Monday, April 13, 2020 by Mike Adams
Editing note: I’ve edited this article after a friend pointed out to me that David Knight recently had a heart attack, which is true, and I don’t want to lend any stress to him, as he’s done excellent work on many issues. So I’ve toned down the article from its original tone which was more harsh. In any case, I respect David’s work on many fronts, but there’s really no match when it comes to who has the better mathematical analysis, as I explain here. I wish David well and don’t plan to comment on this further.
InfoWars host David Knight has now joined the chorus of pro-Trump voices who have now “pwned” themselves with horrible displays of bad math. It’s frustrating, actually, given that David is so good on elections analysis and other geopolitical issues. (I admire his work in other areas, but he’s flatly wrong about this subject as explained below…)
On his April 13th InfoWars broadcast, David Knight incorrectly claimed that the CDC released data claiming a “10% fatality rate” from the seasonal flu. This is false, and it stems from a misinterpretation of basic numbers. The same narrative was first attempted by The Gateway Pundit and then spread to other pro-Trump pundits who couldn’t see the obvious mathematical error (which I explain below).
As the CDC explains right on this influenza surveillance page, “CDC estimates that so far this season there have been at least 39 million flu illnesses, 410,000 hospitalizations and 24,000 deaths from flu.”
In other words, the “24,000 deaths” is a CDC estimate, not an actual number of confirmed, tested samples. And since the CDC is detailing estimated deaths vs. estimated infections, the fraction is structured as:
projected numbers vs. projected numbers
…which comes to an estimated flu mortality rate of 24,000 / 39,000,000 which is less than 0.1%. The seasonal flu, in other words, kills fewer than 1 in 1,000 of those it infects.
But David Knight uses the wrong numbers for his calculation, then claims the CDC announced a regular flu death rate of 10%. They way he arrives at that critical mathematical error and thereby mis-educates the InfoWars audience is by taking the actual number of samples the CDC tested positive for the flu — 247,785 — and running that as the denominator against the estimated number of flu deaths of 24,000 as the numerator.
He’s taking: Estimated deaths / Actual specimens tested, in other words. That’s not the way you calculate anything. He’s using the wrong numbers.
Naturally, using that incorrect structure causes the answer to be at least two orders of magnitude too high. Since 24,000 / 247,785 comes to about 10%, David Knight claims the seasonal flu kills 10% of the people it infects. That his error in this analysis.
Interestingly, if David Knight’s math were correct, then 3.9 million Americans would be dying every year from the seasonal flu (using the numbers from above). How many people actually died from the seasonal flu last year? Just 34,000 according to CDC estimates for the 2018-2019 influenza season (source here).
So David Knight is off by a factor of around 100. If David Knight were correct, then every time you get the flu, you would have a 10% chance of dying. And so would everyone else you know who gets the flu. Are 10% of those people dying every year? Not hardly. David Knight is wrong on this issue, even if he’s insightful on other issues.
After this gross error in the math, David Knight goes on to “pwn” himself by saying, live on air, “Now, you don’t have to know anything about math to understand that you don’t compare estimated numbers to actual numbers.” That’s exactly what he’s doing, of course, calculating a regular flu fatality rate by comparing estimated numbers to actual numbers.
Literally, he comparing estimated flu deaths of 24,000 to actual flu samples collected of 247,785. He’s doing exactly what he claims is “rigging the numbers.”
Here’s the segment:
Here’s the actual transcript of the segment of David attempting to explain his bad math, from his April 13th broadcast:
The CDC estimated up the number of flu cases because just on the cases that they had tested, according to their own data that they released two weeks ago, they said well 10% fatality rate, well we know that 10% of the people that get the flu are not dying, so we’re going to estimate that up. And so they estimated up the number of cases, which they didn’t, it was just an estimate, they didn’t have any tests for that, they just said we’ll just estimate the number of cases up by a factor of 100. That made their case fatality rate 0.1%. But when they compared it to the coronavirus, they went with actual cases and actual deaths. Now, you don’t have to know anything about math to understand that you don’t compare estimated numbers to actual numbers. You don’t compare estimated flu deaths to actual covid-19 deaths, that’s not honest. That’s rigging the numbers.
In essence, what David Knight is claiming is:
So yes, David, you really should issue a retraction on your segment, since you really did do the math incorrectly. I know you care about accuracy, so it’s important to get this right.
As you can hilariously see from the video above, David then says to me on his broadcast, “But if you can’t understand the difference between estimated and actual, then you aren’t a scientist.” He’s actually describing himself, of course, because he doesn’t understand the difference between estimated and actual.
He can’t do simple division. As the CDC clearly releases in their own estimated numbers on this CDC.gov web page:
Estimated Influenza Disease Burden, by Season — United States (for the 2018-19 influenza season)
Symptomatic Illnesses: 35,520,883
Medical Visits: 16,520,350
Do the math. 34,157 deaths / 35,520,883 symptomatic illnesses is: 0.096% case fatality rate, which is less than 1 in 1,000. That’s estimated (numerator) vs. estimated (denominator).
With over 1,500 deaths just today, covid-19 is currently the No. 1 cause of death in America on a day-to-day basis. But David Knight thinks the No. 1 cause of death is the seasonal flu. If 3.5 million people were really dying from the flu every year in America as he claims, there should be 9,589 flu deaths just today.
I don’t see 9,589 dead Americans today, not even from all causes of death combined.
My suggestion is that the pro-Trump indy media hosts should stop claiming this is a “hoax” and start taking this seriously, or Trump will surely lose in November as the true costs of this very real epidemic become impossible to ignore.