Thanks to everybody from yesterday. I saw nothing that made me change my mind, so I’m publishing my analysis as is. There is this new caveat about the data a reader at the Substack mirror discovered. But as I have always insisted, all analyses are conditional on the data.
Simplified summary which doesn’t excuse you from reading the analysis
Perhaps the vaccine killed a few, probably younger, people, though we cannot tell for sure.
The Data
About three weeks ago, Steve Kirsch gave and asked me to look at the New Zealand data. His analysis can be found here, among other posts at the same Substack. Now that he and others have had a chance to look at it and have their say, I thought I should give my additional perspective.
According to Steve, Barry Young, 56, a statistician, has been arrested by New Zealand for leaking the data. Perhaps because the raw data had names in it, or so I was informed. The news story is not clear, except reports said they were denying bail for Young. Incidentally, the data I was given was obfuscated. No names, just numbers, one per person.
The data, which has over 4 million records representing 2,215,725 individuals, consisted of a medical record number, vax batch ID, dose number (from 1 up to an absurd 32), vax date, death date (if any), vax name, date of birth. From these I calculated whether a person was dead or alive during the period of the data, the number of days after being vaxxed until death (if any). Because the number of shots was somewhat suspect in a handful of cases (like that 32), I grouped them into 1, 2, 3, 4, 5+ shots. (See the above top for a caveat on the data.)
The earliest shot date was 30 April 2021, and the latest was 1 October 2023. The earliest death date was 11 May 2021, and the last was 7 October 2023. Which means this is a very limited set of data: only two and a half years. All long term signals, of any kind, will not be here.
There was no cause of death given for anybody. Just death date for those who had at least one shot and died in this window. There can therefore be no certain proof of any cause of death.
Obviously, but since this is the internet I have to say it, all results are conditional on this data, and any other assumptions I make below. I of course could not vet the data, but I take it that it’s likely genuine, especially since they arrested Young.
Analysis
An insurmountable problem in ascribing cause is the lack of data on people who did not get any shots. Their death and age data is missing. There is no comparison group for the people who got shots. The only comparisons we can make is in people who got various number of shots, and with time.
This no-shot control data, should any be forthcoming, has to be from the same time period as the shot data, to be sure to capture covid and covid “solutions” and other deaths prevailing at that time. I mean, we can’t compare the shot data to control groups compiled earlier or later than these dates. Has to be during covid, because death cause distributions changed during that time.
Again, this means there is no natural comparison group and nothing about cause, therefore, can be said with certainty. Which does not mean that nothing can be said. What it does mean is that we’ll never know (using just this data) whether the number of deaths we saw were greater or lesser than the no-shot group.
Still, we can look at time and age of death with time of shot, and things like that. Our results can only be suggestive.
I only looked at the raw data and did all analyses on my own. There were some absurd numbers of shots, as I said, so I grouped all shots 5 and over as 5+. There were 4,193,438 records and the number of records with shots greater than 6 was 108, so this is not skewing results to any appreciable amount. There were about 10 records of people who had death dates before their vax dates. I removed these from all analyses.
First thing is monthly deaths, which shows the monthly count of deaths by number of shots. Obviously, a person who had only one shot and died could not have two shots, or more. Also, just as obviously, as time goes on the number of people with just 1 shot dramatically goes down, as people got more shots.
The red line is the count of deaths by month for those who had one shot and died, at any point, from (to us) unknown causes, and (for the moment) independent of age. And so on for the other shot numbers.
Here from WHO is the weekly count of ascribed covid deaths in New Zealand, over roughly the same period.
New Zealand, you will recall, was in the China-style group of countries whose Experts insisted their “solutions” were effective. China, for instance, reported just under 122 thousand covid deaths, out of a population of 1.41 billion, about half being ascribed to just January 2023! President Xi declared his “solutions” worked, and who could disagree? NZ only reported 3,522 total (as of 4 December 2023) covid deaths, out of population of 5.1 million.
Since, officially anyway, there were almost no ascribed covid deaths in 2021 in NZ, then the August peak in deaths for those who only had 1 or 2 shots could not be covid and had to be something else. Officially. But we have to be careful, here, because the possibility of an artefact of the short-nature of this data and shot-timing can exist. One does, which we will see below.
