Introduction:
The Qanon movement claims that there are over 60k sealed indictments currently sitting in the Federal court system, which are intended for mass arrests on the 'deep state'. I have previously written an article that debunks that claim, which can be found here.
This article will show exactly who started this claim, and how it evolved from a 4chan screenshot of 2 sealed indictments, into the 60k sealed indictment chart we have today.
TL;DR
In order to fully explain the context/timeline/conclusions, this article became quite lengthy. If you want the short answer it is this: Twitter user @Avery1776.
Timeline short version:
October 2017
Robert Mueller/Louise Mensch make headlines due to sealed indictments.
4chan starts discussing sealed indictments.
Avery1776 runs a search for indictments, misreads the results, and falsely claims she has found multiple sealed indictments. She proceeds to make daily updates to twitter, with additional non-indictments.
Those updates go viral, more twitter users join in to help Avery, and it eventually snowballs into the 60k claim.
Context:
In late Oct. 2017, sealed indictments were a hot topic in the media. Louise Mensch had made waves claiming that there were dozens of indictments for Trump and his staff, and Robert Mueller was making headlines with his recent indictments against Paul Manafort and George Papadopoulos.
Papadopoulos's indictment had been sealed, which caused discussions about sealed indictments to increase dramatically when it was unsealed.
On Oct. 30 2017, CNN quoted sources claiming that Mueller had filed new sealed indictments related to the Russia collusion probe that he was overseeing.
Suddenly, everyone was digging into the court records to try and confirm this report, and to uncover more information about the situation. Court filings are available to the public through the PACER online database. Although generally only known to lawyers/paralegals and legal reporters, all the news reports and social media discussions about sealed indictments, quickly informed the general public of PACER's existence.
For example, this was a tweet from a well-known reporter on Oct. 30, showing PACER search results of sealed cases, which prompted a large amount of discussion (17k likes, 11k RT):
All of the discussion taking place in the media, prompted political junkies from both sides of the aisle to start doing their own research into sealed indictments, using the PACER database.
That sets the stage for our discussion here.
Analysis:
On November 4, 2017, a screenshot of a PACER search was posted on 4chan, which showed 2 sealed documents in the US District court of DC:
Notice in the search settings they had selected to only display documents with the case type 'cr' (criminal), which is the specific category where indictments are filed. This means that the 2 sealed documents could be indictments, although they could also be any other type of document with the 'cr' case type in the DC district.
This image started being circulated heavily online, with all sorts of speculation about who the 'indictments' were for. Later that same day, a screenshot was posted on 4chan, of another PACER search from that same district.
This time it showed 12 sealed documents:
Note the search criteria.
Instead of correctly filtering the case type only for 'cr', where the indictments would be, whoever ran this search left that setting on 'All'... which meant that it was now displaying EVERY type of document in that district.
Because this is a rather nuanced detail, and because most people aren't familiar with PACER, this went almost completely unnoticed. People now incorrectly believed that 10 additional sealed indictments had been filed that same day, when the truth was that it was still only the original two (17-206 and 17-207).
Both of these screenshots began circulating heavily on twitter, on Nov. 4 2017. There is no identifying information on them, because whoever performed the searches removed their user id information.
On Nov. 5th, 2017, a link to a full PACER file began circulating on twitter. This file contained the original PACER document for the search that showed 12 sealed documents in the DC district.
This time, the person did not remove their PACER login information, and I noticed that the name matched someone who is credited on the current chart that makes the 60k sealed indictments claim... Avery1776.
When I started researching Avery's tweets from November of 2017, things quickly started falling into place.
@Avery1776:
On Nov. 4, Avery became aware of the 4chan 12 sealed document screenshot that I previously discussed. She tweeted that screenshot 4 times, all within a few minutes.
On Nov. 5th, she decided to start running her own searches in PACER, and she also uploaded the full document.
This is a very critical post.
If you examine the settings that she used for this search, you will see that she copied the same mistake that the person who performed the previous 12 document search made... she didn't filter case type by 'cr', SHE INCORRECTLY LEFT IT ON 'All'.
This set the precedent for all subsequent searches that she and others would perform.
Later tweets will show that she did not understand how the case type classification system worked, so she didn't know she should have filtered for 'cr', in order to only search the category where indictments are filed.
As it was, she was now searching for EVERY type of document in the criminal court system, and incorrectly believing that they were all indictments.
She proceeded to start making daily searches/updates to twitter. The next day (Nov. 6) she reran the DC district, and it now showed 17 sealed cases. She made a tweet and uploaded the file to twitter.
Again: If you open the linked PACER file, the 2 original sealed criminal cases (17-206 and 17-207) are still the only possible indictments on this list... the remaining 15 are sealed NON-indictments.
Because of her misunderstanding about what the results mean, she reports that the count has now jumped to 17 sealed indictments... which caused MANY people to start sharing her false 'discovery'.
By Nov. 8, she was up to 33. This post is significant, because it was the start of where things really started to go viral. Many people shared the information she provided here... 3 of those tweets alone received a total of 20,000 likes and 11,000 RTs.
