Decision-Making Frameworks: OODA, WRAP, and Cynefin Compared

A practical guide to the three frameworks people reach for when a decision matters — and what changes when the decision is irreversible.

Why frameworks at all

A framework is a shape you press onto a moment so the moment stops moving long enough to think. None of them tell you what to do. They tell you what to look at, in what order, before you commit.

The OODA loop

Observe, Orient, Decide, Act. Built by John Boyd for fighter pilots: the operator who cycles fastest wins, because the slower one is always reacting to a world that already moved. OODA shines in adversarial, fast-feedback environments — a sales call, a live incident, a negotiation mid-sentence.

Where it breaks: when feedback is delayed by months or the wrong move can't be undone. Speed becomes recklessness when you can't loop again.

WRAP

Chip and Dan Heath's antidote to the four villains of decision-making: Widen your options, Reality-test assumptions, Attain distance before deciding, Prepare to be wrong. WRAP is built for decisions you'll only make once — career moves, hiring, a big purchase.

Where it breaks: live combat. You don't widen options while a kitchen fire spreads. WRAP wants room.

Cynefin

Dave Snowden's sense-making model sorts a situation into five domains: clear (best practice applies), complicated (experts know), complex (patterns emerge only after you probe), chaotic (act first, then sense), and confusion (you don't yet know which domain you're in). Cynefin doesn't tell you what to decide — it tells you which kind of thinking the moment deserves.

Where it breaks: it's a diagnostic, not a decision. You still need OODA or WRAP once you've named the domain.

A simple chooser

Fast, adversarial, reversible — OODA.
Slow, life-shaped, reversible-ish — WRAP.
You can't tell what kind of problem this is — Cynefin first, then one of the others.

What changes when the decision is irreversible

Every framework above assumes you get another turn. OODA assumes another loop. WRAP assumes you can course-correct after the call. Cynefin assumes you can probe.

Irreversible decisions break that assumption. There is no second observation, no reality-test, no probe. The cost of being wrong is the rest of the story. The honest framework for irreversible choices is much shorter: name the thing you can't take back, name who pays if you're wrong, decide anyway.

That's the question Decision Nodes is built around — an interactive drama where the only decisions that count are the ones you can't undo.