a builder's codex
codex · operators · Morgan Housel · ins_housel-reasonable-beats-rational

People don't want the mathematically optimal strategy — they want one that lets them sleep at night. Reasonable beats rational in practice.

By Morgan Housel · Partner Collaborative Fund; author The Psychology of Money, Same as Ever · 2020-09-08 · book · The Psychology of Money — Reasonable vs. Rational

Tier A · TL;DR
People don't want the mathematically optimal strategy — they want one that lets them sleep at night. Reasonable beats rational in practice.

Claim

The strategy that wins long-term is not the mathematically optimal one — it is the one the operator can actually sustain. Human decision-making is governed by emotional comfort and psychological sustainability, not pure utility maximisation. A "reasonable" strategy that lets the operator sleep at night will be adhered to for years; a "rational" strategy that produces stress will be abandoned at the worst time. For products and behaviour-design, reasonable beats rational.

Mechanism

Compounding requires duration; duration requires that the operator stays in the strategy. Mathematically optimal strategies often produce drawdowns, ambiguity, or counter-intuitive choices that strain the operator's tolerance. The strain accumulates and eventually breaks the strategy at the wrong moment (sell at the bottom, pivot at the trough, switch to something exciting after a flat year). A reasonable strategy is one calibrated to the operator's actual psychology — slightly suboptimal in expected value, dramatically superior in realised duration. The product implication is sharp: design for the user's psychology, not the user's spreadsheet. Personal-finance apps, habit trackers, retirement plans, and onboarding flows all benefit from "reasonable" defaults that produce sustained engagement over "rational" defaults that produce abandonment.

Conditions

Holds when:

Fails when:

Evidence

"people do not want the mathematically optimal strategy, they want a strategy that lets them sleep at night, which is why \"reasonable\" beats \"rational\" in practice"

— see raw/expert-content/experts/morgan-housel.md line 18.

Signals

Counter-evidence

"Reasonable" can curdle into excuse-making for genuinely suboptimal behaviour. The discipline is matching the reasonable-vs-rational gap to the cost of sub-optimisation: a 10% premium for sustainability is usually worth paying; a 50% premium probably isn't. Some categories require rational optimisation regardless of operator psychology (institutional finance, some performance-engineering contexts).

Cross-references

Open the interactive view → View original source → Markdown source →