logoalt Hacker News

cjbgkaghyesterday at 10:28 PM1 replyview on HN

You’re not understanding the point of the hypothetical, it was to illustrate a scenario in that most people would consider evidential and point out that by definition the opposite outcome must also be evidential even if that outcome was far more likely. It’s to highlight an unintuitive aspect of probability formalized as the bayes ratio. The strength of the evidence is dependent on the priors but it exists. Anecdotal evidence is not zero evidence it is weak evidence and with care a lot of anecdotal evidence can be combined to create stronger forms of evidence. Rounding it down to zero or up to one are both incorrect.


Replies

dotancohentoday at 12:44 AM

And yet neither of the sides are strong enough to term as evidence. I have no problem with self-experimentation. But citing a few anecdotes as evidence is a dangerous area that I discourage. Somedau we will understand why some people respond so favourably to the wolverine stack and others don't. And somebody will someday correlate what those hurt by BPC-157 all have in common. But until then, these peptides cannot be recommended to others in good conscious.

But for self-use? Go right ahead.