Privacy violations by a Checkuser in the block log, a permanent record, WP community supports outing in a classist way Options

[SomeRandomAdmin]

Remember what I said earlier about the WP community being
so prideful that y'all cannot accept that you are not infallible?
You are proving my point. An MO gives you a probability, nothing more.

You are quite right; however a recognition of an MO gives you more of a clue of how to approach the issue. Despite the fact that the LTA page says that suspected HeadleyDown accounts can be blocked on sight, in my experience they are pretty much recognized by behaviour and then blocked (or not, in some cases) after more detailed edit analysis and (certainly on the accounts I've looked at) a Checkuser. (Aha! Durova-style Sooper-Sleuthing!). You'd have to ask FT2 and others for details of all the accounts on the LTA page, but I'd be surprised if anyone has ever been blocked as a HD sock when they weren't, and if they have I'd be interested to see the evidence.

[guy]

HeadleyDown is a fairly well-established compendium of socks, meatpuppets and editing from proxies

I'd be surprised if anyone has ever been blocked as a HD sock when they weren't

If Headley Down is a consortium of umpteen people, it's impossible to prove that any given person sympathetic to their views is not in the consortium. Indeed, you could argue that they are in it by definition.

[SomeRandomAdmin]

If Headley Down is a consortium of umpteen people, it's impossible to prove that any given person sympathetic to their views is not in the consortium. Indeed, you could argue that they are in it by definition.

I must use that as the note in my next block "Unable to prove that editor isn't HeadleyDown". Seriously though, note the use of CU, as I mentioned. HD socks last so long not because they are not disruptive, but because (as the LTA page points out) they are often so subtle (including taking opposing viewpoints and arguing against each other on a certain article) that they don't get CU'd until recognized.

[AB]

You are quite right; however a recognition of an MO gives you more of a clue of how to approach the issue.

Well, if overconfidence was your first mistake, then that is
your second mistake. Sockpuppetry investigations should
be as double-blind as possible. It is imperative that a
person who has collected one piece of evidence not share
with a person collecting another piece of evidence.

The human mind has an incredible ability to see patterns.
Andromeda in the stars, man in the moon, canals on Mars,
that sort of thing, Virgin Mary in the windows, that sort of
thing.

Being told to expect something in particular makes the
problem even worse. An investigator looking for Loch
Ness can look at a log and see a sea monster.

See Believing is Seeing by Robert Novella, Assistant Editor,
New England Journal of Skepticism, Volume 4 Issue 1.

but I'd be surprised if anyone has ever been blocked as a HD sock when they weren't, and if they have I'd be interested to see the evidence.

Mathematically, I can show you that you can never be
100% certain. Assuming a random distribution of MOs
throughout the IP range, if you have a 30% chance of
two users being sockpuppets of each other based on
MO data and a 70% chance based on IP data, that
does not add up to a 100% chance. Rather, it is a
.7 + .3(1.0-.7) = .7 + .3(.3) = .7 + .09 = .79 = 79%
chance.

Now, if you block 100 people who have 79% chances
of being sockpuppets, and each time you publicly give
away their IP addresses and say they are sockpuppets,
that's 100 privacy violations and an estimated 100-79=21
defamations.

Of course, you don't even know what probabilities you
are dealing with, but you get the idea. Since nothing
can give you 100% proof of guilt, it's like that old idea
where you go half way, then half of the remaining way,
and never make it all the way there.

Besides, MOs are not evenly distributed throughout the
IP range. Family members are quite likely to share MOs.

This post has been edited by AB: Mon 7th January 2008, 12:33am

[Amarkov]

If Headley Down is a consortium of umpteen people, it's impossible to prove that any given person sympathetic to their views is not in the consortium. Indeed, you could argue that they are in it by definition.

I must use that as the note in my next block "Unable to prove that editor isn't HeadleyDown". Seriously though, note the use of CU, as I mentioned. HD socks last so long not because they are not disruptive, but because (as the LTA page points out) they are often so subtle (including taking opposing viewpoints and arguing against each other on a certain article) that they don't get CU'd until recognized.

As some checkusers have been trying to tell you guys for a while now, checkuser is not magic pixie dust. It has a possibility of error, even when you get a result of "confirmed". Therefore, if the only evidence you have is a checkuser result, you do not have nearly sufficient evidence.

