I am so tired of the narcissistic mutterings of “experts.” What they do is not mysterious and exclusive. Most intelligent people can arrive at their own valid conclusions if they have access to the data in useable form. Thank you for providing this! Merry Christmas!
First off, Thanks so much for your effort and taking the time out to do this quickly. Merry Christmas to you and your family.
Second. These charts are really harrowing. I'm praying there is an error somewhere because I'm afraid now to ask for data from December 7 to Now. I just know what we will see there and it's just unthinkable what's happened and is happening right now. That September spike?!
The "Waning" of vaccine effectiveness was real-and the booster did reverse it, by restoring the high death rate of the susceptible before they could catch Covid. If this is true, it's really astounding that people were sitting on this data and never explained that the denominator changing was the illusion of vaccine efficacy.
Great analysis, thanks! It would be great if you could provide a more extended explanation of different lines in your graphs, esp. what 1st and 2nd vaccinated mean? 2+21 for 2nd? What about 1st? And give your take why sometimes the unvaccinated fared worse than vaccinated on your graphs (Jan-Apr)? To be complete and so ppl's eyes don't glaze over. Also, when did vaxx started in England, what "vaccines" at what times to what age groups?
Thanks, Andreas. All of this is in the bulletin. I would suggest reading that first to get all the definitions. It should be done anyway to get the context of my re-analysis. I haven't changed anything except show the four vaccination statuses aggregated as I said. Oh, I also combined their two flavours of dose 1 and dose 2. I'll mention that too.
OK, 1st is also 1+21. So, I would speculate that the deaths of the "unvaccinated" include deaths of the jabbed withing 21 days of each jab. Is that correct? OK, you say you were able to add those 21 days back to the 1st and 2nd lines. So, I have been mistaken. Why do jabbed fare better than unjabbed, in Jan.-Apr., for a week or two, at different times for different age groups? That is the remaining question.
No, they finally split dose 1 <21 days and >21 days, same with dose 2 so I don't think they are misclassifying this time round.
For the second question, as they say in their bulletin, mortality rates can be skewed when the population denominator is very small and there is great idiosyncrasy in the health of the deceased.
One certain observation from your charts: the mortality of the unvaccinated keeps trending down all the way to Oct.! Whereas 2nd jab, while helping some initially, wore off by Sep., but the price of "helping some" (getting to the bunker) was the sacrifice of such large number of 1st jabbed, that it wasn't worth it even then. It's just the killing fields anywhere you look! Any data on the 3rd jab effects? I guess that spike from Sep. on, for older age groups, is the 3rd jab effect. OMG! Covid "vaccinations" kill, at an accelerating clip! Why isn't it on the front pages of all newspapers? And 1st news of every news broadcast?
The only way for jabs to be beneficial is for the red line to drop under the green line, and from your charts it's obvious that it is not happening, not a chance! That's why the NEED SO BADLY to vaxx every last person, to completely eliminate the controls. So their crime becomes "hypothetical" and impossible to prove, in the long run.
"You are dying, but imagine how bad it would be without the vaccination!" kinda thing.
Suppose you are observing 400 people for 2 years and you want to compare them to another group where you have 800 people for 1 year.
You can do so by using 1 person year as a unit where each person's total exposure time is added. So 400 people for 2 years is 800 person years. And 800 people observed/exposed for 1 year is 800 person-years.
In other words it's trying to standardize the incidence rate of something so that we are comparing the same level of exposure. The obvious problem is that for sometimes exchanging more people for less observation time can skew results or vice versa. However it's the best option we have to try to measure apples to apples.
An example of this kind of skewing would be in the Pfizer trial.
They measured the incidence rate of Symptomatic SARS-COV-2 infection at 28 days after the first dose for people that at baseline (before the trial started) tested positive for N-antibody and PCR positive for SARS-COV-2. Basically, checking what happens to people who are or were asymptomatically infected with antibodies suggesting recovery almost complete and perhaps infection cleared.
The Vaccinated group had one incident and the placebo group also had 1 incident. However the vaccine efficacy for this group calculated was -20%. Why? Because they had to observe a larger number of people in the placebo group for longer to get that single incident. Therefore, it took fewer/shorter "person-years" for the vaccinated to be "re-infected" compared to placebo.
The skew is significant when there is no limiting principle such as the number of people you can observe or how long you can wait because if the total incidence rate will be same if we wait long enough, it will superficially appear that the incidence rate is the same when in fact, in one group something happened faster with fewer people suggesting higher hazard.
