Wikipedia:Wikipedia Signpost/2017-02-06/Special report

Special report

Wolves nip at Wikipedia's heels: A perspective on the cost of paid editing

Article creation by sock puppets

A 2013 sockpuppet investigation led to the determination that the initial version of the article had been created by a sock puppet. Scubadoofeck re-created the article, but was then banned as a sock puppet of User:Morning277. Morning277 had been a central figure in the 2013 Wiki-PR paid editing scandal, in which over 250 accounts were blocked or banned. The article has been deleted twice, has been the subject of at least three additional deletion requests, and has been the subject of extensive edit warring. A biography of the company's CEO was also deleted three times in 2013.

The Cyprus Securities and Exchange Commission (CySEC) licensed BDB as a broker in January 2013, allowing BDB to trade in many European Union countries. Though BDB was fined four times by CySEC, BDB kept its license until January 2017, when the company closed.

Edit wars

In June 2013 two US regulators, the SEC and the CFTC, filed civil suits against BDB and jointly issued a warning against the entire non-US binary options industry.

An external link to these civil suits was placed in the article two days later, prompting a protracted edit war. In May 2014, an editor nominated the article for deletion; that editor was blocked as yet another sockpuppet the next day. During the intervening 11 months, nearly 500 edits were made. An IP editor (traceable to Israel) claiming to be BDB CEO Oren Shabat Laurent, made five identical edits in the same day, all of which were reverted, to include highlights of Laurent's biography, and lists of products and countries served. He also reduced the coverage of the regulators' lawsuits and buried it at the bottom of the article. The article was protected due to edit warring, and vacillated between semi- and full protection until June 2014.

In May 2014, an editor observed changes in web search results relating to BDB, and asserted that the Wikipedia efforts might be part of a search engine optimization (SEO) effort to remove and de-emphasize coverage of the company's bad press from the web. The Times of Israel later noted, under a headline decrying the "wolves of Tel Aviv", that SEO "expertise has plainly been applied by fraudulent binary options firms, whose affiliated sites show up high in Google searches — sending unsuspecting and naive clients their way."

The article was kept after an Articles for Deletion discussion, despite the participation of User:BDBJack, who had declared he was working for BDB, and eight other editors whose comments were discounted as "obvious single-purpose accounts and socks".

User:BDBIsrael, later indefinitely blocked along with BDBJack, also stated that he was paid by BDB. In June he began relaying messages from the BDB board of directors, clearly stating that BDB had been a customer of Wiki-PR and naming several editors who either worked for BDB directly or through Wiki-PR. One of the editors who worked directly for BDB, User:Notsosoros was indefinitely blocked along with 37 related accounts for sockpuppeting. Most of these socks were apparently "sleeper accounts"—deceptive accounts being prepared to be used later—which hadn't yet edited the BDB article.

Word was posted on Wikipedia that BDB had advertised a five figure fee for "crisis management" of the article. This posting rallied Wikipedia editors to fight for an unbiased article.

A sock puppet started the editing in June 2014 with a request for speedy deletion [1]. 106 edits and six days later, the article was placed under full protection. On June 4 a well-known paid editor, who told the Signpost that he never accepted money from BDB, started another article deletion request, but the article was speedily kept. Three days later, he deleted 88% of the article text. The article talk page was even busier, with about 450 edits that month.

On June 16, 2014 the Wikimedia Foundation announced a new requirement for paid editors as part of Wikipedia's terms of use. Paid editors from that date forward were formally required to declare their paid status, their employer, clients, and other relevant affiliations, so that other editors could easily review their work. Paid editing at the BDB article slowed immediately. Only two banned or permanently blocked editors edited the article until 2016.

