Doing EV calculation with personal finances is also tricky, because you don't get repeated trials. In poker, assuming you have an appropriate bankroll, you shou

Doing EV calculation with personal finances is also tricky, because you don't get repeated trials. In poker, assuming you have an appropriate bankroll, you should be able to make risky but positive-EV bets hundreds of times, such that you will realize that EV. For job choices, you're not going to get more than a few cracks at it. You really need to be more concerned with risk of ruin, especially when the ratios are greater. The difference between FAANG and early-stage startups can easily be a 50% pay cut, with at least 5 years until liquidation. If you have a 50% chance of that bet paying off, then maybe it's okay - you can do this twice a decade, and you're reasonably likely to end up at least even over a career. If you have a 5% chance of the bet paying off, even if the rewards are large enough to make the EV even or positive, that's an insane amount of risk to take unless you're already financially independent, because over a career you're not likely to win even once. This is one of the fundamental power imbalances between labor and capital: Capital can diversify. To go back to the poker analogy, VCs are playing cash games - they can take 100 bets of which they expect only 2 to pan out. Employees are in a short-stack tournament - even if you make the highest-EV play, if you're going to lose most of the time, you shouldn't be going all-in.