Hacker News Reader

Comment by underdeserver | Hacker News Reader

Not peer reviewed. Small numbers. Nothing to see here but FUD, move on.

This is quite a bad mindset of dismissing facts and not investigating bad outcomes because they are categorized as FUD. It's the same mindset that never changes your opinion about something if it's made up etc. Very dangerous for the situation we are in.

Not at all. A bad situation does not mean its okay or even beneficial to abandon statistics or the scientific method to scare people.

Publish when there’s data worth talking about. Keep it to yourself if it’s underbaked.

If everybody "kept it to themselves" until every i was dotted and t was crossed, we'd all be dead.

This is not about being rushed. This is about a paradigm already existing that's both better than the scientific method, and ok to use with incomplete information. Bayes. We aren't in the 1700s any more :) the scientific method is pretty cool, but it hasn't been cutting edge for a long time.

Overestimating information is not due to a lack of understanding statistics.

It’s truly awful to take Bayes’ namewhen arguing against the very notion of statistical significance.

Until peer reviewed, these are not facts, they're preliminary data that are subject to change.

Peer review is not magic truth pixie dust. If it's true after peer review, it was also true before, and if you're able to recognize it you can get the benefits of that knowledge ahead of time. And of course sometimes false results get through peer review, too...

Peer review tells me that it's significantly less likely the researchers made mistakes handling the data, analyzing it, controlling for external causes, etc.

A peer-reviewed version also may change the wording of the abstract. The vaccine may be less effective on the South African variant than it is on the common strain, but if it's down to 85% from 95% we'll still do fine once a majority of people are immunized.

That's all true and doesn't contradict what I said in the slightest. And while it does support skepticism, it does not support ignoring as-yet-unreviewed results on that basis alone.

It’s pretty much the same thing with the AstraZeneca study in South Africa that found it to be 10% effective, yet people throw that around a lot.

That the vaccines have lower efficacy for variants is pretty much to be expected and kinda documented by now. Why would the Pfizer vaccine be any different?

Not in the field, but I think it depends how far down and in how many ways the vaccine tries to cut the virus down. My guess is that the leaves can be modified faster than branches ...

Reply on news.ycombinator.com