Is 'The probability of rolling a 3 with a dice is 1/6' a null hypothesis or an alternative hypothesis?

In theory, it is an alternative hypothesis, where to get any number on a die by rolling, it is 1 / 6.





In reality, the probability changes each time you roll, so it is a null hypothesis.|||It depends on what you want to show. If you want to show the die to be biased to have lower or higher probability of a 3 than 1/6, the assumption that it is fair will be the null hypothesis.





Generally, if you want to show that a certain theory holds, you make the inverse of the theory the null hypothesis, and show that, given the evidence, this is so unlikely that you have high confidence that your theory holds.|||It can be either.


If you want to test if the die is unbiased:


H0: prob of 3 = 1/6 --- null


H1: it is not 1/6 --- alternative





If you suspect that the die is biased and want to know that 3 has a probability of 2/3, then the following can be null and alternative hypotheses.


H0: prob of 3 = 2/3


H1: prob of 3 = 1/6