The same is true for the January 2022 peak, which had only a handful of official covid deaths, but which saw a large number of people with 3 shots dying. Only a few people with 1 or 2 or 4 or more died in this month (and, of course, those with just 1 or 2 shots who already died in August 2021 could not die again in January 2022!). The NZ government says these were not covid deaths. Therefore, they were something else.
The next peak in deaths is those with 4 shots (and not any other number), in July 2022. That was the month with the largest officially ascribed covid deaths in NZ. Very few people with other shot quantities died during that month. And the same kind of thing, as you can see, happened with those with 5+ shot in May 2022, which was also a minor high in officially ascribed covid deaths.
Now the number of people who got just 1 shot, just 2, and so on, and the percentage who later died in the dataset’s time window, was this:
Shots, N, % Died in window
1, 149,583, 0.98%
2, 533,835, 0.98%
3, 708,313, 2.0%
4, 453,883, 2.8%
5+, 370,111, 1.0%
The peaks in deaths of 3 and 4 shots also coincided with peaks in official covid deaths. Again, a person who got 1 shot and died could not get 2, etc. And do not forget the range of the data.
Next is distribution of days until death after a person’s last shot, which is just that: the number of days until death for those with just 1 shot, just 2, and so on. The shots=3 sticks up the most because 3 shots was the most. Followed by 2, then 4, then 5+, then 1, the smallest group.
If shots were killing some people more or less right away, we’d expect a bump close to the shot day. But you’d also expect that if the shot was killing people, the ones most susceptible would die after the first, or maybe second, shot. Which the first plot above suggests. People would not be as likely to die after 3 or more if they tolerated the first 2, unless there was a cumulative effect. Which we could not reliably see in this short-time-period data.
Also, as time went on, as we will see, older people got more shots, while younger people stayed with 1 or 2, maybe 3.
At any rate, maybe there’s a small signal in the 1-shot group. But other things can account for the early bump, too. Like covid being newer, and people getting the shot not quite in time. Yet, officially anyway, this can not be the explanation, and the deaths have to be from something besides covid. But that doesn’t logically imply it was only the vax: other natural causes could do this.
Now you might be tempted to infer that because there is a rounded bump in small number of days until death for each shot number, followed by a falling off, that this is conclusive proof the vax killed people. This is not so. Take a look at this picture to see why.
This is the date of the shots by the number of days until death, for each total shot. You will notice the triangular shape.
Consider those who got shot 1 right at the beginning of the dataset. And suppose for the sake of argument the shot had no effect whatsoever (good or bad). Then we would expect a roughly uniform number of days until death in any population, people dying more or less evenly on days from 1 to 850, which is the maximum. Make sure you understand this. Pick an early shot date, and then look vertically at the dots at that date: are they uniform or uniform-ish?
If you’re not following this, here is an example, using the people who only got one shot. I’ve erased all but those who got the shot on about the same short number of days and later died. The distribution of days until death should be more or less evenly spread up and down the vertical axis. If the vax does nothing.
If the vax was healthful we’d expect more dots near the top, with a relative dearth at the bottom days: because people lived longer. If the vax was harmful, we’d expect more dots clustered at the bottom with a dearth at top: because people are dying soon after the shot. All things equal. Which we can’t know, so we cannot claim anything definitively.
Those who got the first shot sometime near the middle of the dataset period could only have a distribution of deaths from 1 to about 425 days. And those who got the first shot near the end of the dataset could only have a uniform number of deaths days from 1 to some low number.
That means two things: (1) the triangle shape is explained because the maximum number of days until death can never exceed the number of days left in the dataset from the shot day, and (2) we should certainly see what we see above, a bump in low numbers of days followed by a tailing off, because people got the shot at various days throughout the time period, which means where must be lower number of days until death.
In other words, this bump in the number of days until death after the shot is entirely expected, if the shot does nothing whatsoever.
If you haven’t seen this yet, look at the 5+ shot group. They were only able to get the shot at the end of the dataset time period, and necessarily the number of days until death for those who died must be small.
That being said, look at shot = 1 for the people who got the shot somewhat early on, blown up here:
The yellow circle indicates a possible departure from the uniformity. It is a cluster of deaths quite close to shot day. The same kind of thing is also present in shot = 2, but with slightly more spread.
