If it is possible to stimulate reality to this level of detail, very low. If it’s possible that the simulation can then run another instance of the simulation with no loss of fidelity and that is true for any simulation within the stack, still low but much more likely than before. If the chance of this simulation existing is higher than the odds for a universe that can sustain intelligent life, then it becomes about even odds.

The people who claim otherwise are mathematicians who forgot how reality works, as they get into an infinite spiral of higher and higher odds without any basis in reality.

The reason why it gets to even-at-best is because the simulation needs to exist in a reality at some point, and it really, really stretches the imagination that someone could build this shit. So, then you’re attaching the odds of intelligent life in a universe to the odds of then some intelligent life understanding literally every aspect of reality and being able to build said simulation (and then that repeats on every simulation).