Steve Kirsch’s “mother of all revelations” about the “deadliness” of COVID-19 vaccines goes poof

Last Thursday, tech bro turned rabidest of rabid antivaxxers Steve Kirsch finally gave a talk that he had been hyping for a couple of weeks as the “definitive” evidence that COVID-19 vaccines are not just deadly, but so deadly that, by his estimation, they’ve killed well over 13 million people. He also made a big deal of giving the talk at MIT in the auditorium that had been named after him before he had turned into a raving antivax conspiracy theorist peddling pseudoscience who has become so proud of his status as “misinformation superspreader” that he had a T-shirt made featuring that saying.

Steve Kirsch: Misinformation superspreader. He means it ironically, but in fact it is an accurate description of what he’s become.

Before I delve into Mr. Kirsch’s massively flawed analysis and findings, let’s briefly take a look at the dataset used and how Mr. Kirsch apparently got his grubby little hands on it.

A New Zealand “whistleblower”

As is the case with many good antivax conspiracy theories that involve torturing an existing dataset until it “confesses” that vaccines kill, Mr. Kirsch’s “mother of all revelations” (MOAR) about COVID-19 vaccines involves a “whistleblower,” who apparently passed on record-level data from Te Whatu Ora (the agency that manages the provision of healthcare services in New Zealand under the Ministry of Health) tallying deaths in vaccinated people, along with COVID-19 vaccination dates. (He also uses the pseudonym of Winston Smith, the protagonist in George Orwell’s dystopian novel Nineteen Eight-Four, because of course he does.) Antivaxxers trumpeting the “whistleblower” also claim that he is a statistician, which, as you will see, he appears not to be, at least not one with epidemiological or clinical knowledge.

Steve Kirsch being Steve Kirsch, a true prolific misinformation spreader who can’t resist anything that brings attention to him, he was hyping this talk and these “revelations” in the lead-up to his talk, revealing quite a bit along the way, because apparently he just can’t help himself. For instance, he revealed that he had illegally obtained the data on November 9 from the whistleblower, whose identity he knew. Of course, he didn’t actually admit that the data had been illegally obtained, but after he had started hyping his talk a number of people told him and his cronies that obtaining record-level data with personally identifiable health information was illegal in New Zealand, the US, and many other countries. At one point on X, the platform formerly known as Twitter, the “scientific advisor” of his antivax org Vaccine Safety Research Foundation (VSRF) even conceded with a metaphorical shrug of the shoulders that the data had been illegally obtained:

Indeed, apparently Te Whatu Ora was sure enough that the privacy breach had been illegal that it rapidly launched an investigation, which resulted in the arrest of a 56-year-old man, who had appeared in a video on Rumble with NZ presenter Liz Gunn to reveal the data, on suspicion of a “mass privacy breach“:

Police have arrested a man in connection with a mass privacy breach of Covid-19 vaccination data.

It comes after Te Whatu Ora- Health New Zealand launched an employment investigation, accusing a health sector worker of spreading misinformation using government data about Covid-19 vaccines.

A 56-year-old man was arrested this afternoon in relation to the “unauthorised disclosure and misuse of data”, police said.

He is charged with accessing a computer system for dishonest purposes and will appear in Wellington District Court tomorrow.

NZ authorities note that the man “had no clinical background or expert vaccine knowledge,” which means that he might be a statistician, but according to his interview as reported by the Brownstone Institute:

Smith was in an unusual position as the database administrator for the payment system. ‘Because New Zealand is a small country, you can get away with one database administrator. I am in a unique position, and because New Zealand is a Tier 1 country with really good IT, I was able to manage and build this system.’

So he was a database administrator, not a statistician, and appears to have no relevant scientific or clinical knowledge of public health or epidemiology, but rather acted on all sorts of assumptions that spikes in daily death rates were “not natural” and “had to be man-made.” Indeed, according to Mr. Kirsch, “Winston Smith” is an Oracle database administrator for the payments database:

The NZMH whisteblower, Oracle database admin Barry Young, is a hero. He knew he would risk his life and could spend the rest of his life in jail, but he made the courageous move to expose the data for all to see. This is a highly commendable act of public service. He basically threw the rest of his life away in order to save the lives of others. Why else do you think he would do that? Nobody can explain it.

