If we're talking about one future spin of the roulette wheel, then how does knowledge enter into it? Maybe my belief that the roulette wheel is functioning correctly is an incorrect belief. Maybe the ball lands on black 75% of the time. Nevertheless the ball might land on red for the one future spin of interest.
Do you accept a distinction between knowledge and true belief? Consider generating a random bit of data that will be used in a predetermined way to construct a statement. For example, the data could be generated via a machine that records details of the timing of radioactive decay of a sample of radioactive material. If the statement is presented as being an oracle's prophecy, then it might be believed by somebody. The belief might be later confirmed to be true. Hence we have an example of true belief.
For something to be knowledge requires more than that. For example, illegal insider stock market trading is conducted when somebody has knowledge, not merely true belief.
Arguably, it's not possible to have knowledge of the outcome of one future spin of the roulette wheel. The argument is that nobody has access to information about the future. Even a shrewd and informed guess about the future that is eventually shown to be correct isn't knowledge. It's true belief.
As I understand it, one can have a true belief about a probability, and one can have knowledge of a probability. How do you get from the topic of probability to the conclusion that you are dealing with knowledge and not dealing with belief?