Also known as survivorship bias.
Someone who practices a skill for 20 hours a week for a year will have had 1040 hours of practice vs 258 hours of practice for someone who practices 5 hours a week, and that linear increase in time will likely produce an exponential increase in skill as knowledge builds on itself. Assuming there's a market for the skill, the odds of success in that market will also go up exponentially. And if demand shrinks (ie economic recession) that filters out the 5-hours-a-week people.
That leaves almost zero chances of success.
Success, however we choose to define it, is in large part a result of luck. Skill, to a much lesser degree.
But if you think that successful engineers are only successful because they just magically got hired one day, and if they never learned any science, math, or got an engineering degree and put in zero or near-zero effort they would have had the same odds of been hired, that's just not a logical argument. Same goes for other professions.
We all start with the hand we're dealt by the universe, but it is possible to improve one's odds quite substantially given time and good decisions (that's arguably the definition of a "good decision").
Is Ken Thompson lucky? Probably was very lucky to have influences to point him towards electrical engineering and computer science, but with his level of skill, it's hard to imagine him not having a huge impact somewhere. Ending up at Bell Labs (let's say that's luck since he could've worked in a variety of environments) is definitely lucky, but he wouldn't have created UNIX if he didn't also have the skill. I think this goes for the majority of figures in computer science (I don't know enough about other fields to say that about them).
Do you just sit at home staring at the wall waiting for luck to happen to you? What is the point of doing anything if all results only come from blind luck?
Anecdotal : every job I've had came from relations, 1st to 3rd degree.
“The luck favors the prepared”