I love the naive assumption that, just because Barry Young risked his freedom to illegally download four million health records, that means there must be something to his suspicions and findings (and therefore to Mr. Kirsch’s “analysis”). No, just because Mr. Young apparently believed strongly enough that vaccines were killing large numbers of people in NZ that he risked prosecution and prison does not mean that he is correct in his belief. Strength of belief does not correlate with correctness of belief, even though a lot of people like Mr. Kirsch think that it does. (Actually, far too many people think that strength of belief correlates with correctness of belief.) Moreover, the revelation that Mr. Young is only a database administrator, and not just a database administrator. but apparently the payments database administrator, tells me all I need to know about his qualifications to judge what these data show: Zero.

According to the NZ authorities:

Earlier tonight, the ministry said the staff member had no clinical background or expert vaccine knowledge, and there was no evidence vaccination was responsible for “excess mortality in New Zealand”.

It’s alleged an individual downloaded a large amount of vaccine-related information, Te Whatu Ora chief executive Margie Apa said.

“The data, as published on an overseas site, appears to have been anonymised. Analysis of the released data is ongoing, but work so far has not found any National Health Index Numbers or personally identifiable information.”

Apa said an injunction had been used to have information taken down from an overseas website and cyber security specialists are continuing to scan extensively for any other places where the information may appear.

This was, of course, a completely expected reaction from a government agency after a computer breach and illegal download of a massive health department dataset, but, of course, Mr. Kirsch predictably saw it as more evidence of a “conspiracy” to “silence” him, particularly when Wasabi, the platform to which he had uploaded the illegal data, revoked his access to his account due to a TOS violation because of too much traffic (although it’s very likely that the New Zealand injunction also played a role in the decision):

Email to the CEO of Wasabi about my urgent situation was not returned.

Wasabi tech support isn’t answering any of my emails either.

The NZMH will do everything they can to stop this data from getting out because it reveals they are killing people with the vaccine.

What did Mr. Kirsch expect? He uploaded a dataset that, he knew, would likely result in a lot of traffic once he revealed on his Substack how to download it and that, he knew or should have known, had been illegally obtained. And, make no mistake, this was illegal. In the US, we refer to health information that has personally identifiable information in it as PHI (protected health information), and we have a law (HIPAA) that governs when such information may be shared and among whom. Suffice to say that anyone with access to PHI can’t just download and share it. Also note that “protected” doesn’t mean that the names have to be associated with the information; any form of personally identifiable information that might allow guessing whose information it is counts. Ever since HIPAA was passed, anyone in healthcare has had to go periodic mandatory training regarding patient data and not even inadvertently sharing it with entities who do not need it to provide healthcare services. Unsurprisingly, NZ has a similar health privacy law.

Of course NZ health authorities were going to act and seek to get an injunction against any private company hosting the data. If I thought Mr. Kirsch were sufficiently clever, I’d suspect that he had intentionally chosen a platform like Wasabi to host the dataset, knowing that once he publicized it health authorities in NZ would act to shut it down. Maybe he even knew that his talk would result in so many people downloading the dataset that bandwidth limits would likely be exceeded. (Hint: I doubt that he’s that clever.) I also note that, while perhaps Mr. Kirsch thought that anonymizing the dataset before releasing it into the wild would protect him and his “whistleblower,” that doesn’t matter. The original dataset, which he presumably possesses, still contains PHI, making its theft—yes, theft—a massive privacy breach, particularly when carried out by someone employed by Te Whatu Ora, which likely included in its employment contract conditions under which the employee was permitted to access the database.

Unsurprisingly, Mr. Kirsch is now bragging that he’s hosted the data on a “bulletproof” platform and that if “they try to take me down next time, they will regret it,” going on to say, “I actually hope they do; we’ll turn the tables on them.” Still, for a tech bro, Mr. Kirsch was either really stupid or more clever than we give him credit for. Given his history, I suspect the former.

In any event, before I move on to the dataset itself, let me just briefly point out that this whole shtick is very familiar. I very much get “CDC whistleblower” vibes from it. You remember the “CDC whistleblower,” don’t you? It was a scientist named William Thompson, who, while working for the CDC, decided that he disagreed with the analysis of a dataset examining whether there was any correlation between autism risk and vaccination with MMR (measles-mumps-rubella vaccine) and then started having phone conversations with an antivaxxer named Brian Hooker, apparently to vent. The result was a “reanalysis” of the dataset by Brian Hooker, a biochemical engineer turned incompetent epidemiologist who had at one point bragged about how he thought “simplicity” in statistical analyses was best and predictably found an increased risk of autism in African American boys due to MMR vaccination. (Of course, “simplicity” to antivaxxers means raw data analyses, without appropriate statistical adjustment for confounding factors.) The whole conspiracy theory was ultimately turned into an antivax epic in 2016, a pseudodocumentary by Andrew Wakefield and Del Bigtree entitled VAXXED. It’s an antivax conspiracy theory that persists to this day, promoted by—who else?—Steve Kirsch. One can’t help but wonder if his MOAR is nothing more than a pathetic attempt on his part to create his own CDC whistleblower.

