r > 0; number of failures until the experiment is stopped
p ∈ (0,1); success probability in each experiment
See Implementation