The collapse

2016 was a rough year for BDB. In January Israel changed its law to make trading binary options with its citizens illegal. In February BDB settled the US regulatory cases with $11 million in fines and restitution, and an agreement not to trade with or even indirectly solicit US residents. The Times of Israel ran a series of detailed exposés on the binary options industry, starting with "The wolves of Tel Aviv: Israel's vast, amoral binary options scam exposed" One article focused on BDB. The prime minister’s office condemned the whole industry, and the Knesset scheduled hearings on stopping all binary options trading in Israel. CySEC fined BDB four times and there were other regulatory setbacks in Australia, New Zealand, France, Belgium, and Belize.

This did not appear to stop the undisclosed paid editing on the BDB article. Four accounts, later banned or permanently blocked, removed content from the article in 2016: FoCuSandLeArN, Tianderni, Beranpolti, and Euclidthalis.

BDB fought hard to push its advertising into the article, and to remove news of its regulatory problems. Wikipedia editors and administrators fought harder and did a good job under the circumstances. Perhaps Wikipedia editors, myself included, could have been clearer: a reader skimming the article might have drawn the erroneous conclusion that BDB was a legitimate business with just a few problems with regulators. Of course we are constrained by what reliable secondary sources say – we could not have written, for instance, "BDB is a scam" unless several sources had printed that.

Other languages

Coverage of BDB in Wikipedia's German, Greek, Portuguese, Russian, and Spanish editions was less informative than in the English edition. All the articles were created from 2013–14. They appear to have been translated from one of the English article versions written by BDB; all but one had a small paragraph on US regulation buried at the bottom.

The creator of the Spanish article got his account globally locked as a "spam only account", but the article stayed pretty much the same until December 2016. The Greek and Russian articles stayed much as originally written. The German and Portuguese articles started in 2013 with word of BDB’s regulatory problems in the lede, but those sentences were deleted in June 2014.

The bottom line

A prolonged conflict around an article impacts Wikipedia's most valuable resource—volunteer time—and impacts the quality of information conveyed to its readers. All of these should be regarded as significant costs to the Wikipedia project. Furthermore, to whatever extent Wikipedia's readers take detrimental action based on faulty information from Wikipedia, those individuals share in the cost, as well.

In total, three editors of the English-language article were banned as sock puppets of Morning277, and 17 others were banned or indefinitely blocked. These editors all pushed BDB's point of view. A couple dozen more were blocked as sock puppets of a BDB employee before they could edit the BDB article.

With 781 edits to the article and 870 edits to the talk page, BDB likely consumed hundreds of hours of our editors' time. Eight related deletion requests, two massive edit wars, 20 banned or blocked editors, and years of page protection likely used hundreds more hours of administrators' time.

The Israeli binary options industry is reported to take in more than US$1 billion per year. BDB reported unaudited revenues of $100 million in 2014. How much of that came from Wikipedia readers can only be guessed. But given the time and money BDB spent trying to manipulate the article on Wikipedia, it must have been a very large amount. Maybe we should all count our kidneys, and keep our eyes on the wolves.

Next "Special report" →

In this issue

6 February 2017 (all comments)

Discuss this story

These comments are automatically transcluded from this article's talk page. To follow comments, add the page to your watchlist. If your comment has not appeared here, you can try purging the cache.

How is this different from the Gamergate controversy, or any other example of unpaid advocacy editing? When Wikipedia was founded, it opened itself up to attacks like this. The problem isn't paid editing, it's our tiny volunteer base. Serendi ^pod ous 11:57, 6 February 2017 (UTC)[reply]

It's not the tiny volunteer base. It's pretty much the entire Wikipedia community and the Wikimedia Foundation who constantly fail to accept how serious these issues are, and how they affect the encyclopedia's reputation. And their refusal to acknowledge and invest in the the need for more, policy based controls on who can edit, and who can patrol the new content. Many Recent Changes reviewers and New Page Patrollers are among the least qualified of all kinds of editors - over 20% of users who join WP:AfC do it to be able to be sure their own articles make it to mainspace.
The tiny volunteer base are the idiots (like me) who organise and do the cleaning up. Kudpung กุดผึ้ง (talk) 12:41, 6 February 2017 (UTC)[reply]