But what about the dataset itself? Let’s, for the moment, ignore the illegal, unethical and immoral methods by which Mr. Kirsch got his sticky fingers on the dataset and just look at the dataset. Then we’ll move on to his talk.

The NZ dataset

The dataset stolen by the NZ Te Whatu Ora “whistleblower” and passed on to Mr. Kirsch consists health records from 12 million doses of COVID-19 vaccines. Specifically, according to Mr. Kirsch, it contains four million of these records from the the “Pay per dose” (PPD) program, the other eight million apparently not being from that program. Mr. Kirsch claims that whether “you got PPD or not is pretty random,” which sent up an immediate red flag to me regarding whether it is actually true or not that PPD was random or not. I could not ascertain that from my reading. However, I did immediately wonder. (Maybe one of our New Zealand readers could help me out.)

In addition, there are numerous other deficiencies that Mr. Kirsch was forced to admit but that, he claims, do not invalidate his analysis, for example:

• There is a disproportionate amount of records for the doses, i.e., they are not in direct proportion to the total number of each dose, e.g., they are not 33% of each dose. Some doses are over-represented, some are under represented.
• Many people will not have all their doses in this database, e.g., there may just be dose 3 data for someone.

In other words, it would appear that some people got some doses through PPD and others not through PPD, which would explain why only some doses were covered. Mr. Kirsch just glosses over this, claiming ridiculously:

The fact that the sampling was uneven doesn’t matter if you analyze it the way I did. The fact that doses are missing is also irrelevant. These are gaslighting arguments made by people who are incompetent to analyze this data.

To which I reacted:

Seriously, if you are going to assert that, for purposes of your analysis, incomplete records and uneven sampling in the dataset don’t matter, you really do need to show the receipts and mathematically prove that these deficiencies in the dataset don’t affect the results of your analysis. Epidemiologists do this all the time by doing multiple different analyses on datasets with missing records and/or uneven sampling, because there are statistical methods for dealing with uneven sampling and missing records that peer reviewers will expect to see if such analyses are to be published in the peer-reviewed biomedical literature. (It’s called sensitivity analysis, as in testing the sensitivity of the outcomes of the analysis to important assumptions made in doing the analysis.) Tellingly, Mr. Kirsch doesn’t even attempt to use them or mathematically justify his claim that these things don’t affect his analysis. He just asks you, with an overconfidence that bespeaks one of the most extreme examples of the Dunning-Kruger effect that I’ve ever seen, to “trust me on this,” so to speak.

Of course, there are other major deficiencies in the dataset, the key one being that the dataset includes only vaccinated people, when they were vaccinated and with what vaccine, as well as dates of death for those who died. We don’t know the baseline death rate in an unvaccinated cohort, which would almost certainly be much higher.

None of these deficiencies stops Mr. Kirsch, though. So let’s look at his analysis. His talk was posted to Rumble. (Where else?) His slides were posted to Google Documents and to his own website as a PDF. Feel free to watch it if you can handle the tsunami of Gish galloping misinformation.

Kirsch’s MOAR goes poof

Watching Mr. Kirsch talk for so long is rather painful. He uses entirely too many slides, and the aesthetics of his PowerPoint design reminds me, more than anything else, of late 1990s Time Cube. Let’s just say that I have a number of old antivax PowerPoint presentations that I saved from 15-20 years ago, and his reminded me a lot of them. But what about the substance? It’s also clear that Mr. Kirsch revels in his reputation as a “misinformation superspreader,” as his early slides claim that he’s written over 1,500 articles on COVID-19 (whoop-de-doo, I’ve probably written at least five times that number over the years on vaccines in general), before he whines:

Funny, but instead of wondering if the problem is with him, instead Mr. Kirsch automatically assumes that the problem is with everyone else and not him. Malignant narcissism, anyone? (Also, “whack job” is probably being kind.)