Later that day, she tweeted that she was thinking about looking into the VA-East courts, due to their close proximity to D.C.
She followed this up a half hour later, with a report from VA-E that she said revealed 50 sealed 'indictments' (really PROCEEDINGS).
While all of this was going on, there was another twitter user who had been closely following Avery's updates... damartin32. He began running his own PACER searches on other districts, and eventually their paths crossed.
This is damartin32's timeline:
damartin32:
damartin32 was very interested in areas where he believed the FBI had been investigating Hillary Clinton. Initially, he did not have a PACER account, and on Nov. 9 2017 he asked Avery if she could look into those locations:
He also asked his followers if anyone had PACER access, to look into those districts:
A few hours later, he signed up for a PACER account, ran several searches, and started posting them to twitter. This is the start of where the large amount of sealed cases really begins:
damartin32 also repeated the mistake that Avery had made, by not properly filtering the case types.
Because of that, he was now reporting that he had found 141 sealed indictments, which was wildly inaccurate. Several hours later, he added more locations to his search, and the count had jumped to 294 sealed 'indictments'.
On Nov. 10, he stated that he was going to check districts where he wouldn't expect to find sealed indictments:
He still isn't aware of the case type classification system, so when he gets a lot of hits, he thinks he has stumbled onto a massive amount of sealed indictments where they shouldn't be. The truth is that the vast majority were routine legal documents, and totally normal.
Nevertheless, he makes an update the following day, and claims that he has now found 842 sealed 'indictments':
This marks the first time that a chart/list of the data is provided. This also includes the numbers that Avery has been reporting, for VA-E and DC.
This is the full chart:
By Nov. 14, the count was up to 1183:
On Nov 21, he brought up the idea of running all 94 districts:
He put the word out for volunteers later that day:
The rest of the current research team joined up at this point, and on Nov. 26 2017, they released their first 94 district search, stating that they had found 4289 'indictments'.
By this point, damartin had discovered the issue with the case type classification system, which I will discuss a little later. Notice that even though he incorrectly said 'indictments' in the tweet, the chart was labeled sealed 'proceedings', which is an accurate description, and not misleading.
There is a huge difference between sealed indictments and sealed proceedings, so it's very important to use those terms properly. For example, according to the FJC report, in 2006 there were 284 sealed indictments, but 24,375 sealed proceedings.
Had they kept this title on the chart, a lot of the current confusion/inaccuracy would not exist... but they didn't.
Side note: The statement in the tweet that says 1077 is a normal yearly amount, is a false comparison. I explained this issue in-depth in my other article, that you can read here.
The next update came on Dec. 31, 2017, and the total count was now up to 9294:
Here is the chart for that update. This one is essentially an outright lie... they changed the title from 'Sealed Proceedings' to 'Sealed Indictments', and gave NO clarification anywhere that the numbers really represented 'proceedings'.
On January 26, 2018, they updated the count to 13,605:
The chart included with this tweet was the final iteration, which is still used today. They kept the title of 'Sealed Indictments', but they added a disclaimer at the bottom in small print, that says the numbers really represent sealed proceedings.
This is still EXTREMELY misleading. It's the equivalent of charting the population of every state in the US, labeling the chart "Male population by state", and then putting a disclaimer at the bottom that says the numbers represent both men and women.
Nevertheless, the research team has continued to keep this chart, and make monthly updates... the current total is now at 61,000 sealed 'indictments' (proceedings).
That is the story of how 2 legitimate sealed indictments, and a mistaken search setting in PACER, turned into the wildly inaccurate claim that there are currently 60,000+ sealed indictments in the legal system.
Below I am going to discuss some other aspects of this issue, including public responses that were made, Avery/damartin's misunderstanding of the case types/terminology, and give some final thoughts on everything.
Public responses, explanations:
The 33 sealed document tweet from Avery, that I referenced earlier, went viral on Nov. 8 2017. Tens of thousands of people were liking and retweeting it, which caused some very prestigious names in the legal industry to take notice of it.
The most popular response was from Benjamin Wittes, who has a very impressive legal background.
He immediately recognized that Avery was including magistrate judge proceedings in her results (i.e. NON-indictments), as well as pointed out that sealed indictments are not exactly unusual to begin with:
(Interesting sidenote... earlier I linked to a picture of tweets that went viral on this same day. Bill Mitchell's second tweet was made immediately after Benjamin's first tweet. Seems like an attempt to suppress the explanation.)
Other notable names, with legitimate explanations:
Brad Heath (Investigative legal reporter for USA Today):
Elizabeth de la Vega (21 year Assistant US Attorney, taught at Stanford Law):
Her tweets didn't get much attention, but they were very informative, and she was on it early (Nov. 6). The '17' she references are the documents discussed in Avery's previous tweet, which claimed to show 17 sealed indictments in DC.
As you can see, legal experts weighed in on this claim in it's very earliest days. They ALL immediately realized that Avery was incorrectly including non-indictments in her searches, which explained the seemingly high numbers.
Avery never directly responded to any of these threads, however, the morning after Benjamin's thread, Avery made this tweet in which Benjamin was tagged:
'Lawfare' is the name of the blog that Benjamin oversees, and that word also has a bit of a negative connotation... so basically Avery is making a disparaging comment about Benjamin.