Thus far, the punishments for paid editing result in bans. What if the WMF chose to send "cease and desist" letters (and publicized such incidents) and if ignored, filed lawsuits against such editors? I know this would lead to a drain on the legal team's resources, but might WP's rules be taken more seriously? - kosboot (talk) 13:24, 6 February 2017 (UTC)[reply]

There are certainly steps WMF legal could take. The key step for the WMF, IMHO, is to widely publicize that we have strict rules against advertising and undisclosed paid editing. Individual editors can do "naming and shaming" pointing out where specific companies are inserting ads into our articles. Smallbones_(smalltalk) 14:39, 6 February 2017 (UTC)[reply]

You've got to get those editors to find out first who the culprits are before they can do the naming and shaming. Since its conception, Wikipedia has had a totally dysfunctional system for vetting new content and/or new users. Every common or garden village residents blog or forum has stricter rules than Wikipedia. Kudpung กุดผึ้ง (talk) 14:46, 6 February 2017 (UTC)[reply]

Re: "totally dysfunctional system for vetting new content and/or new users"—@Kudpung, I know you're into ideas... Do you have any recommendations on this front? Or is it under discussion somewhere? czar 18:42, 6 February 2017 (UTC)[reply]

It's been under discussion for 6 years but we've now arrived at a corner of a blank wall - on one side it's the WMF refusing to recognise it as a priority and on the othr is our volunteer community itself who insist on the 'anyone can edit' meme and continue to regard NPP as a hobbyhorse rather than a policy driven core function. As a result, I'm stepping down, but there is to be an election for people to carry on with the work I've been doing for the past 6 years. Kudpung กุดผึ้ง (talk) 22:18, 6 February 2017 (UTC)[reply]

The point of modern propaganda isn’t only to misinform or push an agenda. It is to exhaust your critical thinking, to annihilate truth.
— Garry Kasparov [2]

It takes a certain wherewithal to edit in high-activity topics, nevertheless with bad-faith actors. But it's insult to injury to then put our time-sinking editors through even more time sinks once it becomes apparent that (1) the actors are indeed bad-faith actors, and (2) the effort is designed to overwhelm our editing capacity. I hope there will be more discussions about article quality now that we're past the point of collecting garbage on every topic and in the stage of refining that garbage into that which is reliably sourced and that which isn't. To that end, it's worth collecting in one place stories like the one above as documentation for the many wasted hours of editor productivity. Juxtapose this situation with the other article in this Signpost on the women's march and how the editors chose to keep IP access because there were more productive IP edits than otherwise. I don't know the specifics of either situation, but that spirit of decentralized guild-like decision-making is emboldening. If our best editors are being overwhelmed in stewarding a topic, the best solution is not always the libertarian option of walking away and hoping someone else will fit in (nor is it to believe that particular stewards are the only ones worth considering), but the stewards should be able to appeal to the community for restrictions that will support their work in what becomes a war of attrition. I don't know how else to describe the above story, yet I especially don't know how to justify a volunteer sticking around for that kind of abuse once the problem gets that deep. Sometimes we fight for free/open in the wrong venues—the encyclopedia is free/open, but every individual article reflects the consensus of those who edit it, and those isolated instances may not be free/open. If Kudpung is correct in his assessment that both the community and admin do not treat these affairs seriously, I hope that in the future we could have more solidarity when our own editors are confronted with efforts so designed to exhaust them, because if those outside efforts succeed in their exhaustion, both our community and the encyclopedia will suffer. czar 18:42, 6 February 2017 (UTC)[reply]