Now let’s look at his key assumption, for which he provides no references, statistics, or math to justify:

When you assume, you make an ass out of…well, you remember the old saying, right?

Before I go on, let me just point out a very important principle. When you see a nonexpert analyze a dataset using methods that no expert uses to analyze the same type of dataset, be suspicious. It is, of course, possible that the nonexpert has stumbled onto an innovative way of analyzing data of this type. But is it likely? No, not really. It’s far more likely that he’s made a rookie mistake. Moreover, when the nonexpert doing the analysis is someone like Mr. Kirsch, who has a history of promoting the most outlandish antivax conspiracy theories, it’s far more likely that the nonexpert is doing nothing more than torturing the data until it confesses what he wants it to confess. Be even more suspicious that this is what is going on when, for example, the nonexpert, after stating his assumptions, doesn’t go straight into the data analysis but instead starts ranting at Moderna co-founder Robert Langer and claiming that as a board member of Moderna he has a responsibility to “stop the shots” and might be liable to prosecution under the PREP act. (That’s exactly what Mr. Kirsch did, referring to the vaccines as “kill shots.”)

Then he looks at another dataset for different vaccines, noting that his assumption is validated for the pneumococcal vaccine:

One of two things could be happening here: Either the vaccines are responsible for excess mortality or the analysis is systematically flawed such that it always comes out the same way. Guess which is more likely.

If you look at the original Substack upon which Mr. Kirsch bases this chart, you’ll note that there are data for unvaccinated people. You’ll also note that he hand waves away why the data are “confusing.” In any event, after Mr. Kirsch had published this analysis back in late February, a number of commenters pointed out that his results could be explained by the healthy vaccinee effect (people who are vaccinated tend to be the most at risk, as in older and less healthy, and those who are less healthy tend to self-select to be vaccinated regardless), seasonality (deaths tend to peak in winter, particularly among the elderly), and COVID-19 waves. It’s no surprise that Mr. Kirsch simply does the same thing with the New Zealand dataset.

But first, he has to cite anecdotes because:

Antivaxxers always value anecdotes over actual rigorous data.

My reaction was this:

When Godzilla gives you the facepalm, you know the failure is monstrous. I also chose this particular facepalm because I can’t wait to see Godzilla Minus One, which just came out on Friday. Unfortunately, I probably won’t get to see it before this weekend.

Think I’m exaggerating? OK, then, let’s look at the next slide:

Jay Bonner? Seriously?

I discussed Mr. Bonnar’s “anecdote” and how implausibly unbelievable in the context of deconstructing Mr. Kirsch’s equally implausibly unbelievable claim that COVID-19 vaccines have killed 3.5 times more people than COVID-19. Mr. Kirsch sure does like to recycle his “greatest hits” of disinformation.

But, surely he must be about to get to his analysis of the NZ data, right? Sort of. First, he has to go on about getting the data on November 9 and then saying:

Steve Kirsch’s Rosetta Stone? Not really. Any dataset is a Rosetta Stone for conspiracy mongers determined enough to torture the dataset until it confesses what they want it to confess.

Actually, any dataset, “record-level” or otherwise, can be a “Rosetta Stone” for conspiracy mongers like Mr. Kirsch, who are determined enough to torture the dataset until it confesses what they want it to confess. That’s exactly what Mr. Kirsch does, but not before extrapolating wildly from his “findings” to claim that COVID-19 vaccines have killed over 13 million people worldwide, a false claim that he repeats later in his presentation with this slide:

Seriously, conspiracy theorists like Mr. Kirsch seem to think that public health researchers are either so clueless that they didn’t notice 13 million deaths or so evil that they covered them up, never mind the implausibility of the claim that anyone could have covered up so many deaths.

I suppose that Mr. Kirsch thought that he was being a good showman in dragging out the presentation. Maybe he thought he was building suspense. To me, what he really did was make me fast forward through his talk because I was getting bored and frustrated that he wasn’t coming to the point. I know, I know, I’m sure it’s not entirely unlike some of my readers, who skim through until I get to the point. The difference is that I’m not entirely oblivious to the flaws in my communication style, the way that Mr. Kirsch appears to be when he decides to throw in some more slides from his bogus Medicare analysis, slides of other antivax “analyses” from other countries, and even slides about his utterly risible “survey” that led him to claim that COVID-19 vaccines have killed 3.5 times more people than COVID-19 into his talk before getting to what everyone wants to see, the NZ data.

