UDWiki talk:Administration/Policy Discussion/Better Vandal Data

Opening

Let us use this space to develop our ideas for improving the current system for handling vandal data. We can all agree that the vandal data is important to the wiki and that it is suboptimal.

If you want to point out a specific problem or propose a solution, please create a new == header so that the discussion can remain orderly. Please refrain from editing this opening section.

In general, A/VD should achieve the following:

• Organization - Escalation history, edit history, and the location of users should be organized
• Ease of use - It should be easy to check the history of specific users
• Automation - Making escalation/de-escalation more automated, reducing errors.

--_GANG Giles Sednik _CAPD 13:47, 30 November 2009 (UTC)

remove deescalations

Get rid of deescalations? They only confuse the issue, and encourage career vandals -- boxy ^{talk • teh rulz} 13:27 30 November 2009 (BST)

^^^ Cyberbob  Talk  13:34, 30 November 2009 (UTC)

But I don't think anyone actually suggested that... --DANCEDANCEREVOLUTION-- 13:38, 30 November 2009 (UTC)

+1 --Haliman - Talk 13:40, 30 November 2009 (UTC)

Your intentions? At the moment, your intentions seem to be for the community to devise it for you. How righteous. Not reactionary and unfounded at all. --DANCEDANCEREVOLUTION-- 13:32, 30 November 2009 (UTC)

I am all for removing descalations if the maximum amount of time a user can remain banned is reduced to a week. After the fourth time the user is banned for a week sysops are allowed to start a simple perma vote on the user. --People's Commissar Hagnat ^{[talk] [wcdz]} 16:10, 30 November 2009 (UTC)

no Cyberbob  Talk  16:15, 30 November 2009 (UTC)

Surely the whole point of the policy, especially deescalation included is to reform vandals, how would the policy work without this? --Rosslessness^{Want a Location Image?} 16:43, 30 November 2009 (UTC)

I think it's pretty clear by this point that the de-escalation system does nothing to reform vandals - quite the opposite in fact, as it more or less allows them to keep on trucking indefinitely as long as they're careful about it. It's a messy, complicated and incredibly dramagenic system that really hasn't achieved much beyond the aforementioned enabling of smart longterm vandalism. Cyberbob  Talk  17:12, 30 November 2009 (UTC)

Hell, even if we're not careful you people suck at catching vandals anyway. I should have probably been perma'd by now!-- SA 22:56, 30 November 2009 (UTC)

If you look through A/VD you'll notice that the great majority of people "reform" (or leave) after a warning or two, and are no more trouble after that. The ones who are regularly deescalated have no intentions of changing, either because there is no incentive too, or because it's all a big game to them. It is also these people whose a/vd history becomes convoluted and confusing. Either striking escalations needs to be much rarer and harder to achieve, or it needs removing altogether -- boxy ^{talk • teh rulz} 02:25 1 December 2009 (BST)

It seems that keeping the users in check at that point, rather than reforming them, would be the idea. Simply lengthening the time between (or entirely removing) de-escalations would do that. Mandating X edits for a de-escalation also seems redundant, since "reformed" users will have escalations fall off over time regardless, while the ones gaming the system will either make the edits anyway, or will begrudgingly make spammy/low-quality edits to reach the goal. In either case, the edits are unnecessary. —Aichon— 06:01, 1 December 2009 (UTC)

To be honest... I don't think the point was to reform vandals. I know I did a lot of stupid shit I shouldn't have as a newbie.... and I see a lot of new people making mistakes to. I think it wasn't so much reform, as it was forgivness for new comers that was the point. -Poodle of Doom^{M! T} 02:14, 1 December 2009 (UTC)

a) Yes it was about "reform", and b) Newbie mistakes don't count as bad faith anyway. Cyberbob  Talk  02:33, 1 December 2009 (UTC)

Since we're out to reform the VB system,... how's it broke? Are we not "reforming" the vandals? If this were the case, it would seem like our warnings aren't doing justice. That being the case, we either need a stiffer punishment, or a new way or reformating people. -Poodle of Doom^{M! T} 03:11, 1 December 2009 (UTC)

Formalize deescalations

It's clear that de-escalations are a big problem with A/VD. As an alternative to removing them altogether, how about formalizing the process of de-escalation? I'm thinking of something like the speedy deletion cue. You would have various criteria for de-escalation.
Crit 1: time passed.
Crit 2: 250 edits
Crit 3: Vandalism ruling was misconduct, etc.
Then people would just submit their name, a link to the A/VB, and the criteria. This would keep de-escalations transparent, organized, and would provide a record of all that has transpired, regardless of history wipes. Also, it would ease the burden of the sysops by providing an avenue for dedicated wiki users to submit de-escalations as part of their wiki maintenance.--_GANG Giles Sednik _CAPD 23:21, 30 November 2009 (UTC)

