long block. Then, the IP logged on only to vandalise again. The fact that the IP logged on only to vandalise again after many weeks means nothing to me. The IP always vandalised, and proved true to form. If the IP logged on only after months or years, the IP has always vandalised. The IP should be treated harshly. If the IP address is not static (not likely, since the pattern of vandalism was always the same) then another user of the same IP could create a user account.

Fail, fail, and fail. The admin who brushed this aside was just to lazy to investigate. We sure as hell don't need lazy admins here. I would hope that anyone who has passed the admin test would be as viligent as I am regarding vandalism. -- Elaich   talk 07:38, 3 February 2008 (UTC)
 * If an IP is as vandalism-happy as you are suggesting, then giving two or three warnings before blocking will result in the same result; they will vandalise until they are blocked anyway. If not, then blocking isn't accomplishing anything, because it is intended to prevent harm to the encyclopedia, not punish the vandal for his edits. As for the quantity of legitimate anonymous edits, your experience does not seem to reflect mine. I've definitely seen more valid IP edits than I've seen vandalism by them. In fact, I believe I remember hearing that the majority of our content comes from IPs. I think the possibility of driving potentially good contributors most definitely qualifies as a good reason to treat them the same as we always have.-- Dycedarg  &#x0436;  08:29, 3 February 2008 (UTC)
 * What Dycedarg said. Your prejudice against IPs is not well-founded and is not supported by any studies that have been done here.  Also, most admins - myself included - disable account creation when they block IPs.  That means that people cannot create new accounts while the IP is blocked.  Otherwise, the real habitual vandals would simply create an account and continue vandalizing.  You're barking up the wrong tree with your line of reasoning here.  —Wknight94 (talk) 14:21, 3 February 2008 (UTC)

In response to "From IP editors, I see maybe 1% useful edits and 99% outright vandalism," you'll be happy to hear that IP edits usually have good intentions. I looked at 25 or so consecutive IP edits just now and saw that 73% of them had good intentions:

edits with good intentions
, ,, , , , , , , , , , , , , , , , 

vandalism
,, , ,

tests
,

Just to set the record straight on "maybe 1% useful edits and 99% outright vandalism." Kingturtle (talk) 17:13, 3 February 2008 (UTC)
 * The road to hell is paved with good intentions. ;) Kafziel Take a number 17:22, 3 February 2008 (UTC)


 * Even if the ratio of bad to good ip editors is 99 to 1 then ip editing works per this. As shown above the ratio is nothing like that. LessHeard vanU (talk) 18:04, 3 February 2008 (UTC)
 * I did a survey myself a while ago - User:Hut 8.5/anon edits. Hut 8.5 18:10, 3 February 2008 (UTC)


 * "Even someone who sits at a public computer is able to register and use an account. Correct?" No.  If the IP is blocked, we usually block account creation as well.  In order to edit, they would have to apply for a third party to create an account for them, and wait for it to be done.  If they're at a public computer (say in a library or internet cafe), chances are their time on the computer is limited, so the notion of trawling through the help and WP pages to find the procedure, follow it, and then wait for a response - that's hardly going to appeal to many people.  We don't want hard or soft administrators, we want fair administrators. Waggers (talk) 22:22, 3 February 2008 (UTC)


 * Amen, brother, amen.  bibliomaniac 1  5  22:41, 3 February 2008 (UTC)

Soft on Vandals?
I think we are "soft" on vandals in some ways.

Problems I see often are:

1) Cluebot is such a sweet-natured and forgiving bot. She leaves gentle messages, even in cases where the previous note in the talk page was another warning or a block.  (for example, User talk:24.79.141.149)

Maybe Cluebot could look back a month or two and go to a higher number warning if there are already warnings or blocks in the recent past.

2) The welcome/warning message is way too long. It doesn't read as a serious warning.  Maybe shorten the overall message, and put the welcome in one paragraph and the warning in another.

3) Editors often fix vandalism but do not leave a warning. Someone can vandalise a lot and get few warnings.  The problem is it is much easier to revert vandalism than to leave a warning.

4) If I know that a contribution record is "pure" vandalism, I will mention that in an AIV report. Problem is, it's a lot of work to check each contrib.  How about keeping a statistic for each editor of the percentage of their edits that are reverted on the next edit???

