Online investigation. How to find “grids” in Telegram

The topic of the rubric is what indicates the connection between telegram channels and what conclusions can be drawn from this

THIS MATERIAL (INFORMATION) IS PRODUCED AND DISTRIBUTED BY A FOREIGN AGENT OF THE BELL OR IS RELATED TO THE ACTIVITIES OF A FOREIGN AGENT OF THE BELL. 18+

Since last year, the state corporation Rostec has been fighting on two fronts: in addition to supplying weapons to the Russian army, it is trying to take control of Russian-language content in the Telegram messenger, whose audience is growing rapidly after the blocking of foreign social networks in Russia. This week, The Bell and Meduza* went into detail about how this happens. And in the “Online Investigation” section, this time we will explain what “grids” of channels are and why they are calculated.

🔎 Much of the investigative journalism is based on open data. In the "Online Investigation" section, we talk about online investigation methods available to everyone. We want to show you how you can check any questionable information with their help. You can train with the help of our “homework”, and send wishes on topics for the future to the author of the column, The Bell special correspondent Irina Pankratova .

* Meduza is recognized as an undesirable organization in the Russian Federation

Tools - tgstat

There are no free methods and services for tracking "grids" of telegram channels. This can only be done using the tg stat or telemetr tools, which we have already covered in detail here . Each service has a set of free functions, but the analysis of connections between channels is definitely not included in them.

Let's tell you how to track the "grid" of channels if you have a subscription to one of the services using tg stat as an example.

- There are two main ways to track the grid: reposts (quoting) and the chronology of the appearance of the same publication in the channels.

— It’s easier to research citations: you need to find a channel by its name through the search, click on the “Statistics” button and select the “Citation” section in the column on the left. In it you will see who is quoting this channel and whom it is quoting. Ignore the ugly picture and click on the "All Mentioning Channels" sections. There you will find the 10 channels that most often mention the channel you are studying.

- The second way is to try to track the distribution of a particular post. To do this, you need to enter a sentence from it in the search for publications and scroll down their list to the very first one where this sentence occurs. It is better to choose a sentence containing some kind of characteristic turnover, the name of the company or the name of the person. If you know approximately when this post was published, you can set a time period to narrow down your search immediately.

Examples of using

Let's look at the Telegram channel of the head of Chechnya, Ramzan Kadyrov, one of the largest in terms of the number of subscribers. Most often, Kadyrov's posts are quoted by a small anonymous channel with 13,000 subscribers "Za Nashi" - more than 900 (!) references in total. And who reads and quotes this channel? It is easy to find that almost no one: the channel's posts gain a maximum of a couple of hundred views. From this we can conclude that 13 thousand channel subscribers are probably bots. Posts from the Za Nashi channel are reposted mostly by the same anonymous empty channels. Even more actively than Kadyrov, Za Nashi quotes the channels of the Ministry of Defense, Operation Z: Military Correspondents of the Russian Spring and Izvestia - each more than 2,000 times, as well as Russia Today - more than 1,500 times.

Going through the list of channels quoting Kadyrov's channel, we will find a dozen such examples. At the same time, even the names of the channels are of the same type. For example, another channel quoting Kadyrov is called “News from the Z|O|V Front”. He has 4.3 thousand subscribers. In addition to Kadyrov, he most often quotes the same channels of the Ministry of Defense, "Operation Z: military correspondents of the Russian spring", "Izvestia" and Russia Today. The same picture with the channels "PMC ZOV", "Work brothers Z", "Nikolay Z Ganchuk" and many others. If a post appears in one of these channels and spreads like a fan through the same citing channels, then we can assume with a high degree of probability that all of them are included in a specially created “grid”.

Another typical example of how networks work is provided in an investigation by correspondents from Meduza and The Bell about how Rostec is crushing Russian Telegram. While the investigation was being prepared, a dozen anonymous telegram channels published fan-shaped posts over the course of a day that Meduza was allegedly going to release a text about the work of Vasily Brovko, a top manager of Rostec, on the “order” of Ksenia Sobchak. On the eve of the publication of a series of these posts, journalists were just taking several interviews with Brovko's acquaintances. All the small anonymous channels that distributed this post ( "Image of the Future" , " Redacted P6 ", " Ask Rasul ", " Telestream Z ", " Informer ", " Oper writes ", " What today? ", " Z|Localka ”,“ Vse obo vsem ”and others), are regularly quoting posts from the political channel with “stuffing” Nezygar.

Precautionary measures

• The first sign of questionable content is posting in an anonymous channel. We described one of the simple and quite effective ways of deanonymization here .
• The second sign of questionable content is a large difference between the number of channel subscribers and the number of views of posts (a small number in the lower right corner of each post). If the channel has under half a million subscribers, and less than 1000 people watch the posts, then most likely the subscribers are bots.
• If you or your company have been the subject of negative telegram posts, before you panic, check if there was any information published in these bots . Gathering information from them is not a tricky business, but it is often used for manipulation and fraud.

What do I get from this?

The ability to check how a particular post diverges along the "grid" of channels usually helps answer several important questions: which channel was the original source, which channels helped spread this information, can they be trusted, does the dissemination of information look like a planned campaign. If you or your company have become the subject of a series of negative posts in dozens of anonymous channels, then a thorough study of the "grid" can help determine the interest of the publications.

Homework

Next week, the editor-in-chief of Novaya Gazeta and Nobel Peace Prize winner Dmitry Muratov was supposed to hold an open discussion “The conflict of generations after February 24” with students of the Moscow Institute of Physics and Technology (MIPT). But on March 23, the rector of the university, Dmitry Livanov, canceled the meeting, citing the impossibility of “depoliticizing the conversation” and “the risk of external provocations.” Muratov linked the cancellation with criticism from far-right nationalists that appeared the day before. On March 22, more than a hundred of the same type of posts with criticism and even threats against Muratov were indeed published in various channels. Set the author of the first post, and also, if possible, the channels from the "grid" around this telegram channel.

What new journalists learned from open data for the week

- The American magazine Wired, together with the organization Lighthouse reports, studied the algorithms for providing social assistance in the Netherlands using the example of Rotterdam. The text of the investigation is under paywall, but its brief retelling is open with a description of the methodology of the investigation (in English). Judging by the description, according to Dutch law, information about the conditions for granting benefits is incredibly transparent and even reveals markers, on the basis of which social workers predict fraud and carefully check certain applicants. Journalists call a number of markers discriminatory: for example, gender, number of children and nationality increase the “fraudulent prediction” regarding the welfare applicant.

— The European Association of Investigative Journalists compiled statistics on murders and violence against women in Europe (in English and other 10 foreign languages). Journalists point out that cross-border data on this type of violence were not available and had to be collected country by country. A surge in violence against women was recorded in Greece during COVID. Journalists also cite examples of the negligence of law enforcement officers, which led to the killing of women a few hours after they contacted the police.

- Eighteen European media have teamed up for a large-scale environmental study of Europe's pollution with the so-called "eternal chemicals" PFAS, which are found in literally everything from disposable cups to children's toys. The results of the project under the general name Forever Pollution Project are published by the French Le Monde (in French, but right-clicking on the screen and choosing "Translate into Russian" even translates texts from infographics), the project also has its own website .