I don't see the need for investigating editors with the goal of naming and shaming those deemed to be bad actors. Focus on content, not contributors. {{Request edit}}, which has been around since September 2007‎, is the editing mechanism for editors with an actual or apparent conflict of interest. That populates Category:Requested edits, which was set up back in September 2007. One can monitor the status of this backlog over time by checking the edit request table's revision history. How has that worked out, are COI-editors satisfied with the handling of that workflow? So the problem is that some COI-editors are either unaware of that workflow, or are choosing to bypass it. The solution for articles such as Banc De Binary, which is currently only semi-protected, is simple. Raise the protection to a new protection level called "COI-protection", so that only administrators and vetted editors who have been granted the privilege (analogous to template editors) may edit it. Administrators and COI-protected editors can then work the Category:Requested edits workflow. Problem solved? wbm1058 (talk) 18:45, 6 February 2017 (UTC)[reply]
- Just to be clear, I'm not calling for Wikipedia editors to be named and shamed. I'm calling for the businesses, the paymasters, to be named and shamed. It does little good to say User:BannedSock was engaging in paid editing. But saying that XYZ Inc.'s article was clearly promotional and infested by socks - that's another thing entirely.
- COI-protection is a very good idea. Smallbones_(smalltalk) 20:14, 6 February 2017 (UTC)[reply]
Re: "widely publicize that we have strict rules against advertising and undisclosed paid editing." You mean like, widely publicize that we have strict rules against illegal immigration and substance abuse? How's that working out? wbm1058 (talk) 18:55, 6 February 2017 (UTC)[reply]
- What have you got against telling people our rules? Everybody in the world should know that paid promotional editing is not allowed on Wikipedia. Smallbones_(smalltalk) 20:14, 6 February 2017 (UTC)[reply]

Everybody in the world should know that paid promotional editing is not allowed on Wikipedia - try telling that to the Wikimedia Foundation, Smallbones. In a video conference on the topic with the WMF two days ago I was accused by them of organising a 'pissing contest'. Kudpung กุดผึ้ง (talk) 23:10, 6 February 2017 (UTC)[reply]

There may be many mom & pop promoters who don't know the rules; sure, no problem with telling them the rules. I'd guess that all of the suspects editing Banc De Binary were already aware of the rules though. wbm1058 (talk) 21:44, 6 February 2017 (UTC)[reply]

Interesting case study. I fear, however, that there are many other cases where we have failed to stop spam/advertisement/COI/propaganda. I wrote about it a while ago at WP:CORPSPAM. Anyway, I thank you Smallbones for bringing this topic up, we need to be constantly on the vigil for more spam/COI editing. --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 19:06, 6 February 2017 (UTC)[reply]

An interesting and well-researched article on a desperately important issue. I'd compare it to developers who will gain by building something that will cause harm, such as a mine in a nature reserve. The developers have professional staff skilled in the process and paid to advance their cause by whatever means (and that can include disinformation or other forms of "political" deception). The nimbies are volunteers, organise in their own time and at their own expense, and may be outflanked in many different ways. Guess who's more likely to win. Anything that can help tilt the balance in favour of truth is to be welcomed.Chiswick Chap (talk) 10:53, 8 February 2017 (UTC)[reply]

I was surprised to see this brought up at such a late date. Most of this happened in 2014. Here's the history on AN/I: Wikipedia:Administrators'_noticeboard/IncidentArchive844. As a result of that, all the Banc De Binary editors and their sockpuppets were blocked. After that, further problems on Wikipedia were not difficult to handle. John Nagle (talk) 20:21, 1 June 2017 (UTC)[reply]

"Naming and shaming" vs. "assume good faith"

While "naming and shaming," as described in this piece and discussed above, has an important and legitimate role, it's an approach that should be used with caution. In some cases, "naming and shaming" can conflict with our assume good faith policy. One clear instance impacted a client of mine, who had diligently observed Wikipedia's policies in both letter and spirit. A columnist called out their clearly disclosed efforts, and Wikipedia founder Jimbo Wales repeated the (largely inaccurate) accusation, explicitly stating that shaming was in order. This prompted extensive discussion on Wikipedia; ultimately, no wrongdoing was found, but the accuracy of Wikipedia's content had already suffered substantial damage in the meantime. Beyond that, the call for "shaming" had negatively impacted our collaborative dynamics, with a variety of accusations, ranging from good faith but misguided comments to outright vandalism and harassment.