Oh, there’s this:

As I’ve noted before, Dr. McCullough has not actually had his board certification revoked yet. I searched the ABIM website. You can search it too if you don’t believe me, but here’s my screenshot:

Oops! As of December 3, Dr. McCullough is still Board-certified!

Finally, there was this:

There’s just one problem here…

I only include this because Mr. Kirsch mentioned me. There’s also just one problem here. He blocked me, and I blocked him back (because I always block back) months ago. This is purely performative. Also notice the number in the lower righthand corner. That’s the slide number. Now go back and count how many slides he included before getting to the point, his big “MOAR” of the NZ data. Yes, it’s a lot.

it’s 11 more slides before he gets to the actual data. Now, before I present the slides, let me note that there’s a bit of a trick here let’s see if you can spot it:

Then, Mr. Kirsch finally presents all ages:

Spot the trick yet?

First, let’s again revisit Mr. Kirsch’s assumption, which is that the death rate should be flat. I immediately wondered about that assumption, particularly in the middle of a pandemic in which a contagious respiratory virus is ripping through the population at various intervals and to which the elderly and those with chronic health conditions are most susceptible to severe disease and death. I had a conversation with Prof. Jeffrey Morris about this, but first lets see what this Professor of Public Health & Preventive Medicine; Biostats, Stats & Data Science at Penn had to say publicly on X/Twitter:

There was, of course, more:

I also like to refer to this graph from Our World In Data:

COVID-19 deaths by vaccination status over time. Notice how much consistently  lower the death rate is among the vaccinated.

Let’s quote a key element of Prof. Morris’ discussion, given that you might not want to have to go to the X/Twitter website to see this and you might not even have an account anymore:

As I would emphasize, any genuine attempt to assess potential causal effects of vaccines requires consideration of controls and adjust for confounding and other sources of bias inherent to these observational data in the pandemic (as may published studies do), but it might be useful for some to consider looking at these data as a basic plausibility filter for assessing whether they think the excess deaths are primarily driven by vaccination or by covid.

BTW the waves of covid infection waves coincide closely with the vaccine deaths in the plot, so even if you question covid death attribution, these times are precisely the times when each place experienced a massive wave of confirmed covid infections.

Note: It should read after the correction of an unfortunate typo:

BTW the waves of covid infection waves coincide closely with the COVID deaths in the plot, so even if you question covid death attribution, these times are precisely the times when each place experienced a massive wave of confirmed covid infections.

Everything old is new again, as I like to say. Basically, Robert F. Kennedy, Jr. “proved” that the CDC was “covering up” data supposedly showing that thimerosal-containing vaccines caused autism because appropriate adjustment of the analysis of the raw data for confounders made the effect go away. Ditto Brian Hooker “reanalyzing” the “CDC whistleblower” data. Basically, it’s a theme in antivax “reanalyses” of datasets like the Medicare and New Zealand datasets; they take raw results and fail to appropriately adjust for confounders. I’ve seen it again and again and again over the years. It’s a rookie mistake, and Mr. Kirsch falls for it like a day one rookie; either that, or he does it intentionally. (Take your pick.) In brief, Mr. Kirsch’s assumption that these curves should be absolutely flat in the middle of a pandemic is just that, an assumption. He does not adequately show that it is a justified assumption either, to put it mildly.

Even worse, Mr. Kirsch’s extrapolation is wildly inappropriate. Where do you think he got his claim that COVID-19 vaccines have killed over 13 million people? Ridiculously simple, although it’s really simply ridiculous. He extrapolated from his “estimate” from this dataset that COVID-19 vaccines kill one in a thousand people who receive them, to estimate that COVID-19 vaccines are responsible for 13 million dead worldwide and 675,000 in the US alone. Again, these are numbers so wildly implausible that they do not pass the smell test.

As Prof. Morris noted to me, the increase in death rate that Mr. Kirsch claims is everywhere in the data is, in actuality, only evident in the older age groups. Of course, the “all ages” death rate is driven primarily by the older age groups. There’s the trick! It’s the main one, but not the only one. In addition, Mr. Kirsch assumes that healthy vaccine effects “last exactly three weeks and not a moment longer,” which leads him to his erroneous assumption that any subsequent increase in deaths can only be explained by vaccine-caused deaths and no other factor. Prof. Morris also agreed with me that you can’t know the baseline death rate without having an unvaccinated control group, but he also educated me by pointing out this:

When you consider the actuarial baseline and the fact that 2020-2021 New Zealand had dramatically lower death rates than historical baseline, it seems like the long increase he sees (in the older groups), is not excess deaths but basically a return to baseline death rate.

