85,000 excess deaths in 2020 over a rolling five year average, the highest number of excess deaths since WW2. Many of those infections could have been passed on by asymptomatic younger people.
Anyone remember the opening ceremony of the 2012 Olympics? Now picture all the crowd in the stadium dead. That's what 85,000 looks like.
Lies, damn lies, and statistics
Case Study A
news.sky.com/story/covid-19-how-mortality-rates-in...5
1. There are no quotes in this article, only the journalists interpretation
2. Headline changes "was almost without parallel" to "unparalleled loss of life. 3rd degree Chinese whispers
3. First graph is cases - irrelevant, but sets the bias
4. First deaths graph which shows covid top is not normalised for population growth, so is meaningless.
5. Second graph a zoom in of the first, to hammer it in. Directly compares Covid to WWI. Because soldiers with 60 years ahead of them deliberately killed are the same as someone beyond average life expectancy dying of pneumonia whose weakened body also picked up Covid. Not normalised for population.
6. Editor taps him on the shoulder about the population problem. Shows excess deaths are up to 10.2 per thousand from 8.9 last year. Shut the economy and hide indoors, the plague!! We are now back to death levels last seen in 2003 (think there was a heatwave then, I guess we should have all been locked up then too.
7. Gives the reason for the gap from 2003 to now was medical improvements. That's just made up, metabolic health has gotten far worse in the interim.
8. So a proper population adjusted graph (but your biases are set, right? In work, all the managers just want to know "the number", i.e. how big I think a field is. So rather than make them wait to the end, I put it on the first slide. Once I've spend the next 50 slides justifying it, they are all nodding along like donkeys, as if they thought of it themselves). Problem here is, 'deaths' are not defined. Covid or lockdown deaths? Also, why the trend increase since 2010 or so, obesity?
9. The next trick on this graph is the scale (used to hide all manner of crimes, which is why I asked someone here for it yesterday). Its not an absolute number, but a rates. Its deaths as a % of the previous 5 year average. So if you have a good five years, and something happens like the March pandemic, you get a spike. This is why the graph is spiky, and not very useful. he relates this back to 1940 "My God this is terrible! We are in a once in an 80 year event". No, 1935-1940 were pretty good, until Hitler started bombing England. What an utterly ridiculous comparison to make. The 1951 flu epidemic also had a quiet 5 year period prior, causing the spike.
10. Same goes for the age adjusted graph up next, showing rates relative to the prior 5 years. Not useful. It doesn't look to be normalised by population either. Here he brings in 'without parallel', when even this graph shows spikes over 10% have happened 7 times during this time period, with the current one having by far the best 5 year prior period.
11. Just for the heck of it, he changes the rate calculation to change in mortality vs the previous year, rather than the 5 year average. Even more statistically stupid. Here, it puts Covid 3rd in the all time (selective) list:
He doesn't mention that 2019 was a year of low mortality (the dry tinder), which is responsible for this. His narrative implies that Covid is worse than Cholera, WWI or WWII.
Then comes a bunch of caveats to cover his backside legally I'm sure, and finished by implying they would have been worse without lockdowns.
If this amateur hour is the best they can come up with, they are running out of ammo.
By all means continue to believe the government, like when they saved us from Saddam and the weapons of mass destruction that never were.
|