Sock Hunting - Done right or not?

[Zironic]

https://en.wikipedia.org/wiki/Wikipedia ... s_Advocate

Do you like apples (T-C-L) and EoT State (T-C-L) have been accused of being TDA socks by PanchoS (T-C-L) due to edit warring and a deletion attempt at Anarcho-capitalism (T-H-L).

The only link to TDA appears to come from Do you like apples, who made their first edits to TDA's talk page in order to remove Liz's "taunting and harassment," which he later apologized for doing on Commons when TDA requested it be restored. This account was indefinitely blocked by Beeblebrox as a CheckUser block, but is unlikely to be TDA as Bbb23 said TDA's CheckUser data was stale.

Most Wikipedia Sockpuppet investigations are quite absurd.

Self-appointed prosecutor: "You share the same opinions as someone I dislike. Prove that you're not that other user or you will be indef blocked."
"Sockpuppet": "Dude, how the fuck am I supposed to prove I'm not another user I know nothing about?"
Self-appointed prosecutor: "The criminal admits to having no defence, the prosecution rests. Block the sockpuppet!"

[Jim]

Most Wikipedia Sockpuppet investigations are quite absurd.
Self-appointed prosecutor: "You share the same opinions as someone I dislike. Prove that you're not that other user or you will be indef blocked."
"Sockpuppet": "Dude, how the fuck am I supposed to prove I'm not another user I know nothing about?"
Self-appointed prosecutor: "The criminal admits to having no defence, the prosecution rests. Block the sockpuppet!"

I'd say "many" as opposed to "most", but only just. Other than that, I agree.

The solution, guess what - don't allow indiscriminate anonymous editing. Oh, have we been here before?

[Zironic]

I'd say "many" as opposed to "most", but only just. Other than that, I agree.

The solution, guess what - don't allow indiscriminate anonymous editing. Oh, have we been here before?

There certainly exist some really good sockpuppet investigations out there, filled with all sorts of data-driven analysis to prove beyond reasonable doubt that two editors are either the same person or at-least indistinguishable in terms of editing habits. However I can certainly understand why most admins don't bother if 'Eh, I feel like he's probably a sock' is enough for a block to stick most of the time.

[Jim]

There certainly exist some really good sockpuppet investigations out there, filled with all sorts of data-driven analysis to prove beyond reasonable doubt that two editors are either the same person or at-least indistinguishable in terms of editing habits. However I can certainly understand why most admins don't bother if 'Eh, I feel like he's probably a sock' is enough for a block to stick most of the time.

The classic is "He's obviously a sock of someone, I'm just not sure who." That works most of the time.

In a system that actually encourages "socking" by making real identities taboo, we should be shocked by this?

[Dennis Brown]

There certainly exist some really good sockpuppet investigations out there, filled with all sorts of data-driven analysis to prove beyond reasonable doubt that two editors are either the same person or at-least indistinguishable in terms of editing habits. However I can certainly understand why most admins don't bother if 'Eh, I feel like he's probably a sock' is enough for a block to stick most of the time.

The classic is "He's obviously a sock of someone, I'm just not sure who." That works most of the time.

In a system that actually encourages "socking" by making real identities taboo, we should be shocked by this?

"Sock of someone" is a valid reason for blocking if it is used right. Someone shows up and from their first edits, they are attacking someone or some idea with insight and meta knowledge that no new user (and most experienced users) would have.... it is easier to block as a sock of someone than to block, then file the SPI, wait 2 weeks, get the linkage, which doesn't matter now. Accounts are disposable, the paperwork just isn't worth it when it is blindingly obvious someone is a sock. The master is irrelevant at that point.

[EthicalWikiUser]

"Sock of someone" is a valid reason for blocking if it is used right. Someone shows up and from their first edits, they are attacking someone or some idea with insight and meta knowledge that no new user (and most experienced users) would have.... it is easier to block as a sock of someone than to block, then file the SPI, wait 2 weeks, get the linkage, which doesn't matter now. Accounts are disposable, the paperwork just isn't worth it when it is blindingly obvious someone is a sock. The master is irrelevant at that point.

This kind of reasoning is a load of shit that completely discounts the very real possibility that someone has been following some drama for a while as an observer and finally made an account to speak up on it. This exact trial-by-assumption crap happened to me once and it's why I just stick to IP editing now. Lord knows I have seen a shitload of other absurd assumed-sockpuppet blocks the past couple years alone as well.

Wikipedia admins seem to have this utopic fantasy of what the process of becoming a new account user is like where a new user will always be this innocent child who will always make tepid edits and always learn the ins and outs of things slowly and steadily. In reality, most people make an account on Wikipedia because a specific subject interests them (e.g., an "SPA") or they've been following something for a while and want to get more invested in it (obviously a sockpuppet!). Your options are damned if you don't learn the rules fast enough and damned if you do.

There are certainly a number of other reasons, but this kind of reasoning is part of the cause of the steady decline of Wikipedia's editor count over the years. When you keep scaring off new, potential long-term quality editors with this clubhouse culture of your-opinion-only-matters-if-you-have-X-edits, this is your result.

[Kelly Martin]

Someone shows up and from their first edits, they are attacking someone or some idea with insight and meta knowledge that no new user (and most experienced users) would have.... it is easier to block as a sock of someone than to block, then file the SPI, wait 2 weeks, get the linkage, which doesn't matter now. Accounts are disposable, the paperwork just isn't worth it when it is blindingly obvious someone is a sock. The master is irrelevant at that point.

