"Luck" at the Olympics is treated like a dirty word. Luck just means "uncertainty" or "volatility." In most events, the extraordinary level of skill among the top competitors is so close that luck can make the difference.
How could natural selection operate so that "the good ones spread in the population" if there were no positive variants in the first place? That is why I am confounded by Jerry Coyne's comment that he can explain natural genetic engineering by "garden variety natural selection."