We do have talk pages you know.-- SA 23:28, 30 November 2009 (UTC)

True, but they're not a central location where past actions can be easily checked. Oh, and I thought it was both 1 month and 250 edits, nor or 250 edits, just to point out one detail. —Aichon— 00:32, 1 December 2009 (UTC)

No Cyberbob  Talk  02:13, 1 December 2009 (UTC)

If I'm not mistaken, this is already part of the current system. I think a better question would be to ask what it is we are trying to reform. See below.... -Poodle of Doom^{M! T} 02:32, 1 December 2009 (UTC)

It is part of the current system, but as SA pointed out, de-escalation requests go through the talk pages of the sysops right now. There are a number of problems with this. Firstly, as Aichon pointed out, their talk pages are not a centralized data-base, so it can be confusing to track down the info regarding de-escalation. Secondly, there is the possibility that a de-escalation request can go overlooked, ignored, or forgotten. By formalizing the process we can prevent such errors and ensure that de-escalations are handled in a timely manner while creating a public record that will provide clarity in future de-escalations/bans.--_GANG Giles Sednik _CAPD 12:16, 1 December 2009 (UTC)

Or we could get rid of the whole concept for reasons which have been stated multiple times yet you seem to have ignored. Cyberbob  Talk  12:20, 1 December 2009 (UTC)

Separate Pages

Just a quick ideas to get us started. How about organizing vandal data into different pages alphabetically? So there would be "Administration/Vandal_Data/A", "Administration/Vandal_Data/B", and so on. And there would be an alphabetical listing on the main A/VD page. This would make it easier to track changes by checking the history of a given page, instead of having to pour through months of data to find the history of a specific user.

Similarly, the pages could be organized like encyclopedia volumes, Aichon - Boberton, Cyberbob - DDR, etc., since certain letters of the alphabet would have more users.--_GANG Giles Sednik _CAPD 13:47, 30 November 2009 (UTC)

i think i suggested once that we should have a separate page for each user, but people complained about the complexity of this. Store it on user:Username/Vandal Data and protect it so only sops can edit it after vandal cases are ruled on. It would make a hell lot easier to see how many times a user was brought to A/VB and how many warnings he got. Sure it add another layer of work, but at least we have an easy access to all this information --People's Commissar Hagnat ^{[talk] [wcdz]} 16:07, 30 November 2009 (UTC)

No then, no now. History wipes pretty much invalidate everything that an individual-page-policy has to offer. Please stop posting Hagnat, you're not particularly smart. Cyberbob  Talk  16:09, 30 November 2009 (UTC)

im not saying we should retroactively archive the vandal records... but starting on the moment this policy gets approved. And even we did retroactively, what history wipe have to do with all this ? Its a simple case of copying specific headers from the current A/VB history into another page... i think you are more worried about what this user-archive would reveal about yourself... --People's Commissar Hagnat ^{[talk] [wcdz]} 16:15, 30 November 2009 (UTC)

If it's not about page histories, what the hell is it about? The current system offers literally the same amount of info as a separate-page system once you remove the history factor. By the way: that hilarious attempt at poisoning the well is, well, hilarious. The gory details of my vandalism history are already where anyone can see them - here, I'll even link it for you. You're even dumber than I thought, to be quite honest. Cyberbob  Talk  16:20, 30 November 2009 (UTC)

Your vandal data record only shows WHEN you was warned, not WHY and all the other cases you were reported but not found guilty of vandalism. This later cases can be used by sysops to identify a trend and punish a user for constant idiocy or something. And i reckon i am not in my high moments, but i have booze and three months of party to blame... what is your excuse ? --People's Commissar Hagnat ^{[talk] [wcdz]} 16:30, 30 November 2009 (UTC)

It's a simple matter to add a quick "Impersonation" or "Page blanking" or whatever note next to a page entry without having to create such a stupendously more complicated system. It already happens to some extent - escalations from breaking arbitration rulings and from misconduct cases get their own notes to distinguish them from the others. Oh, and if you're drunk you really shouldn't be on the computer. That's just about the lamest thing I can possibly think of anyone ever doing short of becoming a camwhore. Cyberbob  Talk  16:35, 30 November 2009 (UTC)

