From: Bruce Johnson
>
> Organisms pass on their genome whole, not piecemeal. A mutation that
> severely inhibits reproductive success will likely rapidly disappear. A
> mutation that enormously improves reproductive success will probably
> spread quickly. 99.999999% of mutations are neither.

If we assume a 1% difference in reproductive success, it should take 70 generations for the prevalence of a given mutation to drop by 50%, and 1,400 generations for it to drop under 1/1,000,000. It's even worse for recessives.