This has always been an utterly bankrupt rationale; it assumes that the first thing a Wikipedia editor does is get an account, then make the usual "first edits" and so forth, and anyone whose behavioral trajectory that deviates from this hypothetical ideal is an "obvious sockpuppet". But virtually nobody actually comes to Wikipedia that way. Many people will edit as an IP for quite some time, and many will "lurk" for a long time before editing at all; in either case, a new user might be quite experienced without having any "edit history" to explain that experience.

The purpose of this practice is to punish people who don't conform with the notion that one isn't really a Wikipedian until one registers a Wikipedia account, and does all of one's edits with that account. I know people who do 90% of their editing while logged out, and use their named accounts only when they need the privileges that accord to a logged-in account. Such people (some of whom are remarkably competent editors) are marginalized by the Wikipedia community, because they have not adopted a proper Wikipedia identity.

A surprising amount of Wikipedia content editing is done by editors who are not part of the community, but, because these people edit in ways that do not draw attention to themselves, the community not only ignores them, but often even believes that they don't exist and are not important.

[Dennis Brown]

Someone shows up and from their first edits, they are attacking someone or some idea with insight and meta knowledge that no new user (and most experienced users) would have.... it is easier to block as a sock of someone than to block, then file the SPI, wait 2 weeks, get the linkage, which doesn't matter now. Accounts are disposable, the paperwork just isn't worth it when it is blindingly obvious someone is a sock. The master is irrelevant at that point.

This has always been an utterly bankrupt rationale; it assumes that the first thing a Wikipedia editor does is get an account, then make the usual "first edits" and so forth, and anyone whose behavioral trajectory that deviates from this hypothetical ideal is an "obvious sockpuppet". But virtually nobody actually comes to Wikipedia that way. Many people will edit as an IP for quite some time, and many will "lurk" for a long time before editing at all; in either case, a new user might be quite experienced without having any "edit history" to explain that experience.

The purpose of this practice is to punish people who don't conform with the notion that one isn't really a Wikipedian until one registers a Wikipedia account, and does all of one's edits with that account. I know people who do 90% of their editing while logged out, and use their named accounts only when they need the privileges that accord to a logged-in account. Such people (some of whom are remarkably competent editors) are marginalized by the Wikipedia community, because they have not adopted a proper Wikipedia identity.

A surprising amount of Wikipedia content editing is done by editors who are not part of the community, but, because these people edit in ways that do not draw attention to themselves, the community not only ignores them, but often even believes that they don't exist and are not important.

It feels like no one actually read my post. As I said, if it is done "right". Many people DO sockpuppet to attack an enemy. When someone comes in as an SPA and has entirely too much knowledge of policy, procedure and such, and their only objective it to under mine an established editor, then it shouldn't be shocking they are usually a sock. CU proves it out, often it is so obvious a CU won't touch it. Often someone knew comes in and attacks, but those genuine new users are pretty easy to spot. Those you help, you show policy, you explain dispute resolution. But anyone that thinks that people do not regularly sock only to undermine an enemy in a dispute is a fool. It's why they have the 500 edit limit in GG, for instance. I won't put a percentage on it, I'm simply saying this type of socking does exist, and blocking for it is perfectly valid. If you want to bitch about it, find a specific example and ping that admin, but theorizing that all blocks without a master are invalid is silly and Pollyanna.

[Zironic]

Shouldn't those users be blocked for being disruptive and NOTHERE rather then being socks? Socking isn't even actually against Wikipedia policy is it?

[LynnWysong]

"Sock of someone" is a valid reason for blocking if it is used right.... it is easier to block as a sock of someone than to block, then file the SPI, wait 2 weeks, get the linkage, which doesn't matter now. Accounts are disposable, the paperwork just isn't worth it when it is blindingly obvious someone is a sock. The master is irrelevant at that point.

You mean like Dreadstar did to me, based on the lies ("from their first edits, they are attacking someone or some idea with insight and meta knowledge that no new user (and most experienced users) would have") of a rogue editor who made all her assertions via email so they couldn't be refuted? The only thing that saved me was that I had used my real name. If I hadn't I probably would have been one of the masses that are driven off by editors that take advantage of new editors ignorance of how they are gaming the system.

[Dennis Brown]

Shouldn't those users be blocked for being disruptive and NOTHERE rather then being socks? Socking isn't even actually against Wikipedia policy is it?

Most of the time, they are indeed blocked for disruption, with only a note of "likely sock of someone", or as a combination of the two. This is a note to reviewing admin as well.

Socking is against policy, period, there is a detailed policy on what is socking and what is just an alternate account. If the alt account is used for abuse of any kind, it is socking, and it makes life miserable for regular editors. This is why socking is treated aggressively, and why errors do happen.

That said, I know several people who are very likely socks but I don't pursue it because they aren't disrupting, they are contributing. It wouldn't improve the encyclopedia, and I'm not there to enforce policy just for the sake of enforcing policy. When someone is clearly socking and disrupting (sometimes passing the bar of WP:DE, sometimes not), they are making life hell for good faith editors, and it is the admin's job to deal with it. We sometimes get it wrong, but that can be said of any admin action. Socks typically only get noticed when they cause drama, generate spam, or act outside of the standard policies. A sock (former banned user, for instance) that just edits and minds his own business could avoid detection for a long time, if not forever.

[LynnWysong]

We sometimes get it wrong, but that can be said of any admin action. Socks typically only get noticed when they cause drama, generate spam, or act outside of the standard policies.

There's no excuse for getting it wrong when blocking someone for being a sock without going through the SPI process.

[Dennis Brown]

We sometimes get it wrong, but that can be said of any admin action. Socks typically only get noticed when they cause drama, generate spam, or act outside of the standard policies.

There's no excuse for getting it wrong when blocking someone for being a sock without going through the SPI process.