In other words, it’s yet another confounder that Mr. Kirsch failed to consider, incompetent rookie epidemiologist and statistician that he is. He also misuses “proof by contradiction,” in which one exhaustively lists all the other possible explanations for an observation, until the “only one standing”—as Prof. Morris put it—must be true.

He also states erroneously:

Gold standard? You keep using that term. I do not think it means what you think it means. In other words, just because Mr. Kirsch claims that this is the “gold standard” analysis for vaccine safety does not mean that it is, in fact, the gold standard analysis for vaccine safety.

Notably, he shows no documentation that this is, in fact, the “gold standard” for determining a safe and effective vaccine. (Also note the slide number. And people wonder why I was getting tired as I neared the end. Hint: He had more than 100 more slides to go, most of which included a lot of rants that are very old antivax tropes about how supposedly there is less autism in unvaccinated children and citing Robert F. Kennedy, Jr.’s false claim that vaccines are making this generation of children the “sickest generation.”) Truly, Mr. Kirsch’s crapfest of a talk was a shining example of a Gish gallop, a.k.a. firehosing and proof positive that Brandolini was an optimist. (Brandolini’s law, also called the Bullshit Asymmetry Principle, states that it takes an order of magnitude more energy to refute bullshit than it does to create it. I say it takes at least two, if not three, orders of magnitude more energy, as evidenced by my going through Mr. Kirsch’s talk and 285 slides, plus his Substack articles.)

The reception to Steve Kirsch’s talk reminds me a bit of Marvin the Martian. Where’s the “kaboom” after his MOAR?

Other antivaxxers react

One of the funny things about Mr. Kirsch’s talk was the reaction. Sure, credulous propagandist hacks at the Brownstone Institute took it at face value, because of course they did. However, Mr. Kirsch is so…out there…that the more “reasonable” antivaxxers—or at least the ones who either have a science background or desperately want to be perceived as “more reasonable” were a bit less enthusiastic.

For example:

Let’s quote the whole thing, shall we:

Steve @ichudov was able to obtain the NZ database. He has been combing through it and has raised serious concerns with the quality of the raw data set. Red flags were first noted by @jikkyleaks followed by others chiming in with established qualifications whom you should recognize. At first look –too much missing data and anomalies. Too many discrepancies with other data points and facts we have about NZ. No wonder the host for the data set wants no one else looking under the hood, rather just uncritical acceptance of the so-called conclusions.

@NickHudsonCT explained having a prior opportunity to look at the NZ data and that project fizzling out because of problems–that event wasn’t disclosed to you apparently by the owner whistleblower. @USMortality thinks the data are so riddled with missing and problematic inconsistencies that it is rendered unanalyzable for purpose. @DowdEdward responded to the conversation indicating there may indeed be a scam underfoot. One aim potentially to undermine you, leaders in the movement. I suspect the focus on high rates of death purported to be associated with “vaccinator” is to shift blame from product to human error and thus muddy the waters of liability. NZ findings don’t match those of other recognized data analyses around the world showing the mRNA C19 product harms. There has never been a robust, large effect size for “vaccinator” that couldn’t be attributed to the confound of bad lots. Just occasional egregious human error cases, but not of this magnitude. There is no concordance offered between NZ lot ID#s and formal Pfizer batch #’s so no analysis has yet been offered showing the same pattern of AEs with bad lots. Why not? That would be straightforward and the pattern and scale of harms by lot known from VAERS and well documented by Craig Paardekooper at http://howbad.info should line up closely, match like fingerprints.