Wanderer57 (talk) 16:19, 4 February 2008 (UTC)
 * For your 1), that's exactly what I don't want to happen. If there is vandalism from an IP from months - or even days - ago, the chances of the latest vandalism being the same person are very small.  If random people keep sitting at a shared computer and make little test edits, there is just as much likelihood that the next person will sit down and add good contributions or start a brand new account, and maybe even become the next Newyorkbrad!  We need to stop treating individual IP addresses as though they are individual people (unless there is an indication that it is the same person every time).  Warnings to stupid kids in November shouldn't cause a block to someone trying out Wikipedia for the first time with a "hi" edit in February.  —Wknight94 (talk) 16:57, 4 February 2008 (UTC)
 * I agree with Wknight94 about #1; far too many IPs are shared, dynamic, or connected to a public computer to treat them all as individual people. Especially for a bot to do so, which can't make any semblance of a judgment call. As far as #2 goes, if you are referring to uw-vandalism1, it isn't supposed to be a serious warning. It's supposed to be a hint for clueless vandals that what they're doing isn't approved of, and a pointer to places where they can learn to contribute constructively should they wish. For obvious cases, you don't have to use it. I for one don't use it particularly often. uw-vandalism2 is plenty kind for a first warning in the case of a vandal who's obviously vandalizing on purpose. You're supposed to use your best judgment when it comes to giving out the warnings, and give whichever one you feel is appropriate for the situation. And there are occasionally situations that call for something that gentle. I agree that #3 is a problem, which is why I like to use Twinkle to automate such things so I don't forget. I think the only thing we can do about that is to gently remind people who frequently revert vandals without warning them that the warnings are at almost as integral part of the process as the reversion itself. If you don't warn them than they'll never stop because admins won't block them until they've been warned. As for #4, I don't think it's necessary enough to be worth the amount of work it would take; you can generally judge how prolific a vandal is by their block log and talk page history. For vandal-only registered accounts, they generally get blocked before their contribution history is long enough to make going through it a problem.-- Dycedarg  &#x0436;  17:08, 4 February 2008 (UTC)


 * Very small ? I'd like to see some figures to back that up. -- John (Daytona2 · Talk ·  Contribs) 17:28, 4 February 2008 (UTC)
 * WP:AGF says I don't need to give figures to back that up. I have to assume that my statement above is true.  This may be a different point but I see a lot of people expecting me to block an IP for 31 hours when the IP only edits every two months.  Even if the IP is the same person, what good is that blocking strategy?  —Wknight94 (talk) 18:35, 4 February 2008 (UTC)


 * I agree. I get the impression that there is a low regard for people spending time and effort reporting vandalism. It's easier for a vandal to vandalise than it is for someone to report them. That's a critical flaw. Repetitive warnings that "you will be blocked" prove the opposite, and I no longer bother to add warning in such circumstance, I can only hope that an admin searches the posting history for themselves. The warning templates do not make it clear that you do not have to start at level 1, this is confirmed by number of level 1 warnings that people give. The proportion of vandalism coming from educational institutions, and the lack of effective action, is utterly dispiriting. -- John (Daytona2 · Talk ·  Contribs) 17:28, 4 February 2008 (UTC)
 * On the warnings front, I'm not as picky about warnings as other admins. If someone has gotten a couple warnings and is clearly continuing unabated, I will block.  I don't count vand1-vand2-vand3-vand4.  I've blocked without warnings before - there I said it!  Not often but sometimes.  Esp. if a particular article appears to be the target of a coordinated attack.  But, to my point above, I don't block for gigantic lengths of time, esp. in those cases.  If instinct tells me that a one-hour block will be enough to make the jackass kid get up and find something better to do (go outside and play), I'll do it.  Conversely, if a school IP has been blocked 13 times already and is obviously host to many future petty criminals, I'll block as close to forever as I feel comfortable - six months, even a year sometimes.  The point is simply to stop vandalism as much as possible while not preventing any good edits or new good editors.  If that means some of your reports go unheeded, don't take it so personally.  If you insist on taking it personally, get away from the Recent Changes screens for a while and find a different part of the sandbox to play in.  Those are my humble suggestions. —Wknight94 (talk) 18:27, 4 February 2008 (UTC)