Lynn, that is just silly. The "SPI process" is just listing it so an admin can look at it and say "Yup, that's a sock" and block him. You say "process" like there is some formal debate at SPI: there isn't. The majority of sockpuppetry cases already go to SPI because they are filed by non-admin, but seldom do any sock cases qualify for a CU, which is seen as a potential invasion of privacy. The overwhelming majority are bases solely on an admin's opinion of the behavior and nothing more. And even when you use CU, half the time the result is useless. I can easily bounce through a proxy, change my headers and a CU would have no way of linking it to me. There is a reason they say "CU isn't magic pixie dust".

SPI isn't some magical place where justice is meted out, or where the result is more likely to be correct. There is zero evidence of that. If I saw a sock, I wouldn't take them to SPI, I would just handle it. I worked at SPI for a long while, know how to investigate, how to check intersects, how to use behavioral analysis to link when there is a known master, and in the rare cases where a master isn't know, when to block and when not to. I'm not the only admin with that kind of experience, plenty have that and more. Most of the time, admin get it right. The key is how do they handle it when they get it wrong. That is where you see their character. But there is no way to have a system that is 100% error-free unless you require everyone to identify and register to edit. Those are the breaks.

So we are diligent, we do our homework, we try to not jump to conclusions, and every admin is different....some better, some not so better. Mistakes will be made, it is unavoidable. That is still better than letting socks run amok and screw up AFD votes and pound topics like P/I and GG, which would run off even MORE editors and undermine the integrity of the encyclopedia. The current flawed system is the lesser of the two available evils.

[LynnWysong]

We sometimes get it wrong, but that can be said of any admin action. Socks typically only get noticed when they cause drama, generate spam, or act outside of the standard policies.

There's no excuse for getting it wrong when blocking someone for being a sock without going through the SPI process.

Lynn, that is just silly. The "SPI process" is just listing it so an admin can look at it and say "Yup, that's a sock" and block him. You say "process" like there is some formal debate at SPI: there isn't. The majority of sockpuppetry cases already go to SPI because they are filed by non-admin, but seldom do any sock cases qualify for a CU, which is seen as a potential invasion of privacy. The overwhelming majority are bases solely on an admin's opinion of the behavior and nothing more. And even when you use CU, half the time the result is useless. I can easily bounce through a proxy, change my headers and a CU would have no way of linking it to me. There is a reason they say "CU isn't magic pixie dust".

SPI isn't some magical place where justice is meted out, or where the result is more likely to be correct. There is zero evidence of that. If I saw a sock, I wouldn't take them to SPI, I would just handle it. I worked at SPI for a long while, know how to investigate, how to check intersects, how to use behavioral analysis to link when there is a known master, and in the rare cases where a master isn't know, when to block and when not to. I'm not the only admin with that kind of experience, plenty have that and more. Most of the time, admin get it right. The key is how do they handle it when they get it wrong. That is where you see their character. But there is no way to have a system that is 100% error-free unless you require everyone to identify and register to edit. Those are the breaks.

So we are diligent, we do our homework, we try to not jump to conclusions, and every admin is different....some better, some not so better. Mistakes will be made, it is unavoidable. That is still better than letting socks run amok and screw up AFD votes and pound topics like P/I and GG, which would run off even MORE editors and undermine the integrity of the encyclopedia. The current flawed system is the lesser of the two available evils.

That's like saying the police should be able to convict someone without a trial. There are other aspects to an SPI, one being that the accused sock can defend themselves against the "evidence". That evidence is documented for all to see, instead of just a "take my word for it, I know a sock when I see one" type of conviction you are defending. Others can evaluate the evidence and speak out if they think it is being misinterpreted or misrepresented. My experience is that blocking suspected socks without an SPI can be too easily abused to be tolerated.

[Dennis Brown]

We sometimes get it wrong, but that can be said of any admin action. Socks typically only get noticed when they cause drama, generate spam, or act outside of the standard policies.

There's no excuse for getting it wrong when blocking someone for being a sock without going through the SPI process.

Lynn, that is just silly. The "SPI process" is just listing it so an admin can look at it and say "Yup, that's a sock" and block him. You say "process" like there is some formal debate at SPI: there isn't. The majority of sockpuppetry cases already go to SPI because they are filed by non-admin, but seldom do any sock cases qualify for a CU, which is seen as a potential invasion of privacy. The overwhelming majority are bases solely on an admin's opinion of the behavior and nothing more. And even when you use CU, half the time the result is useless. I can easily bounce through a proxy, change my headers and a CU would have no way of linking it to me. There is a reason they say "CU isn't magic pixie dust".

SPI isn't some magical place where justice is meted out, or where the result is more likely to be correct. There is zero evidence of that. If I saw a sock, I wouldn't take them to SPI, I would just handle it. I worked at SPI for a long while, know how to investigate, how to check intersects, how to use behavioral analysis to link when there is a known master, and in the rare cases where a master isn't know, when to block and when not to. I'm not the only admin with that kind of experience, plenty have that and more. Most of the time, admin get it right. The key is how do they handle it when they get it wrong. That is where you see their character. But there is no way to have a system that is 100% error-free unless you require everyone to identify and register to edit. Those are the breaks.

So we are diligent, we do our homework, we try to not jump to conclusions, and every admin is different....some better, some not so better. Mistakes will be made, it is unavoidable. That is still better than letting socks run amok and screw up AFD votes and pound topics like P/I and GG, which would run off even MORE editors and undermine the integrity of the encyclopedia. The current flawed system is the lesser of the two available evils.

That's like saying the police should be able to convict someone without a trial. There are other aspects to an SPI, one being that the accused sock can defend themselves against the "evidence". That evidence is documented for all to see, instead of just a "take my word for it, I know a sock when I see one" type of conviction you are defending. Others can evaluate the evidence and speak out if they think it is being misinterpreted or misrepresented. My experience is that blocking suspected socks without an SPI can be too easily abused to be tolerated.