@sasha_latypova @hedleyrees

We don’t have that confirmation. We don’t have analyses that show the overall rate of death and AEs for the population and for age groups by total and number of shots as @denisrancourt excellent new world report lays out. Why wasn’t this sort of analysis performed? Let’s involve @JesslovesMJK @lawrie_dr and protect good doctors like @PierreKory @molsjames @drcole12 @richardursomd @P_McCulloughMD and @MdBreathe from being hurt by this likely psy op. We don’t have reports from NZ provided in a way that provides any sort of denominators. If the data set is legitimate it is poorly administered, full of holes. Perhaps due to widespread faults in monitoring and recording of events that would comprise the data set. Talk to other NZ experts on those processes. Let’s examine the hype and drama that preceded the rollout of the “findings.” A proper well-vetted set of analyses would go through a rational serious process of being reviewed by other experts in the open before being proclaimed the “Mother of All Revelations” and presented in parliament by legislators like @ABridgen

Winston’s demeanor and words are also highly suspect and caught the eye of many including those tagged here, and was my first tell. The full interview with Winston Smith was dripping with sniffly self-serving drama overkill and emotionally overplayed sensitivity to the plight of humanity. He is quite late to the party of outrage and has no believable excuse or response. It is a performance. No meaningful discussion of his real credentials, skills and experience. He is a database administrator but I see no evidence of skills as a statistician or analyst. He has worked with no one else by his own admission. By contrast @naomirwolf and her 2500 strong posse of experts worked months to sift through documents and data to produce interim and final reports. No one other than me has yet pointed out the highly strange clown world “coincidence” that Winston Smith is also a “fictional character and the protagonist of George Orwell’s dystopian 1949 novel Nineteen Eighty-Four…employed by Orwell as an everyman in the setting of the novel, a “central eye … [the reader] can readily identify with” https://en.wikipedia.org/wiki/Winston_Smith_(Nineteen_Eighty-Four) Dr. Malone @RWMaloneMD has been warning us about the advanced state of 5thGen warfare, and this entire event fits in with such a scheme from that playbook. @jjcouey has dubbed such shenanigans “Scooby Doo.”

@HopeRising19 has expressed concern along with @chrismartenson who commit to staying firm on sidelines with respect to buying the conclusions until further inspection and analysis takes place.

This is heartbreaking but fixable if the community works fast and together to self-correct. Take a breath, step back, and take a long second look. Work with Igor and a team of helpers (me included) and determine what really happened. What you discover at the end of that exploration may be much more useful to the world than what you think you found at the outset.

Jikkyleaks? She’s generally an antivaxxer at least as rabid as Mr. Kirsch is. As for Igor Chudov? Holy hell! He’s an antivaxxer whom, for reasons that I still don’t know, I’ve tolerated over in the comments section of my not-so-secret other blog. If these two are balking at Mr. Kirsch’s analysis, it must be bad. I am also very much amused at how so many antivaxxers think that Mr. Kirsch’s analysis is so bad that he fell for a “psyop” in which bad data were intentionally leaked in order to get someone like him to take the bait, do a bad analysis, and thereby discredit antivaxxers.

I mean, Igor Chudov even thinks it’s a psyop:

I do a lot of things. One of them is administering the database for Algebra.Com, a website with millions of monthly visitors and over a million of answered math questions.

So, I understand database administration. The story of a bona fide “leak” does not make sense to me. The data does not have the integrity that a full leaked data set would have.

This is supposed to be a payments database containing information for payments to vaccinators.

How can a payment database have such holes and missing data?

Was data selectively removed from the database before the leak?

How can batch IDs refer to multiple vaccines?

Did both the “whistleblower” and Liz Gunn honestly forget to check that these “deadly vaccine mass murder sites” are nursing homes?

Do the missing records of first vaccinations (doses 1-2) hide real vaccine deaths, making Liz Gunn go on about “deadly nursing homes” instead of looking at deaths actually caused by the COVID vaccine?

Was the “leak” a psyop and an intentional attempt to sow confusion, as it occurred with the old, pro-WEF, and vaccine-crazy NZ government still in place during the last days of it? This question is speculative, but something I would like to clarify.

I didn’t even get much into the claims by Mr. Kirsch that certain batches of vaccines were more “deadly” than others, because of many of the reasons above. I’m also amused by how Chudov thinks that maybe the holes in the database actually hide more COVID-19 vaccine deaths. I was laughing as I read his post. I also note with amusement that Mr. Kirsch, as he characteristically does, is desperately trying to persuade Mr. Chudov that he’s wrong.

As I read his post, I was predicting that, because Mr. Chudov is an antivax conspiracy theorist, he would eventually succumb to Mr. Kirsch’s blandishments and “come home.” In the meantime, though I found it utterly hilarious that antivaxxers were holding him up as some sort of expert on data integrity:

Unsurprisingly…as I was writing this, a reader pointed out to me that Mr. Chudov was already backpedaling, and he did it even faster than I had expected, which is greatly satisfying, as it demonstrates to me that I do know the full depths of his antivax crankitude:

At this point, I believe that Barry Young was more likely to be sincere than insincere in his intentions and actions.

