I'm doing the same analysis here in NZ. Your method similar to that used by the HMD, known as the Shrt Term Mortality Fluctuation. They have a shinyapp that does the graphics. The 5-year trend of weekly deaths suffers from 3-major flaws. I used it myself for a court case against the NZ MOH, where I also pointed the flaws. I now used death rate data, also available on the STMF app, as it takes out population anomalies. The NZ government provides death rates for age bands from 1991 onward. The biggest casualty in 2022-3 was the over 90 cohort, although I have noticed the 60-69 cohort is doing badly in 2023. A lot of the deaths are occurring the warmer months, which is unusual. The older cohorts did better in the winter if 2023, because the vulnerable had died off in 2022.
Glad to see there is at least one other fan of the app out there. 😀
I used to crunch the numbers myself myself, much like ManDownUnder has, it’s a long tedious process and I appreciate what he’s done in his post because I know how long something like his post takes to put together and present, especially doing it as neatly and clearly as he has.
I’ve become a big fan of the app. since I found it. You can also download the raw data if you want to do your own customisation, the app is open source, so you could even build on it if wanted to create your own “custom methods”. If you register you also get access to more raw data, they have actually paid the ABS for data broken down into single year cohorts which is not freely available to the public so could do an even more granular analysis if you want to crunch the numbers yourself.
I could spend heaps time analyzing your post. Firstly, the week specific trend data is prone to too short a range,. This creates a positive slope in the regression because of the bad flu seasons in 2017 and 2019. I've plotted the trends for the summer weeks and many slope down, while the winter weeks slope up. Because they extrapolate these regression lines into the pandemic years, the slope leads to higher baseline deaths. This effect is removed in the week specific average method. Unfortunately, in Australia, they only have weekly data back to 2015, which is very poor, in my opinion, given NZ has data back to 2011. My analysis of NZ data also shows the 85+ were hit hard in 2022 and 2023.
May I suggest you also ook at a paper I co-authired with Aussie, Clare Paine
I agree with what you’re saying. I’ve also plotted all the weekly trends since writing the post to see what they look individually and what you say is true. I was just interested in what would pop out using that particular setting as the baseline because I would never have gone to the trouble of calculating it myself because (if I’d come up with the method myself) I would have thought “nah it’s going to be too noisy to show anything of value so don’t bother.” But, I was surprised the signal just sort of popped up through the noise in exactly the right place to show a temporal association with the beginning of the rollout. It could just be a coincidence, but there seems to be way too many coincidences and they all point to the same thing.
I’ve looked at the excess death from the ABS data many times using all sorts of methods to calculate different baselines.
In general it doesn’t really matter what baseline you use all of them are unfavorable to the mainstream narrative, they all show a large increase in excess deaths no matter how you look at it (some more than others).
If you’re looking for data going further back. There is much more data going back much further for Australia (and other countries) on the main HMD site. It’s even broken down into 1 year age bands so you’d have to “roll your own” datasets to get them into the same age bands used in the STMF app which would be tedious.
Just had a read of your paper it’s well written and a good summary of the situation. It’s pretty similar to what I found on my own before I realized that many others had seen the same thing. The Government’s control of narrative that stopped this sort of information getting any traction has really shocked me. I’d always thought once people began to see this sort of data for themselves they’d wake up and call “bullshit” on the official narrative and it would all be over and the perpetrators would be held to account. Sadly, I don’t think anything is really going to change and no one will be held accountable. They are already pretty much saying there is no point dwelling on the past, mistakes were made and we’ll do better next time.
Glad to see someone else who is dredging the data. The NZ government was using the OWID method of Karlinsky and Kobac, which showed no excess Deaths to the end of 2022. I've dissected that in another paper that was used in a high court case. In any case, I prefer the STMF platform, even though it lags by several weeks.