Complicated cases already go there. Most cases aren't complicated, user:bob gets blocked, and magically user:alice shows up making the same reverts on the same articles, particularly if manually adding the same material. You don't bog down SPI with stuff like this. The community selected us to simply ACT when it is obvious. And WP:ADMINACCT means if I block a sock, I have to justify it, even if it is 6 months later.

If you took the most basic blocks to SPI, it would be months backlogged. No one wants to work SPI, it is a pain. 99% of the time, the accused party doesn't participate and doesn't need to. If you see all kinds of similar articles and edits, and you do get a CU to link them up technically, and the behavior is exactly the same, what can they say that would change the reality that they are a sock? SPI is not a court. SPI blocks CAN be abused, or simply screwed up but honestly, it isn't as likely to be abused as other kinds of blocks, like areas under discretionary sanctions. Those are necessary, but can be gamed as well.

I screwed up one at SPI, Alison called me out on it, it was a genuine error, I went and apologized on their talk page in a large, bold and sincere way. I took responsibility. That is all I can do. Have I made other errors? Maybe. Probably, but I have 2000 blocks behind me, 1500 of them at SPI, mainly chasing spammers and the like. People talk about "abuse", but the simple fact of the matter is that most of the time an admin makes a bad block, it is just that they screwed up. We're human. I'm wise enough to raise my hand up high, apologize and admit the error. A few admin dig in or won't simply agree to disagree and defer to the opinion of the community, so it looks like abuse. It really isn't, it is just bad form. And once in a blue moon, yes, you get abuse.

[LynnWysong]

We sometimes get it wrong, but that can be said of any admin action. Socks typically only get noticed when they cause drama, generate spam, or act outside of the standard policies.

There's no excuse for getting it wrong when blocking someone for being a sock without going through the SPI process.

Lynn, that is just silly. The "SPI process" is just listing it so an admin can look at it and say "Yup, that's a sock" and block him. You say "process" like there is some formal debate at SPI: there isn't. The majority of sockpuppetry cases already go to SPI because they are filed by non-admin, but seldom do any sock cases qualify for a CU, which is seen as a potential invasion of privacy. The overwhelming majority are bases solely on an admin's opinion of the behavior and nothing more. And even when you use CU, half the time the result is useless. I can easily bounce through a proxy, change my headers and a CU would have no way of linking it to me. There is a reason they say "CU isn't magic pixie dust".

SPI isn't some magical place where justice is meted out, or where the result is more likely to be correct. There is zero evidence of that. If I saw a sock, I wouldn't take them to SPI, I would just handle it. I worked at SPI for a long while, know how to investigate, how to check intersects, how to use behavioral analysis to link when there is a known master, and in the rare cases where a master isn't know, when to block and when not to. I'm not the only admin with that kind of experience, plenty have that and more. Most of the time, admin get it right. The key is how do they handle it when they get it wrong. That is where you see their character. But there is no way to have a system that is 100% error-free unless you require everyone to identify and register to edit. Those are the breaks.

So we are diligent, we do our homework, we try to not jump to conclusions, and every admin is different....some better, some not so better. Mistakes will be made, it is unavoidable. That is still better than letting socks run amok and screw up AFD votes and pound topics like P/I and GG, which would run off even MORE editors and undermine the integrity of the encyclopedia. The current flawed system is the lesser of the two available evils.

That's like saying the police should be able to convict someone without a trial. There are other aspects to an SPI, one being that the accused sock can defend themselves against the "evidence". That evidence is documented for all to see, instead of just a "take my word for it, I know a sock when I see one" type of conviction you are defending. Others can evaluate the evidence and speak out if they think it is being misinterpreted or misrepresented. My experience is that blocking suspected socks without an SPI can be too easily abused to be tolerated.

Complicated cases already go there. Most cases aren't complicated, user:bob gets blocked, and magically user:alice shows up making the same reverts on the same articles, particularly if manually adding the same material. You don't bog down SPI with stuff like this. The community selected us to simply ACT when it is obvious. And WP:ADMINACCT means if I block a sock, I have to justify it, even if it is 6 months later.

If you took the most basic blocks to SPI, it would be months backlogged. No one wants to work SPI, it is a pain. 99% of the time, the accused party doesn't participate and doesn't need to. If you see all kinds of similar articles and edits, and you do get a CU to link them up technically, and the behavior is exactly the same, what can they say that would change the reality that they are a sock? SPI is not a court. SPI blocks CAN be abused, or simply screwed up but honestly, it isn't as likely to be abused as other kinds of blocks, like areas under discretionary sanctions. Those are necessary, but can be gamed as well.

I screwed up one at SPI, Alison called me out on it, it was a genuine error, I went and apologized on their talk page in a large, bold and sincere way. I took responsibility. That is all I can do. Have I made other errors? Maybe. Probably, but I have 2000 blocks behind me, 1500 of them at SPI, mainly chasing spammers and the like. People talk about "abuse", but the simple fact of the matter is that most of the time an admin makes a bad block, it is just that they screwed up. We're human. I'm wise enough to raise my hand up high, apologize and admit the error. A few admin dig in or won't simply agree to disagree and defer to the opinion of the community, so it looks like abuse. It really isn't, it is just bad form. And once in a blue moon, yes, you get abuse.

All I can say is that certain editors that have reputations for being serial sock accusers should never be admins.