I view these as suspicious and as real candidates for vaccine injury. We could, if NZ wanted to cooperate, identify all these people, and their exact causes of death could be investigated.
Notice, too, that this kind of early clumping in days until death after the shot disappears (or is not as evident) for those who died with greater number of shots, which is consistent with our theory that if the vax is going to injure you, it would likely do so with the first or second shot. And once these people are dead, they can’t die again after more shots!
We might be able to see more in the next plot, which is age at death. Which is what it says. Age at death by shot number.
There’s a small increase in younger deaths, especially for those with only 1 or 2 shots, and maybe 3. We are comparing the shape of the distributions, here, across those with different numbers of shots, and are not comparing absolute counts. All other things equal, if the shots had nothing to do with age at death, the shapes would be the same. But, of course, all things are not equal, like the timing of reported official covid deaths. And the timing of “advertisements” and official “requests” to get boosted, etc.
In the hopes of clearing up some of that, we have the distributions of age at shot, which shows the ages at which people got the shots, for each number of shots, split by those who lived and those who later died.
Those who lived who got 1 shot then the next, well, you can see how much they aged in between shots right on the plot. And here we see partially that older folks tended to get more shots (see the big peaks in young ages for 1 and 2 shots only). Which is no surprise.
Again, if the shots were killing people, we might expect them to die soon thereafter (in this short-window dataset), especially if they were young and hadn’t had some immunity to shots (and not just the bug) as the old might possibly have by living through any number of infections in their life not very dissimilar to covid or the covid vax. Ccoronaviruses, don’t forget, are one of the causes of common cold.
If there’s any signal, it’s here, as above. Especially after 1 or 2 shots for those 30 and less who died, the second shot coming not that long after the first.
The thing that’s most noticeable is the shape of age distributions at death, which looks the same, roughly, regardless of shot number. Which is precisely what we’d expect if the shots did not cause death. Except, as you can see, for shot = 1 and a little bit at shot = 2, where the young died.
But, perhaps, they could have just died from the bug, too, being sickly or weak. We might be able to see that in these next two (final) plots. The first is the age at shot by shot date, for those who received just 1 shot, just 2, etc.
This is sort of fun because you can see the age-phased rollouts of shots clearly. The youngest did not start getting shots until January 2022—and recall the WHO says covid deaths started ramping up around February–March of that year. The young also got fewer shots: the youngest age rises for each shot increase. This confirms it’s the young, largely, who only had 1 or 2 shots only.
Many of the youngest people (only those from 0 to 10) who only had 1 or 2 shots had a peak in deaths before they got their shots, back in August of 2021. What suspects are there? Not the vax. Government lying about covid? Or is this the same figment of the limited time nature of this data?
On the other hand, those who were 11 and older started getting their shots around the same time of the peak in official covid deaths, in August 2021. So the vax being a cause of death is a real possibility for them. But there not a lot of deaths in this age group: only 6. Here’s the breakdown:
Age bucket, Number of dead
0-10, 6
11-20, 102
21-30, 243
31-60, 2,982
61+, 33,978
As you can see, it’s going to be difficult to guess causes of death in the younger ages because of the small sample size. We might have luck with the oldest, given the larger sample size.
Let’s next look at the age of death by date of death, by number of shots:
The most obvious signal is that the age distribution at death creeps up with shot number, again because the young tended to get fewer shots, and got their first ones later than the older. The 5+ shots came at the end of the time period of the data, which explains why those with the most shots didn’t start dying until the end.
The younger died it seems, roughly, at a fairly routine clip throughout the time period of this dataset. We see the same curiosity of the clustering of deaths in 2021 up through the end of the years, for those who just had 1 or 2 shots (and recall after they died, they couldn’t have more shots!). This reinforces the notion that NZ rulers lied, as many lied, about covid deaths.
Conclusion
Did the vax kill anybody? Very likely. Because all vaxs do, and there is nothing anywhere which shows the various covid vaxs were better in this respect. And there is much (though not necessarily here) that shows they were worse. Like the endless propaganda that vaccines are “safe and effective.” And that if you got the shot, you couldn’t get sick or pass the bug on. Official government agencies told this lie endlessly.