Arbitration and misconduct warnings are noted because they dont actually count in your vandal data in vandalism cases. I can have a month long ban for misconduct, but if i vandalize - and i only had two warnings - all im supposed to get is a 24h ban. That is noted because a sysop can be banned for a period of time different than the one specified in his vandal data, and arbitration violations punish users like they had at least two warnings, even if the user had a clean records. And i stopped drinking and wikiing a loooong time ago... nowadays i only wiki when i get bored, and that only happens when i am sober :P --People's Commissar Hagnat ^{[talk] [wcdz]} 16:44, 30 November 2009 (UTC)

No fucking shit Sherlock - what was the point of going into such pointless detail? You answered absolutely nothing about why my suggestion about adding notes to regular escalation entries might be a bad idea (probably because you can't think of anything). Cyberbob  Talk  17:15, 30 November 2009 (UTC)

Breaking it into multiple pages merely decentralizes a system that should be centralized, and I'm afraid I don't see an immediate benefit. Currently, checking changes to a user's data seems relatively simple: just list the last 500 edits, run a text search, and then manually check any unlabeled ones. Iscariot's case is an odd one, to be sure, but the circumstances that led to it could occur with any system, so we shouldn't let it drive policy, necessarily. Plus...I don't have any vandal data (yet?). —Aichon— 16:27, 30 November 2009 (UTC)

"History wipes pretty much invalidate everything that an individual-page-policy has to offer." Hmm, that's a good point. How often do these history wipes take place?--_GANG Giles Sednik _CAPD 17:03, 30 November 2009 (UTC)

Every few months. It varies, I think. Cyberbob  Talk  17:18, 30 November 2009 (UTC)

Right, well in that case breaking the vandal data into separate pages wouldn't be helpful at all. Skip that idea lol.--_GANG Giles Sednik _CAPD 17:23, 30 November 2009 (UTC)

Watch - it'll take like two more days for Hagnat to drop it. Cyberbob  Talk  17:24, 30 November 2009 (UTC)

IIRC, there have been three history wipes in the whole UDWiki history. --People's Commissar Hagnat ^{[talk] [wcdz]} 17:44, 30 November 2009 (UTC)

I'm almost positive that you are wrong, and the frequency is irrelevant in any case. The fact that they can and do happen negates the utility of a system that is entirely dependent upon one not happening. Cyberbob  Talk  17:48, 30 November 2009 (UTC)

This may seem like a stupid quetion, but I ask because I truly do not know. Does the history wipe really affect a users list of contrabutions? Mine seems to be fairly intact... even though there were quite a few pages that were wiped that I had edited. It would seem to me that a users contrabutions would need to be wiped. I think as long as a users contrabutions aren't wiped, all should be well. Someone stated above just to compare the user contrabutions versus the date of the A/VB... I think this is fair, but had assumed this was already part of the current system. -Poodle of Doom^{M! T} 02:28, 1 December 2009 (UTC)

Last edit and link to the ruling

I've noticed two minor items that might help clear up future questions:

1. A link to the actual ruling mentioned in the data
2. A link to the last edit the user made before the ruling (i.e. where their "250 edits" count begins)

Basically, I'd suggest that each entry in (or modification to) the data contain a link to the ruling that led to that edit, rather than just a timestamp. So, if a user was warned, the "warned" text could be a link to the archived case, and similarly, if the ruling was overturned due to misconduct, the "misconduct" text could be a link to the archived A/M case. Even though it's possible to fact check using the timestamp, adding a link would make it easier.

I'd also suggest that the last edit the person made before the ruling be recorded, that way if there are any questions over when de-escalation should occur, it can be checked without having to do as much cross-referencing. If you want to keep A/VD clean, you could simply make it a note in the text of the ruling itself.

Neither of these is especially burdensome, but they would help to clear up occasional questions that arise later. —Aichon— 17:09, 30 November 2009 (UTC)

The first of those would work, the second would not because of history wipes. Cyberbob  Talk  17:18, 30 November 2009 (UTC)

How are de-escalations currently handled in those cases? Perhaps we should start from there and work backwards. —Aichon— 18:00, 30 November 2009 (UTC)

Well for starters the whole de-escalation system is a load of crap and needs to be ditched... but up until this point it hasn't been a problem as most of the time it doesn't operate over long enough timespans to fall afoul of a wipe. If a wipe happened tomorrow though it would be boned pretty hard - we'd be able to tell how long it had been since an escalation was given easily enough, but we'd have no way of determining how many edits a user had made in that timespan. Cyberbob  Talk  18:08, 30 November 2009 (UTC)

