Should we accept conjectures like Goldbach based on heuristics % probability of being true?

[Anonymous No.9015532]

You can read about all of the evidence to suggest it's true. This isn't water tight proof that it is true, as a very large counter-example could exist. I'm not suggesting we flat out "accept" conjectures without watertight proofs as "true". I meant, basically is it acceptable to have a column of "probably true w/ x arbitrary %" and "probably not true" list of conjectures that we can enumerate in some list and use the evidence we obtain to increase the % of the the likelihood.

Then from there use those "likely but not proven" conjectures to prove other theorems.

[Anonymous No.9015544]

Well Goldbach's only been shown to be true for 0% of integers, so you probably don't want to be working with probabilities

[Anonymous No.9015559]

you should only use proven probabilities

like birch/swinnerton-dyer conjecture is known to be true for something like 80% of elliptic curves

[Anonymous No.9015569]

>>9015559
Never heard of that. Not really in a great place to google for further info. Can you explain? Sounds interesting.

I completely agree that we shouldn't accept any conjecture as "true" based on evidence alone. This thread is sort of an attempt at playing devils advocate

[Anonymous No.9015584]

>>9015569
https://arxiv.org/abs/1407.1826

[Anonymous No.9015644]

>>9015584
>https://arxiv.org/abs/1407.1826
Thanks.