Total probability that resistance evolves in our model. (A–C) $Ptot$ as a function of $μ$ and $δ$ in our standard model, assuming three different effective population sizes (A) $Ne=105,$ (B) $106,$ and (C) $107.$ Larger values of $Ne$ increase the probability of resistance, yet even in the scenario with $Ne=105,$ low values of $δ<10−5$ are still required if resistance from NHEJ is to be prevented. (D–F) $Ptot$ as a function of $μ$ and $δ$ for fixed $Ne=106$ in three limiting cases of our standard model: (D) scenario of a very weak drive with cleavage rate $c=0.2.$ Comparison with (B) shows that this has almost no effect on $Ptot.$ (E) Scenario with low driver costs, $sd0=sdd/2=0.01.$ While this allows for larger values of $δ,$ the dependence on $μ$ does not change much, since $PSGV$ remains largely unaffected by the driver costs (compare with Figure 2B). (F) Scenario in which resistance alleles carry (codominant) fitness costs almost as large (90%) as those of the driver, $sr0=srr/2=0.09.$ In this case, resistance is unlikely to evolve from SGV or de novo mutation. However, it is still likely to evolve by NHEJ unless $δ<10−5.$