In that case, I guess the logical choices are to either ignore the problem with history wipes, as seems to be the case now, or else change de-escalations somehow (e.g. make them time-based only). —Aichon— 18:33, 30 November 2009 (UTC)

Or ditch them because they're crap for reasons other than their vulnerability to history wipes. Cyberbob  Talk  04:49, 1 December 2009 (UTC)

Well, Poodle mentions elsewhere that it looks like personal contribs might be intact, even if page histories are wiped, which means that de-escalations still work from a technical standpoint. Even so, I do agree that they aren't working for other reasons. —Aichon— 05:29, 1 December 2009 (UTC)

(Edit conflicted, it's like you read my mind) - What if we started linking to the vandalism case in the vandal data? This would provide more relevant info to the vandal data and make it easy to check on cases with a glance. So for example, the vandal data would look something like this, but linking to the archives rather than open cases, obviously:
User:Thaedracy

• Warned 23:44, 27 Nov, 2009.

User:Xzsd

• Banned 01:46, 29 Nov, 2009.

--_GANG Giles Sednik _CAPD 17:20, 30 November 2009 (UTC)

Yep, that's pretty much what I had in mind. —Aichon— 18:00, 30 November 2009 (UTC)

You know, this is more or less what I had first thought up of when I saw the Misconduct case on DDR, especially since I couldn't make heads or tails of Iscariot's altered 48h ban-cum-warning. Was it the same as the previous ban (and thus a copy) or a separate case (and thus a second warning)? I promptly got lost searching up the references, although I grant the dates are right there.

I would offer, as a more complicated option, this:

User:Rudolph

Date	Result	Comment
01:23, 01 January 2009 (UTC)	Warned	Struck Feb 28, 2008, 250 edits.
12:34, 03 March 2009 (UTC)	Warned	Impersonation of User:Easter Rabbit.
12:34, 06 June 2009 (UTC)	Warned	Vandalism of RRF Page
00:01, 25 December 2009 (UTC)	24h Ban	Personal threats against User:Dancer and User:Prancer.

-Wulfenbach 20:39, 30 November 2009 (UTC)

It's got more code, but I like your idea Wulfenbach. It's easy to read and it has the links to cross-reference the actual cases. --Maverick ^{Talk - OBR 404} 22:00, 30 November 2009 (UTC)

Needlessly complicated, I hate it. The strike through should actually be striking shit so it stands out more.-- SA 23:05, 30 November 2009 (UTC)

Not pretty enough. :P --ZsL 23:42, 30 November 2009 (UTC)

I like the effort and sentiment but I agree with SA and Zombie Slayer. Also, a big page like A/VB might break with all those templates on it.--_GANG Giles Sednik _CAPD 23:59, 30 November 2009 (UTC)

Way too complicated. Aichon's version is about as far removed from the current format as needs to be done. Cyberbob  Talk  04:48, 1 December 2009 (UTC)

Double the bans?

What if we did:

1. Warn

2. Warn

3. 24 hrs

4: 24 hrs

5: 48 hrs

6: 48 hrs

7: 1 week

8: 1 week

9: 1 month

10: 1 month

11+: 1 month & Permaban vote

And change 2 months/250 edits to 6 months/750.

That way, Deescalations are reduced, but still give the user a fair amount of chances, WITH the option to lose two a year. --Haliman - Talk 02:06, 1 December 2009 (UTC)

Perhaps we could keep it the same as it is now, just subject each deescalation to a vote, like we do with the bans? That way, history, and contrabutions are checked, and what not? What do you think? -Poodle of Doom^{M! T} 02:19, 1 December 2009 (UTC)

A vote on every deescalation is a waste of time. I think if we use the above idea, along with an appeals center on the talk page, it should work. People that qualify for the Dees need to ask. If they don't it won't count. Shitting the page up with spam would be considered vandalism. --Haliman - Talk 02:21, 1 December 2009 (UTC)

Or we could just ditch the whole thing. Can you honestly say that it's actually doing its job of "reforming vandals"? Cyberbob  Talk  02:24, 1 December 2009 (UTC)

Good point, but no dees is very strict. What if someone (who intends good) breaks a few rules? The whole thing is just far too strict... We still need to allow SOME leeway--Haliman - Talk 02:26, 1 December 2009 (UTC)

If a person truly intends good they won't be escalated. We could also go back to three-warnings-before-the-first-ban; IIRC that was changed to two at the same time as the introduction of the de-escalations. Cyberbob  Talk  02:36, 1 December 2009 (UTC)