[AB]

As some checkusers have been trying to tell you guys for a while now, checkuser is not magic pixie dust. It has a possibility of error, even when you get a result of "confirmed". Therefore, if the only evidence you have is a checkuser result, you do not have nearly sufficient evidence.

Long, technical explanation of why up at the bottom of that post.

[AB]

Hell, is no one else infuriated by the rampant
privacy violations and defamations?

[Amarkov]

Hell, is no one else infuriated by the rampant
privacy violations and defamations?

Outrage exhaustion. We found out that Wikipedians have no problem with stealing from banned users, so there's not much outrage left for their engaging in privacy violations.

[AB]

Outrage exhaustion. We found out that Wikipedians have no problem with stealing from banned users, so there's not much outrage left for their engaging in privacy violations.

Hmmm... have they ever run out of outrage about the
(often reciprocal) outing that occurs over here?

Well, I guess it's easier for them, since they have more
people to spread it out between.

They're hurting themselves, really. 'twould be easier
for me to argue that y'all shouldn't out people if they
stopped violating people's privacy in such a proud,
routine manner.

[Docknell]

This set of IPs is interesting

http://en.wikipedia.org/wiki/Wikipedia:Lon...use/HeadleyDown

Clearly HeadleyDown comes from

24.147.97.230
Massachusets
66.176.129.11
Miami
80.189.81.19
London
203.186.238.160
Hong Kong
62.25.106.209
London
And so on

HeadleyDown is ubiquitous? In fact, it looks like FT2 is identifying anyone and everyone in disagreement with NLP, zoophilia promotion and anti-pedophilia. Its not so much an outing of socks, rather an outing of FT2's admin abuse.

HeadleyDown is a fairly well-established compendium of socks, meatpuppets and editing from proxies, and his/their MO is fairly obvious once you get to recognize it. There's a current kerfuffle on an article which has hit AN/I a number of times which is almost certainly one of his/their playgrounds. (And not an NLP, zoophilia or pedo article )

There is enough blatant dishonesty and distortion from FT2 on that list of editors to assume FT2 has conflated it for reasons of self-interest. I have no doubt that HeadleyDown is using a lot of accounts, and continues to edit. From the diffs, its clear that the edits are pretty sound and are working to present facts that FT2 does not want to have on certain articles of unhealthy interest. FT2 seems to add random personal attack accounts and IPs to the list in order to keep a high degree of paranoia over the articles FT2 wants to protect from fact presenting editors.

[LamontStormstar]

See this log.

You will note several occasions when IP addresses were blocked as
sock puppets of particular users, thereby outing the IP addresses
of those users. There were also occasions when pseudonymous
accounts were blocked as sockpuppets of legal names, thereby
outing the real names of those pseudonyms.

From this, we can conclude that the WP community as a whole does
not actually oppose outing. They are simply classist - they protect the
privacy of those with higher rank while sadistically violating the privacy
of those with lower rank. (As disclaimer, I am sure there are many
members of the WP community who do not agree with the community
as a whole on this.)

I don't see anything on the first page of 500 but I think one instance where he said it was a proxy blocked as a sock. Also I have found over the years, Morven is one of the more better checkusers in my opinion.

There are othercheckusers who tagged an IP as Jon Awbrey even though the IP never edited. Why can't checkusers keep this private?

Then there's how vandal pages had lots of IPs listed. Here is the worst one now http://en.wikipedia.org/wiki/User:Jayjg/Alberuni -- this is way worse than the one similar one someone hosted on their user subpage and got deleted because it wasn't in project space. But this one http://en.wikipedia.org/wiki/User:Jayjg/Alberuni is done by Jayjg so Jayjg gets his way. Alberuni I think only edit wars and never vandalizes, but Jayjg marks all IPs the guy has used.


This post has been edited by LamontStormstar: Mon 7th January 2008, 3:56am

[AB]

I don't see anything on the first page of 500 but I think one instance where he said it was a proxy blocked as a sock.

Should I point them out, then?

I feel like I'm outing people for a second time. Dates, then.

Occasions when IP addresses were blocked as sockpuppets
of particular users:
20:58, 25 February 2006
21:26, 25 February 2006
03:14, 27 August 2007

This post has been edited by AB: Mon 7th January 2008, 8:25am

[LamontStormstar]

I don't see anything on the first page of 500 but I think one instance where he said it was a proxy blocked as a sock.