[tarantino]

"Sock of someone" is a valid reason for blocking if it is used right. Someone shows up and from their first edits, they are attacking someone or some idea with insight and meta knowledge that no new user (and most experienced users) would have.... it is easier to block as a sock of someone than to block, then file the SPI, wait 2 weeks, get the linkage, which doesn't matter now. Accounts are disposable, the paperwork just isn't worth it when it is blindingly obvious someone is a sock. The master is irrelevant at that point.

This kind of reasoning is a load of shit that completely discounts the very real possibility that someone has been following some drama for a while as an observer and finally made an account to speak up on it. This exact trial-by-assumption crap happened to me once and it's why I just stick to IP editing now. Lord knows I have seen a shitload of other absurd assumed-sockpuppet blocks the past couple years alone as well.

As it was for me. I was blocked as a checkuser-confirmed sockpuppet of an arbcom banned user after my first edit, because Dominic Byrd-McDevitt said so on IRC.

[Dennis Brown]

"Sock of someone" is a valid reason for blocking if it is used right. Someone shows up and from their first edits, they are attacking someone or some idea with insight and meta knowledge that no new user (and most experienced users) would have.... it is easier to block as a sock of someone than to block, then file the SPI, wait 2 weeks, get the linkage, which doesn't matter now. Accounts are disposable, the paperwork just isn't worth it when it is blindingly obvious someone is a sock. The master is irrelevant at that point.

This kind of reasoning is a load of shit that completely discounts the very real possibility that someone has been following some drama for a while as an observer and finally made an account to speak up on it. This exact trial-by-assumption crap happened to me once and it's why I just stick to IP editing now. Lord knows I have seen a shitload of other absurd assumed-sockpuppet blocks the past couple years alone as well.

As it was for me. I was blocked as a checkuser-confirmed sockpuppet of an arbcom banned user after my first edit, because Dominic Byrd-McDevitt said so on IRC.

Which proves CU and SPI aren't the wonder tools they are claimed to be, and why behavioral analysis and a bit of patience are the best tools. Sadly, no admin is trained for analysing behavior, and most are not very good at it. Even then, mistakes will be made. Again, it is unavoidable.

[Dennis Brown]

All I can say is that certain editors that have reputations for being serial sock accusers should never be admins.

I would agree with that. Some have been barred from filing at SPI because they get it wrong so often. I've known some editors that report a lot of socks, but their ratio is still bad. Others, report at SPI after just a couple of edits, which isn't enough to determine a behavioral match. A few are just annoying as hell in the way they go about it, drumming up drama instead of just reporting and moving on.

Many admin shouldn't make sock blocks, and most understand that and don't unless it is super obvious. (ie: user:bob, user:bob2, user:bob3) We actually get socks that do that, just rotate numbers until we get a CU to privately pull up the IP and do a range block. The percentage of admin that regularly do sock blocks is fairly low.

[MoldyHay]

This doesn't seem like the best of practices from Arbcom.

What do you mean? If this isn't an ArbCom Best Practice™, I don't know what is!

[Vigilant]

So we are diligent, we do our homework, we try to not jump to conclusions.

Dennis,

That's just bullshit and we all know it.
Half of the listed Socks of Vigilant aren't mine.
Many times I've seen other people being accused of being my sock and I've had to step in and say that's not the case.

Admins see socks of the people they have recently fought with and just block the account because it's too damn hard to do a proper investigation with the limited tools and poor understanding of the internet that almost all admins and checkusers have.

[The Joy]

So we are diligent, we do our homework, we try to not jump to conclusions.

Dennis,

That's just bullshit and we all know it.
Half of the listed Socks of Vigilant aren't mine.
Many times I've seen other people being accused of being my sock and I've had to step in and say that's not the case.

Admins see socks of the people they have recently fought with and just block the account because it's too damn hard to do a proper investigation with the limited tools and poor understanding of the internet that almost all admins and checkusers have.

When Jon Awbrey was fighting with the Wikipedians over Charles S. Peirce (T-H-L), anyone that edited that article got blocked because "It's obviously Jon Awbrey!" Well, it wasn't always. Poor The Tetrast (T-C-L) got whacked twice as a supposed sock.

[Kelly Martin]

Which proves CU and SPI aren't the wonder tools they are claimed to be, and why behavioral analysis and a bit of patience are the best tools. Sadly, no admin is trained for analysing behavior, and most are not very good at it. Even then, mistakes will be made. Again, it is unavoidable.

It's not unavoidable. In fact, it could be easily avoided, by the simple expediency of providing training.

Yes, I know, it wouldn't eliminate all of it, but it would eliminate a lot of it.

[Jim]

user:bob gets blocked, and magically user:alice shows up making the same reverts on the same articles, particularly if manually adding the same material. You don't bog down SPI with stuff like this. The community selected us to simply ACT when it is obvious. And WP:ADMINACCT means if I block a sock, I have to justify it, even if it is 6 months later.

We were talking, I thought, about blocking alice as "a sock of somebody, but I don't know who", rather than as "a sock of bob". I see those as distinctly different actions and rationales.
You have "disruptive editing" if alice can be shown to be being disruptive, but isn't undeniably bob (or fred).

[Dennis Brown]

So we are diligent, we do our homework, we try to not jump to conclusions.

Dennis,

That's just bullshit and we all know it.
Half of the listed Socks of Vigilant aren't mine.
Many times I've seen other people being accused of being my sock and I've had to step in and say that's not the case.

Admins see socks of the people they have recently fought with and just block the account because it's too damn hard to do a proper investigation with the limited tools and poor understanding of the internet that almost all admins and checkusers have.

