Deep Dive Of The Deaths Dataset

Checking the Czech data in-depth

Corrections: The incorrect geographical term ‘Czechoslovakia’ was used. It has been changed to ‘Czech Republic’, given they are two different demographs. Updates: An addendum has been added to the end of the article thanks to insights by Maxipes.

Warning: I hope you love numbers because there will be a lot of them in this article.

Precursor

Jessica Rose has recently published an article linking to the Czech deaths and ‘vaccination’ dataset that has been FOIA’d by Steve Kirsch.

I had wondered why the data hadn’t been offered previously and why the same graphs were being reused, but then I found out from Jessica’s article, the data contains a whopping 11,028,372 records (that’s over 11 million records).

I lit up with glee! Processing extremely large datasets is one of my many grab-bags (folks might remember the time I almost melted my CPU doing OCR to convert the massive Fauci FOIA into searchable text).

11 million records will cause most spreadsheet programs to perform horribly or just tip-over entirely. LibreOffice Calc (above) only partially loads the dataset and throws an error (LibreOffice Calc’s maximum row count is just over a million at 1,048,576).

This isn’t the first time I had encountered this. The NHS are notorious for using spreadsheets for large quantities of patient record data.

Fixing The Problem

Normally, what you’re supposed to do with such large datasets is use software and file formats designed for such things, such as SQL or SQLite.

As it so happens, behind the scenes, The Daily Beagle has been developing ‘MIMS’ (Minimal Information Management System), which makes use of SQLite. I still also had the CSV processing code from the time I stayed up until 3am to refute the CDC in order to produce a video showing a vast number of studies showing the harms with the shots.

Punching out some code using the software library I’ve made, I begun converting the CSV into an SQLite file (roughy 1.1 GB, approximately the same size as the .CSV file). A row level count confirming there are 11,028,372 records inserted.

Field titles translated using Google translate and spaces replaced with underscores to prevent typical SQLite issues with naming conventions. Other than that, no further changes were made.

Anomalies In The Dataset

First anomaly: the number of records (11,028,372) exceeds the total population count of Czech Republic.

Citing different sources:

1. Worldometers: 10,504,183

2. Countrymeters.info: 10,732,783

3. World Population Review: 10,503,912

4. According to the Czech government in 2022: 10,533,399

   

5. According to the Czech government’s real time tracker: 10,859,532 (accurate as of 19th June 2024 when above screenshot was taken)

Somehow we have between 168,840 to 494,973 more records for people than there are people in the Czech Republic. The FOIA descriptor states that it is…

Anonymizovaná data všech občanů ČR vedených v datových sadách ÚZIS v letech 2020, 2021 a 2022, 2023 a dostupná data roku 2024 […]

[Translated: Anonymized data of all citizens of the Czech Republic maintained in the ÚZIS datasets in 2020, 2021 and 2022, 2023 and available data in 2024 […] ]

I refuse to believe there is a 100% data tracking rate amongst the entire populace. No medical record system is 100% complete. It is an impossibility; the absence of gender codes for 9,942 records (discussed later) proves that. And yet somehow there are more records than citizens.

Any inaccurate data reporting is the fault of the Czech government. So don’t go blaming the folks who’ve done analysis upon this data.

It is worth noting The Daily Beagle previously called out Czechia’s fraudulent record keeping (which, surprise, involves declaring they had administered more doses than they distributed, somehow).

Even noting these anomalies, I’ve opted to proceed with the analysis.

Metadata Analysis

Using the SQLite query system and some manual analysis, I found:

• Records with Gender_Code (‘Pohlavikod’) of F: 5,580,356

• Records with Gender_Code of M: 5,438,074

• F+M total: 11,018,430 (meaning there are 9,942 records not marked F or M)

• Records with a non-null Date_of_death (‘DatumUmrti’): 389,398

• Records with any vaccine variables [read: total vaccinated]: 6,982,201

• Records with a non-null Date_of_death and at least one variable in any of the vaccinated fields [read: ever-vaccinated who died]: 143,958

• Records without any vaccine variables [note: may include people who received at least one shot E.G. overseas but did not disclose it for record-keeping purposes]: 4,046,171

• Records with a non-null Date_of_death and with no vaccine variables [read: unvaccinated who died]: 245,440 (a suspiciously round number)

  • However the above dataset is misleading because it includes deaths from 2020 from way before the shot was rolled out in Czechia, meaning unvaccinated have an extra year of deaths versus vaccinated

  • Normalising the deaths of unvaccinated so it only falls between 2021-2022 (no data for 2023/2024 was provided): 116,151

    

    • Of which it appears to include roughly 495 still born or died-shortly-after-birth deaths (died the same year born), which in my humble opinion isn’t valid (especially given the mother’s vaccination status is unknown; I’m sure someone can chip in with the SC2 shots causing stillbirths datasets here).