The real scandal is that this group increases in size faster as the pandemic progresses through mass vaccination and the actual vaccine danger is far higher than the blind RCT suggests in actual risk groups who are observed where incidence rate is likely to be high. This time last year, the VIVALDI study in UK care homes showed that we restarted the Alpha variant pandemic by vaccinating our recovered seniors reinfecting them at 2x the rate of unvaccinated. This is real reason why we will never ever see a breakdown of Omicron cases in reinfection by recent vaccination status from either Denmark or England, even as that number shoots through the roof. It's propaganda to convince the naturally immune to give up their immunity and become carriers and incubators of the next variant (like Alpha last year.)
Source for 2x claim- (also demonstrates how order and incidence rates matter in aggregated statistics because "regression models" that assume independent events is clearly flawed. Newly vaccinated transmit and infect unvaccinated through community transmission and therefore the incidence densities in the beginning make all future events not independent unless one proves zero causal pathway between high vaccination infection rates in the beginning, high viral shedding of immune evading and transmissible variants and the increasing hazard for unvaccinated relatively late in the epidemic.
Metatron I am confused about the "Total" line in your graphs. Since we are looking at rates here, not absolute numbers, shouldn't the Total be sort of an average of the other category rates?
Yes but to do it properly I need the numerators and denominators which they do not provide! Besides, there is inevitably little overlap between the two doses. It wouldn't change much. But you are correct, I probably should have taken the average rather than the sum even though it will differ from the weighted average.
That is a different analysis. The 10-59 year old data was not "corrected" for the misclassification error but still showed an apparent increase in ASMR for the vaccinated. For such a wide age range though, this is due to age confounding. There are lots of things to be aware of and control for to get to the truth. This is why it's always a good idea to completely disregard anyone who just puts out an attention-grabbing headline!
I believe Mathew Crawford did attempt to correct by comparing age-based mortality rates, and assumed that the non-vaccinated group became progressively younger through the year, as older people were vaccinated first. However I tend to trust your graphs more since they are closer to the raw data (no fudge factors to correct for age confounding). Thanks again
For a single age cohort it is indeed simply deaths / person-years * 100k. The report gives the person-years for each age group and each month.
For a multiple age cohort you take the weighted average using the European Standard Population 2013.
Unfortunately, the ESP 2013 is done in 5-year brackets and the monthly data is stratified in 10 year brackets so 40-49 was the only one I could re-analyse because the report calculates each part of the age group separately and then provides the weighted average.
For other age groups I could solve for the split but it was quicker to just pick 40 to 49s!
My problem is the following: let's consider for example the age cohort 5-9 years, and let's assume that we have for this age group both the deaths per month and the person-years.
Now, if we see a peak of deaths in a certain month, we cannot be sure that this is due to an increase of mortality, if we are considering the person-years related to the 1st of January (as it could also be for example due to the fact that in that month the number of 5-9 years old increased).
For this reason I was looking for population for each age groups, but divided by month or by week.
Do you know if these data are available? I have been looking for these the last days but I did not manage to find them.
No, person years is not related to the age of each person, it is a function of the time that person spends in the vaccination cohort. So if you had 12 million people and they spent 1 month in the unvaccinated cohort, that would equate to 1 million person years. That's my understanding anyway.
Thanks Metatron. But I don't think you can add up rates in this way. E.g. if you have 50/100k rates for vaccinated 1dose and 2dose, the combined rate (1+2dose) is still 50/100k, as you have to add numerators and denominators. Or perhaps I am missing something ?
Yes but to do it properly I need the numerators and denominators which they do not provide! Besides, there is inevitably little overlap between the two doses. It wouldn't change much. But you are correct, I probably should have taken the average rather than the sum even though it will differ from the weighted average.
I am so tired of the narcissistic mutterings of “experts.” What they do is not mysterious and exclusive. Most intelligent people can arrive at their own valid conclusions if they have access to the data in useable form. Thank you for providing this! Merry Christmas!
First off, Thanks so much for your effort and taking the time out to do this quickly. Merry Christmas to you and your family.
Second. These charts are really harrowing. I'm praying there is an error somewhere because I'm afraid now to ask for data from December 7 to Now. I just know what we will see there and it's just unthinkable what's happened and is happening right now. That September spike?!
The "Waning" of vaccine effectiveness was real-and the booster did reverse it, by restoring the high death rate of the susceptible before they could catch Covid. If this is true, it's really astounding that people were sitting on this data and never explained that the denominator changing was the illusion of vaccine efficacy.
