Play long-term games with long-term people — compounding of money, trust, and knowledge only works if you don't reset the board.
The “with long-term people” half is the part that’s actually hard. You can commit to a long game unilaterally; you can’t commit to a repeated one.
Links
Typed edges. Anything below the line was authored on the other note and derived here.
supports →Write recurring decisions down as principles — a decision made once and codified stops costing willpower and starts compounding.Dalio and Naval are both compounding maximalists, and this is where they agree without qualification: the returns come from not restarting, whatever the substrate.
← has exampleTake accountability under your own name: the downside of being wrong is capped, the upside of being right publicly compounds.
Attaching your name is what makes the game repeated rather than one-shot — reputation is only an asset if it carries between rounds.
← supported byEvolutionarily stable strategies explain behaviour with no planner — equilibria emerge from payoffs alone.Reciprocal altruism is only stable when the game repeats and defectors can be remembered. Naval reached the conclusion by intuition; this is the proof, and it names the conditions under which it fails.
← extended byCompounding rewards duration far more than rate — the dominant variable is how long you leave it alone.Naval argues for long games from trust and reputation. Housel supplies the arithmetic underneath: the exponent is the term that matters, so any reset of the board is enormously expensive.