    • 116,151 is of course less than 143,958; I’m sure some will engage in some divide-by-population jiggery pokery or the classic ‘person years’ fallacy but we’ll see why that isn’t valid shortly

To make independent analysis easier, I’ve isolated the death groups into two datasets:

• ever-vaccinated deaths

• unvaccinated deaths

They’re both below the 1 million row cap so should be viewable.

Analysis Of The Vaccine Deaths

Using the original dataset, I generated new datasets, which included:

A vaccine death summary dataset, including:

• Number of shots taken per record

• Number of days between the last shot and date of death

• When the last shot was taken + date of death

• What the last shot taken was

And some simple tally datasets:

• How many people died, by number of days after last shot

• How many people died, by total number of taken shots

From a temporal (time) analysis, I found:

• 61 died within the same day of their last shot (major red flag the shots kill)

  • The ever-vaccinated have a 1 in ~114,462 chance of dropping dead same day (not ‘1 in a million’ like they frequently invent)

• 2,065 died within 7 days (1 week) of their last shot

• 4,907 died within 14 days (2 weeks) of their last shot

• 8,314 died within 21 days (3 weeks) of their last shot

  • Interestingly, it refutes the ONS and the CDC’s malicious exclusion of “<21 days” from the ‘effectiveness’ datasets, given so many deaths happen before 21 days.

• 12,762 died within 30 days (~1 month) of their last shot

  • Or ~8.8% of all vaccine deaths

  • Ever-vaccinated have a 1 in ~547 chance of dying within 30 days

• 38,640 died within 90 days (~3 months) of their last shot

  • Or ~26.8% or 1/4th of all vaccine deaths

  • Ever-vaccinated have a 1 in ~180 chance of dying within 90 days

• 552 people — the largest single death tally — died at 19 days (within 3 weeks)

• Within 365 days (~1 year) of their last shot, 129,939 had died

  • Or 90.2% of all vaccine deaths

  • Ever-vaccinated have a 1 in ~54 chance of dying within 365 days

From a shot count perspective, I found:

• The bulk of the deaths occurred in people who had taken either 2 or 3 shots (exclusive groups, no double-counting).

  • 2 shots: 62,485 deaths

  • 3 shots: 61,976 deaths

    • Total (2 + 3 shots): 124,461, or ~86% of shot deaths

    • Notably the number of deaths between the two groups is painfully similar

• Fewer deaths occurred in single shot takers: 15,055

• When it comes to deaths, hardly anyone makes it beyond their 5th shot

  • 4 shots: 4,439 deaths

  • 5 shots: 3 deaths (but you’d have to run the gaunlet of deaths in the 1, 2, 3 and 4 shot categories to get there)

Comparisons With The Living

There were no records of deaths for 6th or 7th shot takers. So to better qualify these datasets, I decided to also produce a tally breakdown of the number of shots versus those still living.

• Hardly anyone had taken the 6th or 7th shots (hence the total lack of deaths in those brackets):

  • 6 shots: 1,748 living

  • 7 shots: 49 living (can’t be that successful!)

• The largest shot count groups are the 2 and 3 shot groups respectively.

  • 2 shots: 2,275,352 (~2.2 million) living

    • At 62,485 deaths, you have a 1 in ~37 chance of dying at 2 shots.

    • To give you a rough idea of how likely this is, imagine rolling 2 six-sided die (6*6 = 36 probable outcomes) where getting one specific combination of numbers from the roll means you die.

  • 3 shots: 3,466,688 (~3.4 million) living

    • At 61,976 deaths, you have a 1 in ~57 chance of dying at 3 shots

    • Note: the chances, being calculated separately, stack, so it is a 1 in 37 + 1 in 57 chance.

