I got my first foot in the door via an internship on a referral from a friend who had interned there the year before. I had no formal credentials (just taking CS classes on the side of an Econ degree in college), but I had spent all of my math/CS classes studying with this kid. He knew I was smart and put his own ass on the line to get me an interview.
You can be successful purely by reading books in a room, writing code alone, and building some brilliant startup. (E.g. look at the CD Baby guy, what's his name). But it's SUPER unlikely. You're way better off being a social human being, building relationships/trust, and paying it back/forward to folks that do right by you when you know you deserve it.
A lot of the SV narratives circle around meritocracy (or other ways of 'hacking the signalling game' even. And when you say "it's SUPER unlikely", I think you're 100% correct. And so challenging the meritocracy myth is important.
Anecdata, but: I left academic hard science because I realized it was too difficult to get ahead; moved to the bay since I can code and it was a way to create a runway and find myself, and pitch companies. Got a contract coding gig, because a professor on the board of my roommate's company was impressed that I had independently implemented his prototype as a personal side project. Applied to 10+ startups for "real programming jobs", including through triplebyte, rejected by all, finally got hired by a roommate's friend (on reputation only) whose company needed asses in seats for an investor meeting.
Meanwhile, I'd been pitching 3 biotech ideas and one tech idea (I had written verilog for a hardware prototype of a deep learning chip) and got no bites on any of my pitches.
Finally, a friend offered me a VP product position (because he needed someone he could trust) which I was about to jump ship for, and then my current CEO counter-offered by agreeing to put his personal money as a lead in one of the biotech startups I'd pitched (that was somewhat meritocratic because he has seen how I operate as a coder in the company and was impressed, and I think he figures, probably correctly that if I'm that good a coder then my biotech skills are even better because that's what I'm formally trained in)
Since getting here, I haven't once advanced through pure merit, and have always advanced through connections.