I'm doing the same analysis here in NZ. Your method similar to that used by the HMD, known as the Shrt Term Mortality Fluctuation. They have a shinyapp that does the graphics. The 5-year trend of weekly deaths suffers from 3-major flaws. I used it myself for a court case against the NZ MOH, where I also pointed the flaws. I now used death rate data, also available on the STMF app, as it takes out population anomalies. The NZ government provides death rates for age bands from 1991 onward. The biggest casualty in 2022-3 was the over 90 cohort, although I have noticed the 60-69 cohort is doing badly in 2023. A lot of the deaths are occurring the warmer months, which is unusual. The older cohorts did better in the winter if 2023, because the vulnerable had died off in 2022.
👍 I’ve also been “pimping” this app for awhile.
https://krap.substack.com/p/vax-signal
Glad to see there is at least one other fan of the app out there. 😀
I used to crunch the numbers myself myself, much like ManDownUnder has, it’s a long tedious process and I appreciate what he’s done in his post because I know how long something like his post takes to put together and present, especially doing it as neatly and clearly as he has.
I’ve become a big fan of the app. since I found it. You can also download the raw data if you want to do your own customisation, the app is open source, so you could even build on it if wanted to create your own “custom methods”. If you register you also get access to more raw data, they have actually paid the ABS for data broken down into single year cohorts which is not freely available to the public so could do an even more granular analysis if you want to crunch the numbers yourself.
: End of sales pitch.
I could spend heaps time analyzing your post. Firstly, the week specific trend data is prone to too short a range,. This creates a positive slope in the regression because of the bad flu seasons in 2017 and 2019. I've plotted the trends for the summer weeks and many slope down, while the winter weeks slope up. Because they extrapolate these regression lines into the pandemic years, the slope leads to higher baseline deaths. This effect is removed in the week specific average method. Unfortunately, in Australia, they only have weekly data back to 2015, which is very poor, in my opinion, given NZ has data back to 2011. My analysis of NZ data also shows the 85+ were hit hard in 2022 and 2023.
May I suggest you also ook at a paper I co-authired with Aussie, Clare Paine
www.excessdeathstats.com
I agree with what you’re saying. I’ve also plotted all the weekly trends since writing the post to see what they look individually and what you say is true. I was just interested in what would pop out using that particular setting as the baseline because I would never have gone to the trouble of calculating it myself because (if I’d come up with the method myself) I would have thought “nah it’s going to be too noisy to show anything of value so don’t bother.” But, I was surprised the signal just sort of popped up through the noise in exactly the right place to show a temporal association with the beginning of the rollout. It could just be a coincidence, but there seems to be way too many coincidences and they all point to the same thing.
I’ve looked at the excess death from the ABS data many times using all sorts of methods to calculate different baselines.
Sometimes I feel like this guy. 😀
https://youtu.be/FDKDdVilQT4?si=-dDzOz7f90s9sC_y
In general it doesn’t really matter what baseline you use all of them are unfavorable to the mainstream narrative, they all show a large increase in excess deaths no matter how you look at it (some more than others).
If you’re looking for data going further back. There is much more data going back much further for Australia (and other countries) on the main HMD site. It’s even broken down into 1 year age bands so you’d have to “roll your own” datasets to get them into the same age bands used in the STMF app which would be tedious.
Just had a read of your paper it’s well written and a good summary of the situation. It’s pretty similar to what I found on my own before I realized that many others had seen the same thing. The Government’s control of narrative that stopped this sort of information getting any traction has really shocked me. I’d always thought once people began to see this sort of data for themselves they’d wake up and call “bullshit” on the official narrative and it would all be over and the perpetrators would be held to account. Sadly, I don’t think anything is really going to change and no one will be held accountable. They are already pretty much saying there is no point dwelling on the past, mistakes were made and we’ll do better next time.
Rinse and repeat.
Glad to see someone else who is dredging the data. The NZ government was using the OWID method of Karlinsky and Kobac, which showed no excess Deaths to the end of 2022. I've dissected that in another paper that was used in a high court case. In any case, I prefer the STMF platform, even though it lags by several weeks.
Thank you that’s a massive amount of work 🙏