They shouldn't be blocking anyone they've fought with to begin with, sock or not. Assuming we are talking about legitimate cases, it is possible that someone like you has a lot of false socks, so an admin goes and sees "this is a match" by looking at the most recent socks. In this case, you have two different sock masters, you, and someone confused as you early on. The sock really does match some of the newer socks, but not the master. They are correct in nailing the sock, but not in tying it to the master. I'm not saying it is right, just how it happens in good faith. When matching socks, you can't go back and analyse every single sock, you pick a few "confirmed" socks and compare to those, and typically the more recent ones. Again, I'm just saying how it is done.

And this has nothing to do with blocking without a master, btw.

[Dennis Brown]

Which proves CU and SPI aren't the wonder tools they are claimed to be, and why behavioral analysis and a bit of patience are the best tools. Sadly, no admin is trained for analysing behavior, and most are not very good at it. Even then, mistakes will be made. Again, it is unavoidable.

It's not unavoidable. In fact, it could be easily avoided, by the simple expediency of providing training.

Yes, I know, it wouldn't eliminate all of it, but it would eliminate a lot of it.

I would agree that training would be helpful, but who is the master that could teach? When I joined SPI as a clerk, there wasn't even a single page that had basic instructions on the paperwork. DoRD fixed that later on. I wasn't given any training, just turned loose. Keep in mind, virtually NO admin that do sock blocks have ever worked at SPI. You have to be accepted by the CUs to be an SPI clerk, who then do nothing to train you, and half of them don't know the proper procedures as far as paperwork goes.

In regard to teaching admin proper investigative techniques, there is no course. I have the life experience of being older, plus some experience as a criminal defense investigator early in life, so I try to be a bit skeptical in how I read evidence. It is far from perfect, but it is something. Even if you had a course, it would be difficult to force admin to take the course before sock blocking, and some would learn enough by rote to simply pass and then forget it in the field. If Wikipedia was a for profit corporation, things would be different and you have have a real system in place, but I don't see it changing. Training costs money, and apparently Wikipedia is struggling with only 60 million in the bank, so much so they had to run donation banner ads as big as the entire page recently.........

One of the things I did not like about being part of the SPI team was the reliance on IRC. I was there over a year, and quit in disgust over a case where the Foundation got involved, and which pushed WilliamH to retire and hand in his CU/Admin/Crat bits. That was the case where I blocked over 300 socks in a few day marathon.

[Jim]

I was there over a year, and quit in disgust over a case where the Foundation got involved, and which pushed WilliamH to retire and hand in his CU/Admin/Crat bits. That was the case where I blocked over 300 socks in a few day marathon.

You keep telling us about that, without actually telling us about that. I understand you can't be too specific, but can you give some clues as to why this is such a huge issue for you?

[Dennis Brown]

user:bob gets blocked, and magically user:alice shows up making the same reverts on the same articles, particularly if manually adding the same material. You don't bog down SPI with stuff like this. The community selected us to simply ACT when it is obvious. And WP:ADMINACCT means if I block a sock, I have to justify it, even if it is 6 months later.

We were talking, I thought, about blocking alice as "a sock of somebody, but I don't know who", rather than as "a sock of bob". I see those as distinctly different actions and rationales.
You have "disruptive editing" if alice can be shown to be being disruptive, but isn't undeniably bob (or fred).

If you read my comment, I simply say many admin only block under perfect situations, or simply don't at all, not all admin are the same. Many won't block blind, ie: without a known master.

I can tell this is a hot button for lots of folks, but I won't budge on this. As an admin, you have to be very careful, and not every admin IS very careful, but in the end, there are sometimes very valid reasons to block someone as a sock with an unknown sockmaster. It isn't common, but I'm not going to apologize for it as no one has provided any specific instance of my being mistaken.

I've probably done it 5 or 6 times out of 2000 blocks, but if it arrises again, I will do it again. You are welcome to challenge any specific block at that time. I can't speak for other admin or how careful or accurate they are, but I'm not responsible for their actions, only my own.

I see everyone complaining about the theory of blocking without a master, but I haven't seen a specific example, just condemnation of the practice. Find your examples, go ask that admin. Otherwise, this is just gums flapping in the breeze, as there is no hard evidence, just vague claims.

[Jim]

Find your examples, go ask that admin. Otherwise, this is just gums flapping in the breeze, as there is no hard evidence, just vague claims.

Ok. I was just asking in general. Sorry. Won't happen again.

[Poetlister]

As it was for me. I was blocked as a checkuser-confirmed sockpuppet of an arbcom banned user after my first edit, because Dominic Byrd-McDevitt said so on IRC.

He's definitely one of the worst admins on Wikipedia, a title he achieves despite the stiff opposition for the title.

[thekohser]

As it was for me. I was blocked as a checkuser-confirmed sockpuppet of an arbcom banned user after my first edit, because Dominic Byrd-McDevitt said so on IRC.

He's definitely one of the worst admins on Wikipedia, a title he achieves despite the stiff opposition for the title.

I wonder if he will figure in the (eventual) response to my FOIA request of the National Archives?

[Randy from Boise]

As it was for me. I was blocked as a checkuser-confirmed sockpuppet of an arbcom banned user after my first edit, because Dominic Byrd-McDevitt said so on IRC.

He's definitely one of the worst admins on Wikipedia, a title he achieves despite the stiff opposition for the title.

I wonder if he will figure in the (eventual) response to my FOIA request of the National Archives?

Emphasis on "eventual."

I did finally get my microfilm, so there is always hope.

RfB

[Kelly Martin]

I would agree that training would be helpful, but who is the master that could teach? When I joined SPI as a clerk, there wasn't even a single page that had basic instructions on the paperwork. DoRD fixed that later on. I wasn't given any training, just turned loose. Keep in mind, virtually NO admin that do sock blocks have ever worked at SPI. You have to be accepted by the CUs to be an SPI clerk, who then do nothing to train you, and half of them don't know the proper procedures as far as paperwork goes.