Again, as I have stated above, and had agreed with bob before, in a way,..... I don't think this whole thing was intended to reform vandals necessarily. I think it was aimed towards someone who didn't honestly know better. I think we need to keep that in mind. It sounds like we're trying to differentiate between people who knew full good and well what they were doing, and those who made an honest mistake which is why I think a vote would be good. We could go to a new subpage: A/VB/Deescalation:User. Really,... at that, perhaps we could move the conversation to the newest section I started below,.... who's our focus group,... who are we trying to help? Real vandals? Or the mistake making newbie? -Poodle of Doom^{M! T} 02:41, 1 December 2009 (UTC)

I still don't like the idea of once you get it, you can't lose it, Bob -_- --Haliman - Talk 02:50, 1 December 2009 (UTC)

inb4 STD joke. --Haliman - Talk 02:50, 1 December 2009 (UTC)

I don't like the idea either, but I don't like your suggestion any better. It's like saying "Let's take someone who likes destroying the wiki, and give twice the amount of oppertunities to do it." Seriously, when you tell your kid your counting to three.... you stop at three. You don't give them the option of 4, 5, and 6 before you whoop their asses.... -Poodle of Doom^{M! T} 02:54, 1 December 2009 (UTC)

Well played... What if we split the difference? Same as before, but with appeals on the talk page. 6 months/750edits? --Haliman - Talk 02:57, 1 December 2009 (UTC)

That seems a little high to me. I don't really have a good justification for why I feel that way... I just do. What's this going to prove? Really, how does this solve the problem of the VB being a cluster fuck? -Poodle of Doom^{M! T} 03:09, 1 December 2009 (UTC)

Makes the clusterfucks scares and more spread out. What changes do you propose? --Haliman - Talk 03:10, 1 December 2009 (UTC)

I think that's better than the current system, since it'll slow things down, but it's still not great. If the point isn't to reform vandals, then we don't need the edit count, since it only encourages spamming of talk pages, rather than the positive activity that it was designed to encourage. We could knock it down from 750 to 250, or else remove it altogether. Personally, I'd remove it. I also don't like the idea of appeals, since that's just an avenue for unnecessary drama. —Aichon— 05:43, 1 December 2009 (UTC)

What's being reformed?

So what is it that we're all looking to reform? I think I may have missed the point of this whole thing..... -Poodle of Doom^{M! T} 02:34, 1 December 2009 (UTC)

VD = Shit --Haliman - Talk 02:36, 1 December 2009 (UTC)

Thanks captain obvious,.... now seriously, what's so bad about it. So far half the people want it to stay the way it is, the other half want reform, but their to damn stupid to know what to reform it too. So what's really so bad about it? -Poodle of Doom^{M! T} 02:43, 1 December 2009 (UTC)

The data can be inaccurate, for one. It can also get confusing, as seen in the most recent Misconduct case. --Haliman - Talk 02:51, 1 December 2009 (UTC)

I don't really know what misconduct case your referencing? Got a link to it? -Poodle of Doom^{M! T} 02:55, 1 December 2009 (UTC)

*Clicky!* --Haliman - Talk 03:13, 1 December 2009 (UTC)

That wasn't a case of the data being confusing, that was a case of DDR having a rather large brain explosion. Cyberbob  Talk  04:46, 1 December 2009 (UTC)

In terms of VD itself, I think it's just a matter of better transparency and ease of fact-checking. Along the way though, we've also started discussing reforms to the de-escalation policy, which, while being a separate issue, is responsible for much of the difficulty in dealing with VD. —Aichon— 05:27, 1 December 2009 (UTC)

^ ^ ^ This! ^ ^ ^ --_GANG Giles Sednik _CAPD 12:11, 1 December 2009 (UTC)

Unnecessary

There's nothing wrong with A/VD, the problem is the way that sysops are managing it.

The system isn't broken, the systems operators are. There's no parity in random de-escalations, there's no pro-active intent to rectify past errors. Leave the system alone and focus on the actual problem, which doesn't require new policies. -- . . <== DDR Approved Editor 08:53, 1 December 2009 (UTC)

While I disagree with your opening premise, for the sake of argument I'll assume that the sysops are solely to blame. But even if we believe the sysops are the problem, surely we can agree that a system which can be so easily manipulated by them must have some inherent flaws. By creating a more organized, transparent system we can reduce the level of mismanagement and ensure that everyone is treated fairly. --_GANG Giles Sednik _CAPD 12:10, 1 December 2009 (UTC)