Should I point them out, then?

I feel like I'm outing people for a second time. Dates, then.

Occasions when IP addresses were blocked as sockpuppets
of particular users:
20:58, 25 February 2006
21:26, 25 February 2006
03:14, 27 August 2007

Hmm I think that might be past the first 500 and require digging. It took me a long time to search through the first 500 just that.

[AB]

I don't see anything on the first page of 500 but I think one instance where he said it was a proxy blocked as a sock.

Should I point them out, then?

I feel like I'm outing people for a second time. Dates, then.

Occasions when IP addresses were blocked as sockpuppets
of particular users:
20:58, 25 February 2006
21:26, 25 February 2006
03:14, 27 August 2007

Hmm I think that might be past the first 500 and require digging. It took me a long time to search through the first 500 just that.

It all fits on one page for me.

Does your browser have a search function? Most do. You
should be able to activate it with / or Ctrl+F, then find future
occurrences of the phrase you are looking for with n, F3, or
Ctrl+G. (Which keys will work depend on your browser. And,
if it's a graphical browser, you could also use the menu.)

I suggest the phrase 'sock'.

Alternatively, you could skim for IP blocks and then check
their blocking reasons.

[Cynick]

As far as I can see, admins with CheckUser permissions, have no accountability or transparency. This is unacceptable since personal IP addresses have clearly been revealed, in contravention of Wikimedia's privacy policy.

While the details of the result of a checkuser should be private, USE of Checkuser should be logged, publicly, with an indication of WHO performed a CheckUser, when, WHY, and on whom. The type of results returned should also be given, showing whether they are conclusive, negative, or how they have been interpretted.

I'd also like to see a requirement for all CheckUsers to require a second Admin to confirm a check, with both names logged.

[Lar]

As far as I can see, admins with CheckUser permissions, have no accountability or transparency. This is unacceptable since personal IP addresses have clearly been revealed, in contravention of Wikimedia's privacy policy.

While the details of the result of a checkuser should be private, USE of Checkuser should be logged, publicly, with an indication of WHO performed a CheckUser, when, WHY, and on whom. The type of results returned should also be given, showing whether they are conclusive, negative, or how they have been interpretted.

I'd also like to see a requirement for all CheckUsers to require a second Admin to confirm a check, with both names logged.

One argument advanced against this is that in some cases, even revealing that there is a check can out someone, and therefore that some checks should be kept private for the protection of the innocently accused. I can give an example of how that would work if you don't follow it.

I want to reiterate something here that I said, but that think has gotten lost. I don't think Checkuser is perfect. I am very reluctant to mark anyone as "confirmed", and if I do, I try to remember to caveat that as meaning "it's our highest probability judgement, but it's still not a certainty". I don't think I am perfect either, and I welcome people trying to help me be better/fairer/more respectful of privacy concerns/etc at what I do.

To the log issue I confess I'm at a loss as to how log things in such a way as to avoid a determined analyst making a connection between IPs and users. I'm open to suggestion as to how to log things going forward but I'm not sure the suggestion to omit mention of socks completely is workable, that is, it doesn't solve the correlation problem.

Consider this somewhat simplistic (but fairly realistic) scenario: Serious vandalism is discovered, and editing patterns strongly suggest that vandal users A, B and C might be related. When a check is run, it's discovered that they all edit from the same IP address (or small range), and that the only other users have ever edited from that address on that wiki are also vandals D, and E not yet discovered as being associated with A, B and C, and sleeper accounts F, G and H, which have never yet edited. I choose not to block the IP at all to avoid association, and to avoid any possible collateral damage. Blocking just A B and C without blocking the IP subsequently results in creation of new vandals I, J and K, more vandal activity by D and E (arguably they should have been blocked the first time), and (on investigation, discovery of creation of 20 new sleepers) as well as activation of previous sleeper accounts F, G and H, so the case is revisted. It's now clear that the IP needs to be blocked, as the IP is clearly not the source of anything useful at all. If I now block F, G and H and the IP at the same time, since F, G and H carry tagging showing them as socks of A, B and C, anyone looking at the block record is going to deduce that hey, the IP must go with F, G and H (and by extension A, B and C since both are tagged) because it was the only IP blocked by Lar at the same time F, G and H were.

