Probabilistic vacuum: dirt is removed with probability .
What's the probability a square would be still be dirty after 10 sucks, given that it begins dirty?
We are assuming that dirtiness now next conditionally dependent only on dirtiness now: Markov property.
How solve?