This is a dynamic I frequently encounter in less dramatic examples, when helping friends, teaching students, and advising clients. Shaming is a powerful tool; it has great power to advance our goals when used with care, and great power to damage our social dynamics when used recklessly. -Pete Forsyth (talk) 22:21, 6 February 2017 (UTC)[reply]

There is a certain truth in what you are saying: Naming and shaming, while necessary to set an example, adds more to the negative publicity that puts the accuracy of Wikipedia in question. The encyclopedia is already being lampooned by the serious newspapers and TV in the globally powerful and respected UK media. So, Peteforsyth, what do you suggest we do about it? Perhaps next on the Signpost agenda should be an article about what's been done, actually being done, what's being attempted, and what's deliberately being ignored and how the community is polarised and utterly divisive about it. Kudpung กุดผึ้ง (talk) 23:06, 6 February 2017 (UTC)[reply]

Obviously, I can't assume good faith when it comes to the editors employed by Banc De Binary. I stand behind my decision to name and shame Banc De Binary. There very well may be better ways to deal with others, but we are not using them. For the "mom and pop" coffee trucks that have articles on Wikipedia, I think a clearly stated publicity campaign would do the trick. For those in between these two extremes other methods should be used. We just have to get off our duffs and do it. Smallbones_(smalltalk) 01:21, 7 February 2017 (UTC)[reply]

Smallbones, yes, I think that is obvious -- so much so that I neglected to say so explicitly :) I agree, your story offers a clear illustration of a case where "naming and shaming" is appropriate. My comment was intended to inspire caution around the idea of generalizing that the approach is a good core strategy for dealing with undisclosed -- or, especially, disclosed -- paid editors.

To Kudpung, I think the general approach has to be that we continually encourage each other to take care and exercise good judgment in specific cases, rather than trying to find a "one size fits all" approach. I would hypothesize that part of the problem is, when a highly prominent figure like Wales advocates naming and shaming, perhaps some of our younger or less experienced Wikipedians hasten to find instances where they can take part. Whether conscious or unconscious, the idea that we can gain social status by accusing and embarrassing other users is IMO something we ought to be very cautious about encouraging. I believe the stance Wales takes frequently and prominently around COI editors is at odds with Wikipedia policy and works against our parallel efforts to improve our social dynamics and reduce harassment and hounding.

I don't think there's a simple answer, but I do think the right answer includes acknowledging that these situations are rarely as simple as they might have been with BDB. -Pete Forsyth (talk) 02:12, 7 February 2017 (UTC)[reply]

When you find something boring or unnecessarily time consuming, just walk away. We are volunteers, no one forces you to investigate or patrol anything.--Catlemur (talk) 16:44, 7 February 2017 (UTC)[reply]
- Perhaps I'm misunderstanding you. You seem to say that when you see a Wikipedia article turned into an advert for a company that intentionally rips off at least $100 million a year from innocent people, we should just walk away. If that's what you are saying, I'll suggest that you should just walk away from Wikipedia. Smallbones_(smalltalk) 18:32, 7 February 2017 (UTC)[reply]
Well I have an idea that I keep thinking about but it would not solve everything. Just make all (good) edits paid edits is my idea. Make it micro-payments that add up, and then editors could keep the payments or donate them back to the project. Maybe trying this as an experiment on a small scale to see if doing something like this would divert editors, paid editors from competing with each other for higher paying requests? Even though it has been talked to death, with the advent of "fake news", (which is a very real problem), new vigilance is needed to protect Wikipedia. At some point it is going to get ugly and astroturfing/spam is going to probably become a legal problem where it is not clearly disclosed. I think that WP needs to get on top of it by forcing disclousure as-if that time was already here.TeeVeeed (talk) 16:58, 7 February 2017 (UTC)[reply]
- One problem here is that BDB and other scam artists can pay much more than micropayments. Reports by the UK police put the average loss to individual customers at 20,000 pounds. For all language versions there were about 300 page views per day. If only 1 victim per day bit on BDB pitch on Wikipedia ... well, I'll let you do the math. Clearly they can put a lot of money into manipulating the article. Smallbones_(smalltalk) 18:32, 7 February 2017 (UTC)[reply]
- On this point, I think the software should let editors collect public donations for our work through the site, or at least have an explicit policy for allowing external links for such. If volunteer software developers can do this, why can't we? Oftentimes our work is plenty complex, and that work (sometimes) deserves more than 'thanks' or barnstars. I don't think this would directly solve the "paid editors" problem, though, but it might generate more volunteer editing that would work against them. Stevie is the man! ^{Talk • Work} 18:57, 7 February 2017 (UTC)[reply]

