You shouldn't believe or disbelieve anything. You can & should always know.

Something is known Beyond a reasonable doubt if the laws of cause and effect make it impossible for it not to be true

I KNOW with certainty that a 2 headed coin will come up heads if flipped.

If the coin has a heads and a tails then I dont know what it will come up as but I KNOW the probability is 50/50.

If I have 3 coins in my pocket, one with 2 heads, another with 2 tails, and another with one head and one tail and I pull one out at random and, without looking at it, flip it then I dont know the true probability for that particular coin (which could be 0 or 50 or 100 percent) but I still KNOW the a priori bayesian probability for that particular coin is 50/50.

If I continue flipping that particular coin and observing the results then the Bayesian probability for that particular coin will gradually change and approach the true probability for that particular coin.

Put simply if I keep getting heads then I will start expecting to see heads. But even after a thousand heads there is still a very small chance that I just got lucky 1000 times.

But at no point is it ever necessary to believe anything. You can and should always know