Great work thank you
Have you seen the AIER email release? Or , read the smoking Fauci gun piece in Brownstone? Or my substack...
Mandatory information...
Re response to GBD...
Great analysis, thanks! It would be great if you could provide a more extended explanation of different lines in your graphs, esp. what 1st and 2nd vaccinated mean? 2+21 for 2nd? What about 1st? And give your take why sometimes the unvaccinated fared worse than vaccinated on your graphs (Jan-Apr)? To be complete and so ppl's eyes don't glaze over. Also, when did vaxx started in England, what "vaccines" at what times to what age groups?
Thanks, Andreas. All of this is in the bulletin. I would suggest reading that first to get all the definitions. It should be done anyway to get the context of my re-analysis. I haven't changed anything except show the four vaccination statuses aggregated as I said. Oh, I also combined their two flavours of dose 1 and dose 2. I'll mention that too.
OK, 1st is also 1+21. So, I would speculate that the deaths of the "unvaccinated" include deaths of the jabbed withing 21 days of each jab. Is that correct? OK, you say you were able to add those 21 days back to the 1st and 2nd lines. So, I have been mistaken. Why do jabbed fare better than unjabbed, in Jan.-Apr., for a week or two, at different times for different age groups? That is the remaining question.
No, they finally split dose 1 <21 days and >21 days, same with dose 2 so I don't think they are misclassifying this time round.
For the second question, as they say in their bulletin, mortality rates can be skewed when the population denominator is very small and there is great idiosyncrasy in the health of the deceased.
One certain observation from your charts: the mortality of the unvaccinated keeps trending down all the way to Oct.! Whereas 2nd jab, while helping some initially, wore off by Sep., but the price of "helping some" (getting to the bunker) was the sacrifice of such large number of 1st jabbed, that it wasn't worth it even then. It's just the killing fields anywhere you look! Any data on the 3rd jab effects? I guess that spike from Sep. on, for older age groups, is the 3rd jab effect. OMG! Covid "vaccinations" kill, at an accelerating clip! Why isn't it on the front pages of all newspapers? And 1st news of every news broadcast?
The only way for jabs to be beneficial is for the red line to drop under the green line, and from your charts it's obvious that it is not happening, not a chance! That's why the NEED SO BADLY to vaxx every last person, to completely eliminate the controls. So their crime becomes "hypothetical" and impossible to prove, in the long run.
"You are dying, but imagine how bad it would be without the vaccination!" kinda thing.
What are "Person-Years"?
Suppose you are observing 400 people for 2 years and you want to compare them to another group where you have 800 people for 1 year.
You can do so by using 1 person year as a unit where each person's total exposure time is added. So 400 people for 2 years is 800 person years. And 800 people observed/exposed for 1 year is 800 person-years.
In other words it's trying to standardize the incidence rate of something so that we are comparing the same level of exposure. The obvious problem is that for sometimes exchanging more people for less observation time can skew results or vice versa. However it's the best option we have to try to measure apples to apples.
An example of this kind of skewing would be in the Pfizer trial.
They measured the incidence rate of Symptomatic SARS-COV-2 infection at 28 days after the first dose for people that at baseline (before the trial started) tested positive for N-antibody and PCR positive for SARS-COV-2. Basically, checking what happens to people who are or were asymptomatically infected with antibodies suggesting recovery almost complete and perhaps infection cleared.
The Vaccinated group had one incident and the placebo group also had 1 incident. However the vaccine efficacy for this group calculated was -20%. Why? Because they had to observe a larger number of people in the placebo group for longer to get that single incident. Therefore, it took fewer/shorter "person-years" for the vaccinated to be "re-infected" compared to placebo.
The skew is significant when there is no limiting principle such as the number of people you can observe or how long you can wait because if the total incidence rate will be same if we wait long enough, it will superficially appear that the incidence rate is the same when in fact, in one group something happened faster with fewer people suggesting higher hazard.
This is an excellent explanation. Thank you.
The real scandal is that this group increases in size faster as the pandemic progresses through mass vaccination and the actual vaccine danger is far higher than the blind RCT suggests in actual risk groups who are observed where incidence rate is likely to be high. This time last year, the VIVALDI study in UK care homes showed that we restarted the Alpha variant pandemic by vaccinating our recovered seniors reinfecting them at 2x the rate of unvaccinated. This is real reason why we will never ever see a breakdown of Omicron cases in reinfection by recent vaccination status from either Denmark or England, even as that number shoots through the roof. It's propaganda to convince the naturally immune to give up their immunity and become carriers and incubators of the next variant (like Alpha last year.)
