Hacker News Reader

Comment by thehappypm | Hacker News Reader

Same take for me as well. Screams of cherry picking. The stats in favor of vaccines take a sentence to explain (like “of the 1000 people with the vaccine 2 died, vs the 1000 people with placebo, 187 died”). This is a long term rambling wall of text.

Another part is how individual doctors mention individual patients who develop problems after the vaccine. Statistically, that is guaranteed to happen. There is a guarantee that someone will have their first heart attack today. The overlap between vaccine doses and heart problems is statistically guaranteed but does not imply causation.

View on HN

> The stats in favor of vaccines take a sentence to explain (like “of the 1000 people with the vaccine 2 died, vs the 1000 people with placebo, 187 died”)

Are those real numbers?

At least according to table S26. Death summary in Blinded Phase (https://www.nejm.org/doi/suppl/10.1056/NEJMoa2113017/suppl_f...) it should be “of the 15184 people with the vaccine 16 died, vs the 15162 people with placebo, 16 died”. Unless I'm mistaken.

Read the results in the paper:

https://www.nejm.org/doi/suppl/10.1056/NEJMoa2113017/suppl_f...

Group A and Group B had the same amount of deaths.

Group B also had more adverse reactions, some serious and some minor.

Given these details in isolation, I would want to be in Group A. But then you look at what that Group A is, and it's actually the control group.

Some statistics about the table:

There were 30,000 participants, and in both test and placebo, there were 16 deaths. The test arm had 3 covid deaths, and the control arm had 1, but it was an HIV patient who was not fully vaccinated so they are arguing it’s 0 deaths for the fully vaccinated cohort. So, the numbers we’re looking at here are pretty small.

So what can we learn from this? Basically, if you remove all the covid deaths, it’s 13 deaths in test and 15 in placebo. These are statistically equivalent. In both test and control you’re equally likely to die of something other than covid in that window.

Now onto the covid deaths. There were 3 in placebo and 0 in control. These are horribly small numbers. More data would be useful.

However.. you can do the stats anyway. The placebo group suggests that the death rate is .02%, with some error. Having 0.00% death rate in the test group is unlikely, but within that error.

So this data is too weak to draw much of a conclusion.

But why would you go look at the results of a year old study with 30,000 participants when there is recent data covering many more people where protection from death among the vaccinated is blatantly obvious?

That's the trick, focus on the evidence that paints a particular picture instead of actually evaluating the available evidence.

Firstly, as others pointed out, your numbers are wrong so your one sentence explanation is just misinformation. The number of deaths in the trial arms were about the same, actually a bit higher in the vaccine arm. The official explanation for this is that COVID is not deadly enough to show up in a trial population of 'only' ~60-70,000 participants especially because the trials failed to recruit many over 80 year olds, who are by far the most at risk.

Secondly, the article is broken down into sections with proper titles, and many links to evidence. It's not a "wall of text" and it's not "rambling", it's well structured.

Thirdly, the doctors and nurses in question are saying explicitly that they found the sudden prevalence of these events surprising and concerning. They are front line workers, they aren't sitting in an office reading Excel for Dummies, they're saying they noticed a qualitative change in the situation.

Finally, I only got about half way down but I actually found statements that were underplaying the extent of the problem, like this one:

"the English data, though difficult to interpret, contains some worrying signals that all-cause mortality could be higher in the vaccinated, especially in those under 60 years of age"

It doesn't contain "worrying signals" that all cause mortality "could" be higher. It shows outright that the all-cause death rate in the vaccinated has been much higher than in the unvaccinated in 2021. You can find a graph of it here:

https://bartram.substack.com/p/the-impact-of-time-on-non-cov...

And that article links to the original ONS report, if you want to read that.

that comment is an excellent example of why these articles get flagged rather than debated. They become an incessant Sisyphean slog through every minor detail that has between twisted or cherry picked in ways that really seems intentional on the part of the pro-death anti vax movement in order to prove their narrative at any cost.

210 million America are vaccine as of today. Are you trying to say the number I mentioned in my comment was wrong because I said “few hundred million” instead of “a couple hundred million plus change” when even if it were just “one” hundred million my point would be valid? Quibbling over whether it should be a couple or few when either of them support the point is 100th level nitpicking on the unimportant. Of course the point I made has been ignored because it was valid.

> show up in a trial population of 'only' ~60-70,000 participants

Why are you looking at the number of participants? It’s not relevant to your point there. The number that had covid is needed to answer the question you’re posing.

"Are you trying to say the number I mentioned in my comment was wrong because I said “few hundred million” instead of “a couple hundred million plus change”"

No. You claimed "of the 1000 people with the vaccine 2 died, vs the 1000 people with placebo, 187 died" which is just factually wrong. The trials didn't show anything even close to this disparity.

"Why are you looking at the number of participants? It’s not relevant to your point there. The number that had covid is needed to answer the question you’re posing."

Trial size determines ability to detect small but real effects, which is why the larger number of deaths in the vaccine vs placebo arm was ignored.

Oh oops wrong level of comment. You weren’t saying that my numbers were wrong, my bad. But the other points still stand. Those numbers the person quoted do appear off.

So you agree with my points about the few hundred million getting vaccinated in a highly anxious climate with tons of misinformation about side effects is essentially guaranteed to have high levels of reported side effects simply due to psychology. That alone means nearly all of the articles points fall apart so quibbling with sub comments in an effort to salvage the article is pointless.

Also the article is a long mass of poor quality text. Having section dividers doesn’t make bad references and argumentative fallacies good logic.

For the question you posed about Covid and the excuse giving, yes the number of Covid patients matters.

The greatest point of failure in these studies is the quality of data. If it isn't ensured to be accurate any further analysis is useless. And this is the core point the article tries to make.

Not saying it is correct or not, but you did not respond to the criticism.

Numbers are cherry picked on both sides. Read the results and draw your own conclusions!

Reply on news.ycombinator.com