Not blocking the IP isn't an option. (maybe it's a small range rather than a single IP so autoblocking isn't going to work).

What, in this case, should be done? As I say I am very very open to suggestions and to changing what I say in block messages, and if it makes sense, lobbying for other CUs to adopt the change. So tell me what you think should be done?

[GlassBeadGame]

What, in this case, should be done? As I say I am very very open to suggestions and to changing what I say in block messages, and if it makes sense, lobbying for other CUs to adopt the change. So tell me what you think should be done?

Again, why do Wikipedians insist on addressing this on such a low level? Require IRL identities for editing. Either allow a trusted outside entity to provide exceptions for editors with well founded fears or just do not create any expectation of privacy at all. Make participation in the project a public act. IRL identity for editing is the single, easiest reform that would profoundly improve Wikipedia. It would end most vandalism, end the MMORPG like "patrols," frustrate POV pushing and provide a modicum of respect for people with real credentials. I believe WP admins will fall over themselves with "how can I make it better" nonsense and can not address a true solution because it lessen their influence in the project.

[Lar]

What, in this case, should be done? As I say I am very very open to suggestions and to changing what I say in block messages, and if it makes sense, lobbying for other CUs to adopt the change. So tell me what you think should be done?

Again, why do Wikipedians insist on addressing this on such a low level? Require IRL identities for editing. Either allow a trusted outside entity to provide exceptions for editors with well founded fears or just do not create any expectation of privacy at all. Make participation in the project a public act. IRL identity for editing is the single, easiest reform that would profoundly improve Wikipedia. It would end most vandalism, end the MMORPG like "patrols," frustrate POV pushing and provide a modicum of respect for people with real credentials. I believe WP admins will fall over themselves with "how can I make it better" nonsense and can not address a true solution because it lessen their influence in the project.

I'd like to think I don't really care about influence, or power, or whether everyone fears me or respects me or likes me or whatever. I can't say that's always true, but I do think I at least try not to care... (after 25 years in online communities I think I've gotten better at not caring, I'd say), and I try very hard to do the right thing if I possibly can, because I care about doing the right thing.

Personally, I like the idea of real names only. My real name is out there for all to see, after all. I wish WP was set up that way and wish it had been so from the very start. It might cost WP some very good editors to be sure, but on balance that's my preference.

But it's not set up that way. And I'm not going to stop contributing to WP because of that. I got a Citizendium account but I haven't contributed much there, I think it has different challenges.

Given that larger context, I'm asking how others think, specifically, I should do things differently in a specific limited context. If no one gives me any constructive suggestions I can't well change how I do things or lobby for others to change, can I? If you don't want to give input, that's fine, but I do want to make that point clear. Remember I'm here at least in part to understand valid concerns and try to where possible act on them.

Within the framework of "WP has anonymous users" and "I'm a CU and a steward and I have no plans to stop being either of those things", if you have things I ought to try to do differently I want to hear about them.

(To draw a real world analogy, I personally believe the US should abolish the income tax and drastically cut the size of government (libertarian macho flash warning!!!), but I'm a realist and don't expect that to happen any time soon, and despite that belief, I'm not going to move out of the US in protest... but I can and will solicit advice about how to improve my tax filing ability or how to maximise my deductions, and I don't see that as a compromise of my principles to do so)

[GlassBeadGame]

Given that larger context, I'm asking how others think, specifically, I should do things differently in a specific limited context. If no one gives me any constructive suggestions I can't well change how I do things or lobby for others to change, can I? If you don't want to give input, that's fine, but I do want to make that point clear. Remember I'm here at least in part to understand valid concerns and try to where possible act on them.

This type of activity is better suited for an internal WP process than a critic site.

[Lar]

Given that larger context, I'm asking how others think, specifically, I should do things differently in a specific limited context. If no one gives me any constructive suggestions I can't well change how I do things or lobby for others to change, can I? If you don't want to give input, that's fine, but I do want to make that point clear. Remember I'm here at least in part to understand valid concerns and try to where possible act on them.

This type of activity is better suited for an internal WP process than a critic site.

Perhaps I'm wasting my time here then? I'd like to think not...