What about this data? There are hints, as described above. I think the signal of Expert and ruler lying is stronger than any signals of vax-caused death, except where noted in the first and possibly second shot. But this surely reflects my bias against lying and boastful rulers and Experts. Well, we’re all used to governments lying by now.
We cannot tell cause from this data. Not with any certainty. We also can’t tell if the shot prolonged anybody’s life. No control group again. But, if New Zealand could release the age and date of death of people with no shots for this same time period, then we’d really be in business. There’s no reason not to, not after the shot data is already out there.
Control data could also prove, as Experts claim, if the vax prolonged lives. Why should we trust them that it did? Answer: there is no reason to trust them. Even better is to have data with official causes of death. Think we’ll get it? As Santa would say: Ho Ho Ho.
This analysis, which is in conjunction to all the other work out there, will please no one. It does not say the vax was a great killer, nor does it say the vax was harmless. Because, using this data alone, we are very limited in what we can say.
Yet I could well be wrong. There is no control group. I am not claiming cause, either. I also did not like at vax batch or manufacturer, which might have signals not apparent in the main data.
Note that this analysis is just looking at the data. No formal models. No survival analysis or cohorts or anything like that, which I believe is just not needed here. Because it would be too likely to obscure what happened. We have more than enough data to say “Here is what happened” and do not need to smooth things over with formal models.
Subscribe or donate to support this site and its wholly independent host using credit card click here. Or use the paid subscription at Substack. Cash App: $WilliamMBriggs. For Zelle, use my email: matt@wmbriggs.com, and please include yours so I know who to thank.
Covid was overblown and vaccine safety concerns are overblown. One of my takeaways from your excellent analysis is to confirm my choice not to get the vaccine because it’s pointless and potentially dangerous.
Any correlation between with batch?
I had a seat of the pants calculation in 2021 that vaccines could/would kill ~1/10000 per shot. Since I already had the ‘rona and have an immune system, it seemed pretty dumb thing to take. My doctor (not in America) said the same thing.
Of course all of this is quite damning: even if vaccines killed nobody, they and the other interventions pretty obviously didn’t do any good.
Leo,
I didn’t look seriously at it, but so far, no.
Scott,
Exactly so.
Thanks for deviling into the data.
It’s a bit late now but would it have been helpful for your analysis if the age binning step (0-10, 10-20, etc) documented/recorded at the start the patients record keeping had been skipped and replaced with a month and year of birth metric for each patient?
Back in my youth Marty always wanted our clinical trial data tracked this way as he and marketing wanted to know how our products worked in the neonate wards as well as in the hospitals and nursing homes. One can’t say much about neonates if their results were averaged out in a 0-10 bin.
The vax did kill somebody. You can get the data from the CDC. The viruses killed unvaccinated people too. Were there more deaths among unvaccinated people? We know the answer and can easily imagine why or explain it away.
What are measures of vaccine efficacy? (FYI: https://en.wikipedia.org/wiki/Polio_vaccine) I wouldn’t want to spend time on the data here, however, I probably would on the clinical trial data.
I don’t think you can make inferences using ceteris paribus. There are many factors to be considered, some of which cannot be fixed. For example, it is not feasible to keep the pathogenicity and virulence of various strains constant. Univariate analysis might give you directions for further investigations but won’t entail any definitive conclusions.
My hourly charge is $500. It took me 40 minutes to read this post and write comments.
I don’t disagree with your conclusions, nebbish though they might be, but I would rather have seen survival curves, hazard functions, half-lives, etc. possibly regressed against covariates like age and jab date (easy in R). The foggy cluster visual model doesn’t appeal.
JH,
I charge $5,000 an hour. And by the time I figured out you did not disagree with anything I said, you racked up quite a bill.
I’ll send an invoice for the difference, unless you’re interested in some kind of barter.
Uncle Mike,
No models needed here, which would only obfuscate things. Recall I am nowhere interested in parameters, but observables.
Wow, deja vu all over again. Sensationalized hyperbolic claims about a deeply concerning topic, a curiously uncurious press parroting the state party line, finishing up with yet another frustrating case of scandalus interruptus when the adults show up. Yes, the New Zealand Ministry of Health freaking out over the data leak just might have more to do with protecting their lying posteriors than covering up a smoking vaxicide gun.
