3 ms·
Statistical models work great for well defined problem spaces like the probability of rolling a die in a particular way. You can reason in your own mind the pro
by passivate 5y ago
Statistical models work great for well defined problem spaces like the probability of rolling a die in a particular way. You can reason in your own mind the probability of aliens or extinction events or lab leaks or anything really, but it doesn't get you very far (IMO) in proving anything.
To paraphrase Wernher von Braun - "Hard data is worth a thousand expert opinions." :)
- lovemenot 5y agoYou seem to be conflating subjective and opinion on the one hand with statistical on the other. Actually, Bayes statistics works great in poorly defined problem spaces where we can update our priors as new information becomes available. Just like in the issue under discussion. Your example of rolling dice is Frequentist, not Bayesian. We wouldn't use Frequentist stats in this domain, for the reasons you mention.
- passivate 5y ago>Actually, Bayes statistics works great in poorly defined problem spaces where we can update our priors as new information becomes available. Just like in the issue under discussion. Can you give me a few comparable scenarios where it worked great?
- lovemenot 5y agoOne of the techniques is Bayesian search https://www.inf.ed.ac.uk/teaching/courses/dmr/Notes2019/7-Bayes-case.pdf https://www.inf.ed.ac.uk/teaching/courses/dmr/Notes2019/7-Ba... I don't know whether Bayesian search is currently being used to search for the unknown reservoir species from which SARS-COV-19 jumped to infect humans (assuming a natural cause). Under this approach, the longer the search goes on, the more we may lessen our confidence in the prior assumption that it was a natural infection.