You are viewing a single comment's thread.

view the rest of the comments →

Men13 ago

If anyone wants to know where OP actually lied - I copied both links here so you can follow them and look for yourself.

TL;DR: For 2020, in addition to missing the "future" (which OP accounted for) it's missing the "near past" (because the CDC gets death notices from local morgues a few weeks after death). OP "forgot" to account for that.

The CDC's data has the correct number of "deaths compared to previous years" that OP claims to calculate and it shows MASSIVE INCREASE IN DEATHS. But OP decided to hide these numbers from you and instead wrongly calculate them himself to spread misinformation.


Like all good lies - they intentionally fudge 2 different ways of counting and comparing them.

First page - mortality per week:

https://www.cdc.gov/nchs/nvss/vsrr/covid19/index.htm

Note that there's a column called "percent of expected deaths" that does exactly what OP claims to do - compares the number of deaths to previous years. And you can see it's significantly over 100% almost every week - PEAKING AT 142% FOR SEVERAL WEEKS!

That's 40% EXTRA DEAD compared to previous years! It's HUGE!

Anyway - IF OP were honest, they'd use this data: data counted the same way in multiple years.

But OP is a liar - so they instead decided to use this page for comparison:

https://www.cdc.gov/nchs/fastats/deaths.htm

They got the total deaths in 2018, and did "math" to compare the numbers.

Why is it wrong? They are both deaths, so why can't we use both numbers?

Well, let's read the comment on the first page:

Percent of expected deaths is the number of deaths for all causes for this week in 2020 compared to the average number across the same week in 2017–2019. Previous analyses of 2015–2016 provisional data completeness have found that...

The important part:

... completeness is lower in the first few weeks following the date of death (<25%), and then increases over time such that data are generally at least 75% complete within 8 weeks of when the death occurred

Meaning real-time data is hard to get. The data takes a few weeks to get from local morgues etc. to the CDC. Meaning the data for the last few weeks is lower than the real data!!!

WE CAN SEE IT IN THE ACTUAL TABLE! Look at the last 4-5 weeks in the table, you'll see the "percent of expected deaths" is extremely low there - like 32% in the last week, 77% in the previous week etc.

Why? Because it's only partial data!

THIS IS OP'S LIE!!! He knows that the data is incomplete (real-time data always is) but pretends in his math that it is complete and compares it to complete data from 2018. Instead of using the week-by-week comparison that would show the 142% of expected death in multiple weeks.

OP IS A LIAR!

Glory_Beckons ago

You're right, the OP is either misrepresenting or misunderstanding the data.

However, you're making mistakes as well. Chiefly among them:

Not all excess death can be attributed to Covid-19

This is apparent from the data in the CDC table, because the excess percentages do not match the total Covid deaths. However, the CDC glosses over this, or at least fails to highlight it. Similar to how whoever made the slide on the OP glossed over some crucial details to make the data support his views.

I took the liberty of slapping together a table of my own, based on the same data.

The first four columns (including date) are lifted straight from the CDC table. To address your concerns about incomplete data, I removed both the first eight weeks (Covid hadn't really taken off yet) and the last eight weeks (incomplete reports). To highlight what the CDC glossed over, I added four columns of my own.

See Table Here

The additional columns are calculated as follows:

  • E = C / (D / 100)
  • F = C - E
  • G = F - B
  • H = (G / F) * 100

First thing to note is that cutting the irrelevant and incomplete weeks yields a much more representative dataset. Which shows that, during the time Covid has been active, there have been, on average, over 20% excess deaths compared to previous years. However:

This is in spite of the fact that, as we know, they attribute every case they possibly can to Covid, including motorcycle crashes and heart attacks, because doing so brings more handouts from the government. If you die and you ever tested positive for Covid, recently or not, symptomatic or not, you'll likely get labeled with the Covid code. Note their fine print; their "All Covid Deaths" column includes "presumed" cases.

Even setting all that aside, even taking the numbers at face value... 33% of excess deaths being caused by something other than Covid seem pretty significant to me. What mysterious nameless affliction could be killing thousands of Americans every week? What else, besides Covid, is so very different in 2020 compared to previous years, and could be driving so many people to their deaths?

Ah, yes. "The Lockdowns."

The great big elephant in the room that nobody wants to talk about: How many people have the locksdowns killed? We know it has to be quite a few, between missed essential healthcare, suicides out of desperation, and everything else. But nobody seems interested in keeping track of it.

Well, you can get a pretty good idea here: Over 3,500 every week, on average, in the US alone.

That's just under 100,000 in just over half a year.

For every 2 deaths blamed on Covid, the lockdowns took 1 more.

And what for? Would Covid have killed more than that, without them? How do we know? Even if we assume, would it have been significantly more than that? Enough to warrant everything we've given up? Everything we've been forced to give up? Enough to justify taking it upon ourselves, to impose measures that would kill so many, instead of letting nature run its course, the way it always has? And what of all the suffering caused that hasn't (yet) resulted in death?

Was it worth it?

26565698? ago

Would be interested in demographics of the dead. I would wager the covid deaths are largely old, fat or other pre-existing conditions and the non covid deaths have a comparably larger percentage of younger and healthier people.