This got me thinking about the unintended consequences of the US FOIA. There’s the obvious, such as thoroughly corrupt politicians like Hillary Clinton choosing to break the law with an illegal email server to keep evidence of her treasonous pay-to-play get-rich schemes out of reach of FOIA requests. Less obvious perhaps, but certainly less visible, is the information that was not generated or collected in the first place out of fear of the FOIA. One would think that Covid and Covid vaxes would be the most thoroughly documented subject in human history, given the initial panicked mass-mortality predictions. Instead it supercharged the distinction between good truth and bad truth, with a determined effort to hide the latter. Hopefully enough data has been collected and will someday be public domain, letting the chips fall where they may.
On a completely unrelated note totally unrelated to this particular blog post or any of the significant effort that went into producing it, does anyone know of a good way to donate anonymously to a stranger? If you have an address, you can still stick some cash in an envelope with a fake return address. Does the Internet have anything to offer along these lines, or is anonymity a necessary victim of a cashless society? (Patreon comes pretty close, but I don’t know if they can be trusted – PayPal clearly can’t be trusted).
Briggs, your analysis makes me happy. People aren’t dying like flies from the vexx. Some are dying, some are sickened, but it isn’t wholesale mass-murder. Not yet anyway. Most jabbees seem to be just fine. I don’t know anyone who croaked from it, or got sick. Looks like we were spared the v’axe, in spite of so many misplacing their trust in the demonic, money-power malthusians and their fool followers. Hooray, and thanks be to almighty Providence.
Hopefully, this experience has made more people aware that our overlords are not to be trusted. Hopefully, this mild betrayal has immunized people against greater betrayals to come. If so, two cheers for the vexx. And three cheers for Briggs and his reliably honest insights.
Hagfish, If I can make you happy, then I am happy.
There are theoretical scenarios where a mostly vXed country would have mortality per time period in any vX status looking lower for the vXedvwhen compared to that countries unVxed pop but at the same time would overall have higher mortality rates than another similar but mostly unVxed country with both countries being subjected to similar possible initial infection dynamics before any vXing programs get started.
Briggs,
Funny… I thought you mainly made non-definitive conclusions in this post. I did pick one that I have a definite answer to. And made a quick comment. (If my comments are wrong, I welcome corrections.) There are reasons why some analyses, e.g., univariate time series plots, are called explanatory analyses. Possible explanations are not “beliefs” to be agreed or disagreed.
Ah.. you have a prejudice against me.
I read somewhere that your rich friend Kirsh has been giving money to researchers to find what he wants to hear. So, take his money. I am willing to spend time reading your analysis and making suggestions. Ain’t I great?
“exploratory” not “explanatory”
Online goblins are real.
JH,
Alas, Steve has not paid, and indeed does not love my analysis. I shall instead get rich by relying on you honoring the invoice I send.
The purposes of the “vaccines” and even vaccines are fertility reduction. Gates has made that clear for one. The authorities and their authorized experts are never going to fess up to the “vaccines” and vaccines being in any way harmful, except maybe around the edges to cover the bases. In the case of “vaccines” the coof will be blamed and who’s to say otherwise, what with PCR tests and such scientism being deployed. They can always roll out another plandemic, after all how can genpop complain when they bent over so fulsomely last time. Nothing has changed the same crooks and liars are in charge or identical replacements. The body counts and side effects of government action have to be astronomical for genpop to get a clue and (only temporarily) snap out of their trance, QED ad nauseam. I enjoy peeking behind the curtain as much as the rest of you but this is entirely for entertainment purposes for heretics like us. The comments section here appears to me to show that the regular commentators are here for the schadenfreude of watching the world burn as am I. I re-watched “Escape from LA” last night and Snake pressing the TV remote to EMP the world back into the stone age at the end seemed so much the right and moral choice, even more so than when I first watched it. “Welcome to the human race.”
This was fun! But I have some – dare I say – insights and a quibble. First the quibble.
> Many of the youngest people (only those from 0 to 10) who only had 1 or 2 shots had a peak in deaths before they got their shots, back in August of 2021.
Where did you get this? o.O Is this from some outside dataset, like the news? Since the dataset you analysed doesn’t include those who didn’t get shots, this bit of information you shared can’t possibly come from the dataset you analysed.