My previous questions and comments about Liz Gunn’s statements about nursing home deaths and data quality still apply, with one exception: the partial nature of the data is explained by the fact that some shots were not paid through the system that Barry Young was supposedly administering. (I hope more clarity emerges).

This clarification is vital since I questioned the sincerity of the person who possibly risked his life to disclose data.

I greatly hope that, after thorough analysis, the data will yield useful information!

Antivaxxers are so predictable. Also, once again, Mr. Young’s sincerity or lack of sincerity is completely irrelevant to whether the dataset is complete or incomplete, biased or unbiased, and, most importantly, whether his interpretation of it (and Mr. Kirsch’s “analysis” of it) show what antivaxxers claim.

Then there’s Norman Fenton, he of the p-hacking to find badness in COVID-19 vaccines fame (not to mention fame from his inability to distinguish postviral bacterial pneumonia from COVID-19), who notes about Mr. Kirsch’s claim of “deadly batches”:

What we can probably discount is the claim concerning batches with exceptionally high mortality rates. The claim that these batches were especially deadly due to the contents of the vaccine or its delivery is confounded by their very different age and time of vaccination profiles.

Of course, when you see one claim that is so easily dismissed, you have to wonder about all the other claims. Moreover, Prof. Fenton clearly very much wants to believe Mr. Kirsch’s findings, so much so that he says that there is evidence “of increased risk the more doses one gets” but is oblivious to obvious confounding that comes from the fact that those at a higher risk of death are more likely to get the later boosters. That means that an increased death rate associated with a higher number of boosters does not necessarily imply that increasing number of doses causing increased risk of death. (Hat tip again to Prof. Morris, who helped me clarify something that I suspected but couldn’t quite put my finger on.)

In any event, it frequently amuses me when the self-identified “reasonable” antivaxxers appear to recognize a nonsensical analysis when they see it but, because they want to believe its results so badly, just can’t quite bring themselves to call it what it is, nonsense. That’s Fenton. He’s calling Mr. Kirsch’s analysis crap without actually being blunt about calling it crap.

I, on the other hand, have no such compunction. I call ’em as I see ’em, and I call this crap. I don’t know if it’s crap because Mr. Kirsch is so incompetent or if he’s a bit more competent than he appears and dishonest instead of clueless. Take your pick. Moreover, just because antivaxxers are claiming that this whole thing might be a psyop doesn’t mean that it isn’t organized. Barry “Winston Smith” Young was clearly working with groups that wanted to weaponize the dataset:

This actually reminds me of the whole “CDC whistleblower” conspiracy theory too. Remember that the “CDC whistleblower” William Thompson had been venting to Brian Hooker, who, apparently unable to resist chasing clout, had let Andrew Wakefield know that he was recording phone conversations with a CDC scientist. Andrew Wakefield, being Andrew Wakefield, couldn’t resist doing a brief video with clips from the phone conversations, thus blowing the lid off of the secrecy that Hooker had been keeping on his phone calls in the hopes of finding out still more information. Ultimately, Hooker decided to go all-in with Wakefield to publicize the “CDC whistleblower,” and, less than two years later, the saga produced the conspiracyfest of a pseudodocumentary VAXXED. I’m getting strong “CDC whistleblower” vibes, with Barry “Winston Smith” Young playing the role of William Thompson, whoever he had been working with playing the role of Brian Hooker, and, of course, Liz Gunn playing the role of Andrew Wakefield. It’s not perfect comparison, of course, because Mr. Young did apparently approach Mr. Kirsch, but you get the idea.

Conspiracy theorists tend to be proof positive of how difficult it is for any group of people to keep a secret, and Mr. Kirsch is just doing what antivaxxers have long done, “reanalyzing” a dataset based on incorrect assumptions and failing to correct for obvious confounders. To that tradition, he adds using a stolen dataset of uncertain provenance—remember, the CDC whistleblower’s dataset was the one ultimately used by the CDC to publish its results—whose acquisition involved a massive data breach and the violation of the privacy of millions of New Zealanders. Here’s hoping that Te Whatu Ora investigates fully, and we find out exactly how antivaxxers got their hands on such a dataset illegally and who besides Mr. Young was involved.