• It is clear people don’t trust the shots as far fewer people get their fourth or fifth shot:

  • 4 shots: 620,787 living

    • At 4,439 deaths for 4 shots, you have a 1 in 140 chance of dying with 4 shots

    • Remember, probability still stacks

  • 5 shots: 288,487 living

    • 3 deaths; 1 in 96,163 chance of dying with 5 shots

    • Remember, probability still stacks

• Single shot had 329,090 living

  • Of which 15,055 died, or a 1 in 22 chance of dying

• The estimated probabilistic odds of dying at each stage (exclusive) are:

  • 1 in ~22 (1 shot)

  • 1 in ~37 (2 shots)

  • 1 in ~57 (3 shots)

  • 1 in ~140 (4 shots)

  • 1 in ~96,163 (5 shots)

  • In total, the odds would be 1 in ~48 chance of dying at some point after taking any number of shots

  • Hardly anyone is crazy enough to take 6 or 7 shots (I wonder why?)

Deaths By Vaccine Brand And Type

The deaths breakdown by type and manufacturer are eye-opening. For this I am going to simply show a screenshot of the tally.

The largest number of deaths are from Pfizer, at a whopping 106,235 deaths:

Moderna had 19,724 deaths, and mRNA in total had 125,959 deaths, or ~87% of the total number of shot deaths. There were way more makes and models of shot in the tally of the living. The biggest risks, however, were with AstraZeneca.

• Of mRNA:

  • 1 in ~33 chance of dying if you take a Moderna shot

  • 1 in ~57 chance of dying if you take a Pfizer shot

  • 1 in ~53 chance of dying if you take any mRNA shot

• Of Genetically Modified Adenovirus:

  • 1 in ~5 chance of dying if you take AstraZeneca

  • 1 in ~66 chance of dying if you take Johnson&Johnson/Janssen

  • 1 in ~18 chance of dying if you take either genetically modified adenovirus shot

• Of Saponins (note the transfection risks here):

  • 1 in ~416 chance of dying if you take Novavax

    • Note this may be skewered by having an extremely small sample, that said, even if it wasn’t skewered, it shows just how badly the mRNA and genetically modified adenovirus shots are performing

Manually compiled breakdown as a screenshot:

1 in 416 chance of dying still are not good odds. It only looks good because the ratios for Pfizer, Moderna, AstraZeneca and J&J look so bad.

To put these figures in sharp contrast, your risk of dying during surgery from general anaesthesia is 1 in 100,000 (source).

To visualise the risk, here’s a chart from the Royal College of Anaesthestists:

These deaths are not “very rare” or “rare”. AstraZeneca’s death rate of 1 in 5 would make it extremely common, and Pfizer’s death rate would be quite common. That’s extremely alarming!

This is all the analysis time I’ve got for now. You can find the uploaded repository of data I’ve extracted from the dataset here.

Disclaimer: There may be oversights in my compilation and/or reporting, as I cannot see what the SQL query is doing when extracting the information (it’s impossible to manually vet the scanning of 11 million records), however hopefully you can see this was a good faith effort in accuracy.

Would you trust any of these shots, dear reader?

---

Addendum

Commenter Maxipes, who is naturally fluent in Czech, has taken the liberty of greatly expanding context. Please take the time to read and like their comment.

They’ve highlighted there’s a discrepency between what was requested (data on citizens) and what was actually supplied (data on persons; which would include non-citizens) — these are the sort of technicalities machine translations will miss.

They go on to note that this is a merged (compiled) dataset, merging vaccinations and deaths — it isn’t a single centralised database. They note during this time (Feb 2022 onwards) Czechia received a lot of Ukrainian refugees (typically wouldn’t have been subject to the usual rigourous controls).

This raises an issue of transitory receipt of shots. The problem with this is, non-citizens could have received shots in Czechia, then travelled on (and died/gotten ill) in another country. The supplied dataset by the Czech government does not actually comply with the narrow request for vaccination data on citizens. It shouldn’t be hard to comply with that criteria; can’t help but wonder if the non-citizens are being used to fuzz the data.

---

Found this informative?

Help inform?

Thoughts, dear reader?