That was her only response. Note that she received 1 like on her post... from damartin32.
This shows that Avery and damartin were aware of the issues with their numbers at the very beginning, but they didn't change their methods. It's possible that they didn't understand the details. As the tweets below will show, not understanding the terminology played a big role in all of this.
Misunderstanding terminology:
Avery made several tweets early on that implied she didn't fully understand the terminology of the cases she was discussing... specifically how sealed indictments differed from things like magistrate judge proceedings or criminal complaints.
For example, consider this tweet from Nov. 7, where she examined 6 so called 'unsealed indictments':
Semantically, there is no reason to believe that these cases were ever 'sealed'... they are simply public cases on the docket. However, look at the case numbers for all but the first case... the 'mj' in the case number stands for 'magistrate judge'... which means they are clearly NOT indictments.
It's one thing to not understand the case type classification system on sealed documents, as those docket numbers contain no identifying information (Ex: 17-206)... but to have the full case number displayed, literally showing 'mj', and still refer to it as an 'indictment', indicates a significant lack of understanding about the data being examined.
Misunderstanding case types:
The discovery of the case type cr/all classification issue -- which is the big issue here -- came from damartin.
On Nov. 18, he stated that he had 'broken down' the DC/VA numbers.
He followed this up a few hours later, and as you can see, his 'deep dive' entailed searching for each case type individually.
This made a massive difference in the numbers, in regards to sealed indictments.
Where he had previously been reporting 136 sealed indictments in DC, the number was really 6. Where he had previously been reporting 210 sealed indictments for VA, the number was really 5.
Keep in mind that those 11 sealed cases aren't even confirmed indictments... they are merely sealed documents in the category where indictments are filed. It indicates a maximum of 11 indictments, with a possibility of 0.
The difference compared to what he had been reporting was so large, that a person who had been following his updates even made a comment expressing disappointment at how low the 'real' number was:
When damartin posted this breakdown, Avery asked him where he was getting the data from... he explained it was from filtering the case types.
This shows that Avery did not understand the significance of case types when she started making her claims. This could explain why they didn't react to the rebuttals I previously posted... they possibly didn't understand the details of what the legal experts were saying.
Despite becoming aware that the amount of sealed indictments was dramatically lower than what they were reporting, they continued to search for ALL case types, continued to use the term 'sealed indictments' in their discourse, and continued to falsely compare the total to 1077.
Labeling the chart sealed indictments:
The biggest reason that people mistakenly refer to the 60k cases as 'indictments', is because of the way the chart is titled, which I discussed above. This is a conversation between two of the research team members, after someone questioned them on their use of the word 'indictments':
This is acknowledgment that the research team was aware of the issues with using the word 'indictment' when they made the chart, and it implies that they think it was not the right move. They could correct it at any time, but they choose not to.
Conclusions:
Avery and damartin did not start the 60k Sealed Indictment chart with bad intentions... the evidence clearly shows that it was just a matter of not understanding the data they were looking at.
That being said... the evidence 'also' clearly shows that at some point they became aware of their mistakes, and they have made no real effort to correct them. They know that labeling the chart 'sealed indictments' is misleading, and they certainly know that 1077 is a false comparison. At this point they are knowingly pushing bad info.
I reached out to both of them for clarification, but I was blocked after bringing up the case type/1077 issue.
Final note, regarding 4chan's role:
There is still one factor at play here that is a bit harder to determine... and that is how big of a role did 4chan play in this.
The first 4chan 2 sealed document screenshot was totally legitimate, and Avery's 60k sealed indictment chart was started because of a mistake. The mystery is the 2nd screenshot from 4chan, that links those two documents.
Was the mistake to leave the case type on 'all' intentional? Was it meant to bait someone into running searches incorrectly, in the hopes that it would set off a chain reaction of incorrectly identified 'sealed indictments'? Or was it merely an unintentional mistake, made by a person who also wasn't aware of the case type classification system?
In my opinion, due to all of the talk about PACER, and all of the armchair investigative reporters researching the topic at the time, I think it's plausible that leaving the case type setting on 'all' was simply an unintentional mistake.
That being said...
In the course of researching this article, I discovered a PACER screenshot that was definitely edited in a deliberate and misleading way. I wrote about it here. That is concrete proof that there were some hi-jinks taking place on 4chan, in regards to the sealed indictments.
It's hard to say if the person that edited that screenshot specifically did it to mislead people, or if they edited it for some other reason... but if they did edit it in hopes of misleading people... well... it might have worked.
Hi m.l.jackson. I answered your question on another forum where I saw you ask it. Here is the link: https://www.metabunk.org/posts/228816/
Posting this here because the other sealed indictment thread is closed(?)... Sorry, TL;DR. Can you give me the bottom line? Despite the poor comparisons used by Storm Watcher et al, a) what is the ratio of the average number of sealed *proceedings* preceding 2018 compared to the final count for 2018, and b) based on that ratio and historical data, what is the average number of sealed indictments we can predict? Isn't that the bottom line based on available evidence? Thanks!