 * In response to your assertion that the templates do not make it clear that you don't have to start at level 1: Actually, every single page related to vandalism work says that you don't have to start at level 1. It says right on Template messages/User talk namespace: "You are responsible for ensuring that the template's text is appropriate to the violation: if the template's tone isn't appropriate, don't use the template." This doesn't imply "Put the templates up in order regardless". It implies "Give warnings consistent with the severity of the offense." WP:VANDAL says "If you are not certain that an edit is vandalism, always start with uw-test1. Conversely, if you are confident that a user is aware of the disruption he is causing, you may start with a stronger warning such as uw-test2 or uw-test3." Same deal. Guide to administrator intervention against vandalism says "Vandals should always receive enough warnings before being reported. What constitutes "enough" is left to your best judgment. Consider the user's past edits, warnings and blocks, the severity of their offense, the likelihood that their edit(s) could have been made in error or otherwise in good faith, and the type of user in question (IP addresses may be shared or dynamic, and old warnings could be irrelevant to the current situation)." Where is this misconception that vandals should always receive all the templates in order before being reported or that some page somewhere states/implies that you should coming from? It isn't stated anywhere in any page related to this matter.-- Dycedarg  &#x0436;  18:36, 4 February 2008 (UTC)

We are all expected to assume good faith. Kingturtle (talk) 17:31, 4 February 2008 (UTC)


 * I know that I see and understand far far less of this problem than the administrators commenting here. My comments starting this section did not suggest administrators were "soft" on vandalism. I was looking at other issues, and trying to suggest solutions.  I do not take it personally if my report does not result in a block.  (though I do take it personally enough that I CHECK if a block resulted.)  Actually most of my reports have led to blocks.


 * I wonder about cases like the IP I listed above. Total career stats:  has edited on 14 articles, vandalized 13 of them, one looks legit (Rick St. Croix, May 2007‎).  Has received 17 warnings and 3 blocks, leaves the edit summary "Blanked the page."  What does "assume good faith" mean in this case?  Wanderer57 (talk) 18:48, 4 February 2008 (UTC)


 * Re this text quoted above: "Vandals should always receive enough warnings before being reported. What constitutes "enough" is left to your best judgment. Consider the user's past edits, warnings and blocks, the severity of their offense, the likelihood that their edit(s) could have been made in error or otherwise in good faith, and the type of user in question (IP addresses may be shared or dynamic, and old warnings could be irrelevant to the current situation)." How can I know if the IP is shared or dynamic, and how would that affect warnings?  Wanderer57 (talk) 18:57, 4 February 2008 (UTC)


 * (edit conflict) ClueBot was right in this case to give a fairly low-level warning. The block was more than a month earlier, and the IP is probably shared (it's registered to "Shaw Communications Inc." which sounds like an ISP to me). Therefore it it unlikely that it is the same vandal back again unless there is further evidence. We hand out blocks per person, not per IP. Suppose someone sees a mistake in a Wikipedia article, tries to edit the page to correct it and gets a big angry block notice accusing them of vandalism. They aren't going to come away with a positive impression of Wikipedia.
 * Another problem with getting ClueBot to give higher level warnings is that it isn't infallible. It reverts when it sees obvious things like blanking the page (the edit summary of "blanked the page" is automatic, the user didn't add it btw). It is possible that the blanking edit was made in good faith - if a newbie sees some content they don't agree with or is inaccurate, they may well remove it. Potential good faith edits aren't vandalism and should be given gentle notices rather than vandalism warnings. --Hut 8.5 18:59, 4 February 2008 (UTC)


 * I agree with Hut 8.5. In this case, the first edit could have been accidental, as it was merely a blanked page with an automatic edit summary. The second edit warranted a stricter warning than was initially given in my opinion, but not the first. Dealing with your second question, it is difficult sometimes to tell if an IP is shared or dynamic, but one can often tell by the variety of edits it makes. Sometimes people will find out through various methods and put one of the templates at the top of the page that tells you what kind of IP it is. If you don't know for sure, default to treating it as shared or dynamic: treat every case of fresh vandalism as it's own entity, and take histories of vandalism with a grain of salt. You don't have to ignore previous instances of vandalism altogether, and if the pattern of vandalism (such as if the methods of vandalism or pages edited are similar) you can safely assume that the person involved is the same, but otherwise try to give warnings appropriate to the immediate situation at hand. In my opinion it's just something you get a feel for; I don't think there's a way of writing down exactly what level of warning is appropriate to every situation.-- Dycedarg  &#x0436;  19:17, 4 February 2008 (UTC)


 * Thanks muchly for all the feedback. It is a lot clearer now.  Wanderer57 (talk) 19:36, 4 February 2008 (UTC)