There are a few reasons why offering paid editing for all productive editing would have an advantage over BDB-types. Obviously if someone is asking you to do something sneaky or risky they cannot be trusted to fulfill any bargins. Someone who is looking to make $$ editing WP, if given a choice between working for bad actors or working for the project, might be more inclined to devote their efforts to the project.TeeVeeed (talk) 19:14, 7 February 2017 (UTC)[reply]

Weird how this article made no mention of the Wikimedia Foundation's clumsy financial support of paid promotional editing when the Stanton Foundation director was assisted by the WMF in sending money to a paid editor in support of the director's husband at the Belfer Center. That was a $50,000 project, if I recall -- very few paid editors on Wikipedia ever make that kind of money; yet, with the Wikimedia Foundation's backing, it's entirely doable! Initial attempts to "name and shame" the Foundation for that disaster were silenced by Wikipedia admins and WMF staff. - 108.216.213.21 (talk) 15:20, 9 February 2017 (UTC)[reply]
- I find it weird that anybody could possibly equate the Belfer Center, a unit of Harvard University, with Banc De Binary, a broker that's been accused of scamming innocent people out of their pensions. One of the most obnoxious things that paid editors do is harass other editors. Checking your recent edit history it looks like you are harassing Jimbo Wales. May I ask if you are the guy who was banned 10 years ago for paid editing by Jimbo, and who has been harassing him ever since? Are you still running an advert on your own website for paid editing services? Smallbones_(smalltalk) 17:29, 9 February 2017 (UTC)[reply]

No, what IP 108 suggests is absolutely off base in its implications, even if some of the facts are accurate. I was, over time, the most consistent and vocal critic of the WMF on this 2013-14 incident, and was centrally involved in it -- I know this episode well. A few points:

We're talking about disclosure, so the choice to post anonymously is more than a little ironic.
The Wikimedia Foundation did not intend for the position to be a paid editing position; that happened due to a lack of oversight and some naive decisions, not due to a desire to "send money to a paid editor." That's an oversimplification, and a significant one.
Neither the WMF nor admins "silenced" anyone (nor could they have); in fact, they did the opposite. The first public posting about the issue was from odder on his personal blog (which he has since deleted). Following some open and candid email list discussion, the WMF's then-executive director Sue Gardner and her deputy Eloquence proactively created a thorough report on the incident, on a public wiki, inviting input and feedback. And afterwards, KLove (WMF) elaborated on what lessons the WMF had learned from the episode, and how it has adapted its grant programs accordingly.

This is an example of a fairly isolated mistake that was handled with care, and in order to interpret it as indicative of the WMF's disposition in 2017, you'd have to ignore a lot of important stuff. -Pete Forsyth (talk) 18:49, 9 February 2017 (UTC)[reply]

Readers of this article might also be interested in Wikipedia vs. Banc De Binary: A 3-year battle against binary options ‘fake news’

Smallbones_(smalltalk) 23:17, 29 May 2017 (UTC)[reply]

Get the latest headlines on your user page – just add {{Signpost-subscription}}.

Home

About