Source for 2x claim- (also demonstrates how order and incidence rates matter in aggregated statistics because "regression models" that assume independent events is clearly flawed. Newly vaccinated transmit and infect unvaccinated through community transmission and therefore the incidence densities in the beginning make all future events not independent unless one proves zero causal pathway between high vaccination infection rates in the beginning, high viral shedding of immune evading and transmissible variants and the increasing hazard for unvaccinated relatively late in the epidemic.
https://twitter.com/mahmudme01/status/1448644758872621069?s=20
Metatron I am confused about the "Total" line in your graphs. Since we are looking at rates here, not absolute numbers, shouldn't the Total be sort of an average of the other category rates?
Yes but to do it properly I need the numerators and denominators which they do not provide! Besides, there is inevitably little overlap between the two doses. It wouldn't change much. But you are correct, I probably should have taken the average rather than the sum even though it will differ from the weighted average.
Yes, then I'd say just take out the "total" all together. What is important in the comparison between the different categories, right?
Can't reproduce the results except for 18-39 age group
Hi Joel, I just discovered your substack. Excellent work !
In an earlier study Mathew Crawford, showed that there was virtually no difference in all cause mortality between vaccinated and non vaccinated. Do you have an explanation for the difference ? https://roundingtheearth.substack.com/p/uk-data-shows-no-all-cause-mortality
That is a different analysis. The 10-59 year old data was not "corrected" for the misclassification error but still showed an apparent increase in ASMR for the vaccinated. For such a wide age range though, this is due to age confounding. There are lots of things to be aware of and control for to get to the truth. This is why it's always a good idea to completely disregard anyone who just puts out an attention-grabbing headline!
I believe Mathew Crawford did attempt to correct by comparing age-based mortality rates, and assumed that the non-vaccinated group became progressively younger through the year, as older people were vaccinated first. However I tend to trust your graphs more since they are closer to the raw data (no fudge factors to correct for age confounding). Thanks again
Interesting analysis, as usual by you. Thank you for the extensive job you are providing on this.
A question.
I guess you calculated for each month x, the mortality / 100k Persons Years as it follows:
mortality / 100k Persons Years (x) = ( deaths_age_group (x) / population_age_group (x) ) * 100000
Is it correct?
In positive case, where did you get the data for population for the age group, for each month?
Or did you assume population_age_group(x)=population_age_group(January), that is, you set it constant for the year?
Thank you in advance.
For a single age cohort it is indeed simply deaths / person-years * 100k. The report gives the person-years for each age group and each month.
For a multiple age cohort you take the weighted average using the European Standard Population 2013.
Unfortunately, the ESP 2013 is done in 5-year brackets and the monthly data is stratified in 10 year brackets so 40-49 was the only one I could re-analyse because the report calculates each part of the age group separately and then provides the weighted average.
For other age groups I could solve for the split but it was quicker to just pick 40 to 49s!
I have no explanation for the differences though.
Ok. I see.
My problem is the following: let's consider for example the age cohort 5-9 years, and let's assume that we have for this age group both the deaths per month and the person-years.
Now, if we see a peak of deaths in a certain month, we cannot be sure that this is due to an increase of mortality, if we are considering the person-years related to the 1st of January (as it could also be for example due to the fact that in that month the number of 5-9 years old increased).
For this reason I was looking for population for each age groups, but divided by month or by week.
Do you know if these data are available? I have been looking for these the last days but I did not manage to find them.
No, person years is not related to the age of each person, it is a function of the time that person spends in the vaccination cohort. So if you had 12 million people and they spent 1 month in the unvaccinated cohort, that would equate to 1 million person years. That's my understanding anyway.
Alright, I see. Thank you for the explanation
Personally I find charts are racist and bigoted. I prefer numbers in a table.
Thanks Metatron. But I don't think you can add up rates in this way. E.g. if you have 50/100k rates for vaccinated 1dose and 2dose, the combined rate (1+2dose) is still 50/100k, as you have to add numerators and denominators. Or perhaps I am missing something ?
Yes but to do it properly I need the numerators and denominators which they do not provide! Besides, there is inevitably little overlap between the two doses. It wouldn't change much. But you are correct, I probably should have taken the average rather than the sum even though it will differ from the weighted average.
Acus Mortis
Why did you accuse Mortis for?