Live data from Hacker News

In Praise of Observational Evidence

asteriskmag.com

51–56 of 56 posts

Re: In Praise of Observational Evidence

#51
post #50

Earlier quoted context omitted.

Uhh... the question was "can the market detect efficacy." My answer is: Generally no, and it also can't detect negative effects either. That's answering the question. > this only require information regarding if the thing works or not for the people who use it as long as that information is not censored. Can "the market" currently detect whether there is complex life in the Messier 54 star cluster? No. Is it because…

>because that information is indistinguishable from noise. Not if the signal is strong. Whether a thing works or not, is a pretty strong signal. Your equating it to bad effects of smoking is flawed and is not comparable. You other arugments are also pretty bizzare, but not going into that for sake of focus.

Well, with an argument like that!

> if the signal is strong, it will be visible

> your specific example of it taking decades of massive deployment and close scrutiny to detect an effect significantly stronger than most drugs is ummm... flawed and not comparable for no reason I can articulate

okay buddy

That's okay, you're someone who thinks the FDA is in cahoots with insurers and drug companies to approve ineffective or unsafe drugs at scale lmao. And here you are arguing that the market would actually naturally detect exactly that.

Re: In Praise of Observational Evidence

#52
post #50

Earlier quoted context omitted.

>because that information is indistinguishable from noise. Not if the signal is strong. Whether a thing works or not, is a pretty strong signal. Your equating it to bad effects of smoking is flawed and is not comparable. You other arugments are also pretty bizzare, but not going into that for sake of focus.

Well, with an argument like that! > if the signal is strong, it will be visible > your specific example of it taking decades of massive deployment and close scrutiny to detect an effect significantly stronger than most drugs is ummm... flawed and not comparable for no reason I can articulate okay buddy That's okay, you're someone who thinks the FDA is in cahoots with insurers and drug companies to approve ineffective…

> to detect an effect significantly stronger than most drugs..

What the heck are you smokin? Did paracetamol have a weaker affect on temperature than smoking causing cancer?

Re: In Praise of Observational Evidence

#53
post #52

Earlier quoted context omitted.

Well, with an argument like that! > if the signal is strong, it will be visible > your specific example of it taking decades of massive deployment and close scrutiny to detect an effect significantly stronger than most drugs is ummm... flawed and not comparable for no reason I can articulate okay buddy That's okay, you're someone who thinks the FDA is in cahoots with insurers and drug companies to approve ineffective…

> to detect an effect significantly stronger than most drugs.. What the heck are you smokin? Did paracetamol have a weaker affect on temperature than smoking causing cancer?

That is correct. Smoking has a stronger effect on cancer rates than paracetamol does on fever (which itself is an extremely strong effect that can be proven in RCTs of just a few dozen people - far stronger than most drugs on the market).

Thank you for demonstrating my point.

Re: In Praise of Observational Evidence

#54
post #52

Earlier quoted context omitted.

> to detect an effect significantly stronger than most drugs.. What the heck are you smokin? Did paracetamol have a weaker affect on temperature than smoking causing cancer?

That is correct. Smoking has a stronger effect on cancer rates than paracetamol does on fever (which itself is an extremely strong effect that can be proven in RCTs of just a few dozen people - far stronger than most drugs on the market). Thank you for demonstrating my point.

And thank you for demonstrating mine..
Post reply on HN