Now the other thing: when I look at graph/chart/whatever of “days to death vs. date of shot for one shot” (the one with the yellow circle) and after that look at “age of death by date of death, by number of shots”, it seems rather obvious to me that the cluster of people who died soon after getting the shot came from the cluster of people who were old when they died after getting their shot. The number of people who were young when they died after getting their one shot, in that timeframe (time slice), doesn’t seem big enough to generate the cluster of people who died quickly after the one shot. To my eyes, it seems more plausible to explain the yellow-circled cluster as old people who were about to die anyway but who got the shot.
The anomalous bumps in the age distribution of dead for one shot and two shots can be seen in the graph “age of death by date of death, by number of shots”. We can see there is a greater concentration of the dots in the bottom part of the graphs for 1 and 2 shots compared to the bottom parts of the 3 and more shots.
But I would caution against just looking at the “age of death by date of death, by number of shots” without also taking into account the total number of people who had “by number of shots”. If there were a lot of people who had the right number of shots to generate the deaths, the deaths might get skewed up. Was the vax propaganda more effective on young people? Perhaps the semi-coercive methods of restricting access to movies and concerts to the unvaxxed meant something to the young and therefore they went and got themselves vaccinated to a greater degree than the older, married, people? Suggestion for analysis: perhaps if you were to normalize the deaths in an age group of X shots in a time slice to the number of people of that age who had X shots in that time slice, you would see something. This can be represented as a table, perhaps of numbers or colored tiles.
I don’t know if I was clear and understandable in my comment. There are many attributes in these terms, it might be best to talk in formulas… that we can’t write in plaintext. :/ I reread and fixed the comment, I hope I managed to communicate my ideas well.
P.S. Where can one get the dataset to play with it? Is the dataset contraband? If you were to kill the medical record number, would the dataset be clean and therefore shareable?
There is a problem with finding an “unvaccinated” cohort to use as a control group; which government data set might actually have such a category at all, and in a form that is searchable? The officials have certainly been claiming that deaths and medical events are “higher in the unvaccinated than the vaccinated” but what they don’t know, is how many people are outside their database altogether? If they are using, say, a “health records” data base, it is quite possible that there are people in their country but not even on that data base. They WILL have included everyone who is vaccinated because by default, they have all those people on their data base. But they don’t know who they don’t have on their data base, and every single person they don’t have on their data base will be unvaccinated. If such people die or suffer a health event, they will end up in the records; but what no one knows is how many such people have not died or suffered a health event. Whoever is calculating the mortality or health debacles of “the unvaccinated” needs to be either leaving out the people they didn’t know about until the death or the medical event, or making a credible estimate of how many more people exist outside their data base and not vaccinated.
A further massive problem that needs to be honestly teased out, is the principle that the officials use almost everywhere, to classify as “unvaccinated” someone who dies or suffers a medical event during a 2, 3 or even 6 week period AFTER the vaccination – “because the vaccination didn’t have time to provide the expected protection”. Briggs and his readers will instantly see the problem with this. Barry Young’s NZ data set obviously avoids this problem because it makes no such distinction; the data is for “vaccinations paid for”. But if officials are left to do the “vaccinated versus unvaccinated control group” comparison, what kind of illogic and dishonesty might result?
Oh, and by the way, Barry Young does identify a massive “batch” problem; and also a “vaccinator” problem – but this surely has to be the same thing as the batch problem, as vaccinator incompetence surely can’t match the elevated outcomes for bad batches in the event that the vaccinator was using good batches. I understand that inadvertent IV injection is a problem, but surely even a skilled nurse malevolently trying to find a blood vessel in the deltoid muscle with the needle, would not be able to achieve the elevated bad outcomes that would match “bad batch” outcomes?
Dr. Briggs:
Speaking of Kirsch, what do you make of his proposed time-series cohort analysis of record level vaccination/death data? He has issued this challenge to the state keepers of such data to do his analysis and claims it will definitively settle the question of covid vaccine safety. Is his approach sound?
I may have an opportunity to pitch this study to my state’s HHS, but I was hesitant until I learn more about it.
Pingback: “Why Didn’t Use You The [fill in the blank] Model On That NZ Data, Briggs?” – William M. Briggs