That's a cultural problem with Wikipedia. It doesn't have to be that way.

The problem with Wikipedia is that far too many of its functionaries are zealous of their knowledge and refuse to share it with anyone else except in the context of a quid pro quo. Knowledge is power, and to give away power in exchange for nothing is just stupid.

There are arguably three people most to blame for this state of affairs in Wikipedia: Jimmy Wales, Erik Moeller, and David Gerard.

[Dennis Brown]

I would agree that training would be helpful, but who is the master that could teach? When I joined SPI as a clerk, there wasn't even a single page that had basic instructions on the paperwork. DoRD fixed that later on. I wasn't given any training, just turned loose. Keep in mind, virtually NO admin that do sock blocks have ever worked at SPI. You have to be accepted by the CUs to be an SPI clerk, who then do nothing to train you, and half of them don't know the proper procedures as far as paperwork goes.

That's a cultural problem with Wikipedia. It doesn't have to be that way.

The problem with Wikipedia is that far too many of its functionaries are zealous of their knowledge and refuse to share it with anyone else except in the context of a quid pro quo. Knowledge is power, and to give away power in exchange for nothing is just stupid.

There are arguably three people most to blame for this state of affairs in Wikipedia: Jimmy Wales, Erik Moeller, and David Gerard.

I've only crossed paths with Eric Moeller once, where I asked for him to be desysopped. My opinions are pretty clear, as are those by others. Never crossed paths with David. Jimbo...I've expressed enough on him publicly, that isn't hard to find. https://en.wikipedia.org/wiki/Wikipedia ... nnis_Brown

[Dennis Brown]

I was there over a year, and quit in disgust over a case where the Foundation got involved, and which pushed WilliamH to retire and hand in his CU/Admin/Crat bits. That was the case where I blocked over 300 socks in a few day marathon.

You keep telling us about that, without actually telling us about that. I understand you can't be too specific, but can you give some clues as to why this is such a huge issue for you?

Will was a friend, and he was universally liked by everyone. Wikipedia lost a very experienced and well trusted Crat for starters. He was one of the good guys, and the system ran him off. That is why it is a huge issue. I truly liked working with him. He played by the rules, he wouldn't do CU checks as "favors" so he was a clean CU, but he would explain the criterias and work with you to meet them. He would share in the tasks and didn't just lord over admins as a CU/superadmin. He treated it as a partnership, more than anyone else I had worked with. So yeah, it still pisses me off that he left, and it was a blow to the system, and a win for paid editing sockpuppetry.

[Kumioko]

Sock hunting is an often abused past time on Wikipedia that many admins do just because they can run up their numbers. Its easy to accuse someone of being a DUCK because they know too much for a new user.

The fact is a high number of the current active contributing community has used another account prior to the one they currently use and most won't admit that. Even if they edited as an IP or some other legitimate account, they still had previous editing.

Additionally, the checkuser tool is garbage and is prone to false positives and far too much trust is put in the tool and the users of it. Its often used when there is no reason too as a fishing expedition and many times even in cases where they admit there is no evidence to prove it is a sock they block it anyway because they can't prove it isn't either. It makes a mockery out of the project,, the people and the tools.

[collect]

Sock hunting is an often abused past time on Wikipedia that many admins do just because they can run up their numbers. Its easy to accuse someone of being a DUCK because they know too much for a new user.

The fact is a high number of the current active contributing community has used another account prior to the one they currently use and most won't admit that. Even if they edited as an IP or some other legitimate account, they still had previous editing.

Additionally, the checkuser tool is garbage and is prone to false positives and far too much trust is put in the tool and the users of it. Its often used when there is no reason too as a fishing expedition and many times even in cases where they admit there is no evidence to prove it is a sock they block it anyway because they can't prove it isn't either. It makes a mockery out of the project,, the people and the tools.

More to the point is not the deliberately vague "duck test" which is far too often pure bs, but whether the net effect of the person's edits so exactly mirrors another editor's edits as to arouse doubt as to actual identity - to that end, the single best "secret system" for finding genuine socks is typos! "Joe Jobs" generally so exaggerate the person they are designed to harm that they are ludicrous to anyone who has had to read literally millions of messages over the years. But they are routinely ruled to be socks of the person they are intended to harm.

SPI has, to my best estimation, over a 1/3 "false positive" rate when the "duck test" is used. And also note, genuine new editors are almost invariably blocked for some reason or another without any actual evidence still. Just mention any major sock master who vaguely edited on the same topic and poof! the annoying newbie is gone. In the meantime, "newbies" who post at ANI on their second edit - who are pretty clearly not newbies, get a free pass if they have "approved" opinions.

Frankly, the bureaucracy of the current project would benefit to a great degree by actually paying genuine professional editors and journalists to oversee it, rather than having benighted (beknighted? <g>) "elected officials", even elected by the plebeians of Wikipedia.

[Jim]

In the meantime, "newbies" who post at ANI on their second edit - who are pretty clearly not newbies, get a free pass if they have "approved" opinions.

It would be amusing to see an example of that. Do you have one?

[Tarc]

Frankly, the bureaucracy of the current project would benefit to a great degree by actually paying genuine professional editors and journalists to oversee it, rather than having benighted (beknighted? <g>) "elected officials", even elected by the plebeians of Wikipedia.

Well, equally frankly, it gets a bit tiring to read this sort of thing over and over, because it will never...never never, ever ever...happen. You're talking about changing the bedrock of what it is to be the Wikipedia. They are not going to move away from their idealized vision of "Web 2.0".

I'm not championing or cheer-leading the Wikipedia model here, but these fanciful what-if ideas are pointless.

