Whats the answer b?

No, 2/3. Remember, case one has already been discounted. In two of the three cases, we get a second gold ball.
The 4th case is just as likely as 2 and 3.
But we can't tell the difference between two and three.

yes, so?

Could probably get the noise to an acceptable level after just 50 tbh.

you cannot just discount the initial case, you picked before you knew about the outcome

Correct

So in 180 tests you take each ball an equal amount of times, 30

In 60 tests total you get 2x G
In 30 tests you get 1x g
In 90 tests you get Silver and the problem is a non starter.

In any scenario where you select a box and the order of the balls is not known or predetermined the outcome is that the GG box is held twice as often

See:

10,000 is plenty if you were right. The numbers would be close to 50%, and they simply arent. Experiments are worth more than theories

So that means it's 2/3 is the correct answer, not 1/2. Objective reality is pointing towards the actual mathematical answer.

Sure they can. Couldnt we flip a coin to determine the probability of it landing on tails? Cant we roll dice to figure out which total is the most likely?

But case one doesn't satisfy the condition set "Given your first draw is a golden ball"
So case 1 must be removed as a possibility.
In reality, there are six cases total, but cases 5 and 6 are from the double silver box, and are equally ignored

I already did this test, 100 times, and I only got gold each time, therefore the chance of gold is 100%