The observation that many people claim that god exists, is positive evidence that god exists, i.e. P(God exists|Many people claim god exists, other background information) > P(God exists|Not many people claim god exists, other background information).
[pollid:738]
Given all other information, what is the likelihood ratio of the observation that many people claim that god exists, with respect to the hypothesis that god exists? I.e. what is the ratio: P(Many people claim god exists|God exists, other observations) / P(Many people claim god exists|God does not exist, other observations) ?
[pollid:739]
Edit: The second poll might be restricting to numbers between zero and one, so I’m adding this one for bigger values. Please answer this third poll question as you would answer the second, but only if there was a problem answering the second one, to try to avoid double-counting.
The observation that many people claim that Vishnu exists is evidence against the proposition that there is no god but Allah. But the observation that many people claim that there is no god but Allah is evidence against the proposition that Vishnu exists.
We could try to factor this as “a divine being exists” vs. “that divine being has properties X, Y, and Z”, but different traditions don’t even agree on what a divine being is. Mormonism and some Dharmic traditions, for instance, consider that humans could be reinstanced as divine beings after death, whereas most Abrahamic traditions do not.
I think you messed up the second poll item; don’t Bayes factors run from 0 to infinity? Not 0-1.
Anyway, I strongly believe that people claiming there is a god is indeed a great deal of evidence for it, in the same way that being told ‘China’ exists by a lot of people is a lot of evidence that such a place exists. Of course, afterwards, when you start digging into their reasons for believing in god, all the other gods they didn’t mention to you but that still other people believe in, etc, then you start to undo it.
I used the likelihood poll type I found on this page hoping (and I was aware it was hope and did worry it might be broken!) that it would work properly. I did see it says ‘probability’, but does it actually insist on a zero to one value, or are you guessing based on that it says ‘probability’? (I actually found it difficult to answer my own question so only answered the first question so far so have not tested the second poll.) I’ll look at making a third poll of a different type in case the second eats positive evidence.
I see. I wonder if the aliases should just be deleted from the help page? There’s no real need to have a bunch of aliases for a poll-command, and as you demonstrate, it’s possible to read the page and not realize they’re aliases rather than differently-typed polls.
The observation that many people claim that god exists, is positive evidence that god exists, i.e. P(God exists|Many people claim god exists, other background information) > P(God exists|Not many people claim god exists, other background information).
[pollid:738]
Given all other information, what is the likelihood ratio of the observation that many people claim that god exists, with respect to the hypothesis that god exists? I.e. what is the ratio: P(Many people claim god exists|God exists, other observations) / P(Many people claim god exists|God does not exist, other observations) ?
[pollid:739]
Edit: The second poll might be restricting to numbers between zero and one, so I’m adding this one for bigger values. Please answer this third poll question as you would answer the second, but only if there was a problem answering the second one, to try to avoid double-counting.
[pollid:740]
The observation that many people claim that Vishnu exists is evidence against the proposition that there is no god but Allah. But the observation that many people claim that there is no god but Allah is evidence against the proposition that Vishnu exists.
We could try to factor this as “a divine being exists” vs. “that divine being has properties X, Y, and Z”, but different traditions don’t even agree on what a divine being is. Mormonism and some Dharmic traditions, for instance, consider that humans could be reinstanced as divine beings after death, whereas most Abrahamic traditions do not.
I think you messed up the second poll item; don’t Bayes factors run from 0 to infinity? Not 0-1.
Anyway, I strongly believe that people claiming there is a god is indeed a great deal of evidence for it, in the same way that being told ‘China’ exists by a lot of people is a lot of evidence that such a place exists. Of course, afterwards, when you start digging into their reasons for believing in god, all the other gods they didn’t mention to you but that still other people believe in, etc, then you start to undo it.
I used the likelihood poll type I found on this page hoping (and I was aware it was hope and did worry it might be broken!) that it would work properly. I did see it says ‘probability’, but does it actually insist on a zero to one value, or are you guessing based on that it says ‘probability’? (I actually found it difficult to answer my own question so only answered the first question so far so have not tested the second poll.) I’ll look at making a third poll of a different type in case the second eats positive evidence.
Well, if you try entering, say, ‘2’ and hitting Enter, it emits a little red error message: “Probability must be between 0 and 1”.
Ah! I only skimmed the page I linked and confused likelihoods (which are probabilities) with likelihood ratios (which are...ratios).
I see. I wonder if the aliases should just be deleted from the help page? There’s no real need to have a bunch of aliases for a poll-command, and as you demonstrate, it’s possible to read the page and not realize they’re aliases rather than differently-typed polls.