[Ross McPherson]

The ‘community’ is actually made up of ‘communities’ so small and autonomous they are really just ‘gangs’. Socks are guys in hoodies trying to get about the ‘hood unmolested or maybe whacking rival gangsters. The cops generally end up working with the established gangs because some order is better than none.

The significant fact for me is this: nobody is forced to be part of this underworld experience. The city lights look impressive from a distance and that is all the outside world sees.

You crazy bastards.

[Wonderer]

I don't remember in what thread Poetlister said this, so forgive me if I'm misquoting. Poetlister said his socks that got caught are socks he allowed to get caught. And someone else said something about chess, that you have to sacrifice some pawns every now and then.

So of course the sock hunters are going to get it wrong most of the time. They're too busy chasing the pawns while the bishops and the rooks are doing some serious damage.

And in yet more retread of many things that have been said here before, Wikipedia is like thermonuclear war, only winning move is not to play. And hopefully not get dragged in (e.g., Seigenthaler).

[Ross McPherson]

Most of the time, they are indeed blocked for disruption, with only a note of "likely sock of someone", or as a combination of the two. This is a note to reviewing admin as well.

Socking is against policy, period, there is a detailed policy on what is socking and what is just an alternate account. If the alt account is used for abuse of any kind, it is socking, and it makes life miserable for regular editors. This is why socking is treated aggressively, and why errors do happen.

That said, I know several people who are very likely socks but I don't pursue it because they aren't disrupting, they are contributing. It wouldn't improve the encyclopedia, and I'm not there to enforce policy just for the sake of enforcing policy. When someone is clearly socking and disrupting (sometimes passing the bar of WP:DE, sometimes not), they are making life hell for good faith editors, and it is the admin's job to deal with it. We sometimes get it wrong, but that can be said of any admin action. Socks typically only get noticed when they cause drama, generate spam, or act outside of the standard policies. A sock (former banned user, for instance) that just edits and minds his own business could avoid detection for a long time, if not forever.

An invitation to sock! Thanks but no thanks. I would only go back to challenging the editorial control of your ‘regulars’.

[Stan Dixon]

This whole discussion reminds me of the notorious, gun-packing, sock hunting expert Durova.

From reading this thread it seems to me that Wikipedia uses the same methods rather a lot.

Sock hunting is part of the MMRPG that we see at Wikipedia.

[NativeForeigner]

I'm glad that it seems SPI has moved off of IRC.

That being said, I think outcomes of officially filed SPI (in the end) tend to be more correct. However, filing an SPI is pretty damaging in itself, and there have been several cases where I've been inclined to block the individual filing an SPI for an absurd implication of socking. (I don't think I have though)

Also a lot more of the socking at SPI you see is really silly "guy makes second account to contest AfD on his brother's band" than controversial/TDAesque. I think a lot more people get indeffed at SPI than should, but that's just a personal view. Even though there is now a general procedures page, a lot of it is on discretion.

[Kumioko]

I'm glad that it seems SPI has moved off of IRC.

That being said, I think outcomes of officially filed SPI (in the end) tend to be more correct. However, filing an SPI is pretty damaging in itself, and there have been several cases where I've been inclined to block the individual filing an SPI for an absurd implication of socking. (I don't think I have though)

Also a lot more of the socking at SPI you see is really silly "guy makes second account to contest AfD on his brother's band" than controversial/TDAesque. I think a lot more people get indeffed at SPI than should, but that's just a personal view. Even though there is now a general procedures page, a lot of it is on discretion.

Of course some that are blocked are there for disruption but using my own personal experience that is not always they case. Several have been blocked and blamed as me that weren't. Many that were me have gone completely unnoticed. The other problem is that Checkusering has become a blank check for harassment. There are no rules prohibiting checkusers from checking any account they feel like or worse, not even both to check say its a checkuser block and not check at all.

The checkuser tool is garbage, prone to abuse, prone to error and for some reason has become trusted even when everyone from the programmers on down know that its a marginal tool at best.

[Wonderer]

Several have been blocked and blamed as me that weren't. Many that were me have gone completely unnoticed.

Exactly. Because it's based on secret evidence, no one else can examine it and say "Hey, wait a minute." Of course you can't put a "citation needed" on an SPI.

[Kumioko]

Several have been blocked and blamed as me that weren't. Many that were me have gone completely unnoticed.

Exactly. Because it's based on secret evidence, no one else can examine it and say "Hey, wait a minute." Of course you can't put a "citation needed" on an SPI.

You're right, and as long as only a handful of people can see it, nothing will change. Most of them wouldn't question it and even if they did won't question one of their own. Even if they did notice something out of the ordinary they wouldn't say anything for fear of people catching on to the abusiveness and untrustworthiness of the tool and the checkuser system. All someone would need to do would be to use any Windows machine with IE or Firefox and add a wikiproject tag or do an edit to one of the Medal of Honor recipient pages and they accuse me of doing it. Especially if they edit comes from the Massive Navy, Verizon or T-Mobile networks. Because you know, everyone knows that I am the only one editing Wikipedia from those networks that does any edits to Medal of Honor recipients or adds WikiProject banners.

[thekohser]

This whole discussion reminds me of the notorious, gun-packing, sock hunting expert Durova.

'member when she wrassled a Diamondback to the ground, then caught a mangy bobcat out of the corner of her eye and spit nails at it to drive it back into the brush?

Those were the days. The high-water mark for Wikipediot self-aggrandizement. Jorm came close with his "steel wool and vinegar" quip, but he was no Durova. Man, I miss her.

[Jim]

You're right,

Things that make you go wow. (You edited that, didn't you?)

[Poetlister]

The high-water mark for Wikipediot self-aggrandizement.

She was quite busy until March 2013, and then made just three more edits, in April 2013, April 2014 and December 2014.