/sci/ - Science & Math

>>9624366
P(Success|True) = P(True|Success) P(Success)/P(True)

If P(True|Success) > P(True) then P(Success|True) > P(Success)

So the answer to your questions is yes.

>>9624366
Truth implies it's success. If the success of any theory implied it's truth, there would be no need to constantly test theories. You can find a counter example to show this
>>9624375
>Bayesianism
fuck off you complete ignoramus

>>9624366
>>9624380
To make this point more clear, OP, a theory can describe many phenomena (and thus be success) but once a contradiction is found, the theory cannot be true.

>>9624386
How do you prove a contradiction?

>>9624375
>P(Success|True) = P(True|Success) P(Success)/P(True)
>If P(True|Success) > P(True) then P(Success|True) > P(Success)
This is not well-defined.

>>9624387
If what there is conflict between what happens in reality and what a theory tells us should happen, there is a contradiction

>>9624417
At what point does the theoretical/measured difference become a conflict? What we measure from reality always has limitations based on the experimental design and technology used to measure

>>9624423
>At what point does the theoretical/measured difference become a conflict?
If the measurements don't fall within calculated uncertainty, I think. A physicist would be able to answer this question better than I can; I'm just a mathematician

>>9624366
What's your definition of success?
>Literally arguing semantics

Bump

>>9624366
There are statistical confidence limits.
"The odds of getting these results by random chance are less then such-and-such."
For a crucial discovery, like the Higgs Boson, they didn't make an announcement until they'd collected enough data to reach the golden "six sigma."

The more data you have which agrees, the more confidence you can have that the theory is "right".
But you can never be totally sure.
Newtonian gravitation passed all tests for several centuries -- except for one tiny irregularity in the orbit of Mercury.
Several "explanations" were offered -- from a subtle flattening of the Sun to an undiscovered planet inside the orbit of Mercury -- to "explain away" the discrepancy.
Ultimately, Einstein came up with an entirely new explanation of gravity. The concepts and equations are quite different, but they agreed with all the observations which had "confirmed" Newton.

So, in a sense, ultimate proof can never be attained. Not on the real world. Absolute certainly can be achieved in areas where WE'VE defined the rules. No one is ever going to prove square root of two is a rational number.

The trouble is, laypeople use "theory" in a different sense. They don't understand that Evolution, for example, is as solidly established as gravity. No one denies the existence of gravity, because it doesn't threaten their dogma.

>>9624366
>Or the truth implies it's success?
That must be true.
A true theory has to also be successful, if "success" is defined as the theories predictions are the same as the outcomes of correctly posed experiments.

>Does the success of a theory implies it's truth?
Absolutely not, I flat earth theory can "explain" many things, that doesn't make it true.