OODA loop

John Boyd's observe–orient–decide–act loop: where it came from, what Boyd actually drew (not the four-box cycle on most slides), each phase with the questions to ask and the ways it fails, tempo and "getting inside the opponent's loop", and how to use it for competitive response, incidents, product work and sales. Decision tools in general are in decision-making; how intent lets a team run fast loops without a boss in every one is in mission command; the ownership mindset that goes with decentralized loops is in Extreme Ownership.

John Boyd in one table

E-M theory compares aircraft by their ability to gain or shed energy, using specific excess power:

where $V$ is speed, $T$ thrust, $D$ drag and $W$ weight. A positive $P_s$ at a given speed, altitude and g-load means the jet can climb or accelerate there; comparing $P_s$ plots for two jets shows where each wins a turning fight. The habit it built (turn a "black art" into measurable trade-offs, then design around them) runs through everything Boyd did afterward.

Where the loop comes from

Boyd never wrote a book. His ideas live in one essay and a set of slide briefings, some of them hours long, that he revised for two decades.

Patterns of Conflict speaks of the adversary's "observation-orientation-decision-action time cycle or loop" and Organic Design of the "O-O-D-A loop", but for about 20 years Boyd described the loop only verbally, usually as a simple circle. When he finally drew it in 1995 it looked nothing like the four-box cycle people had copied from his talks (Chet Richards, Boyd's OODA Loop, 2020).

Boyd's loop vs the four-box cycle

Boyd's note on the sketch: "Note how orientation shapes observation, shapes decision, shapes action, and in turn is shaped by the feedback and other phenomena coming into our sensing or observing window." In Organic Design he is blunter: "Orientation is the Schwerpunkt." (Schwerpunkt: the German term for the focus of main effort.)

Implicit guidance and control is the important arrow. A trained pilot, surgeon or on-call engineer does not deliberate over most moves: well-built orientation goes straight to action, and the explicit Decide step is for the cases the repertoire does not cover. Boyd's organisational conclusion is that a team whose members share an orientation needs fewer explicit orders, which is why he tied the loop to trust and to German mission tactics (Auftragstaktik; see mission command).

The four phases in depth

Observe

• Widen the aperture: customers, frontline staff, logs, competitors, regulators, your own failed experiments.
• Separate signal you pulled (you went looking) from signal pushed at you (alerts, the loudest voice).
• Orientation decides what gets observed (the implicit-guidance arrow back to Observe). If the team "doesn't see" something obvious in hindsight, the fix is usually in Orient, not in more data.

Orient

Boyd lists five inputs to orientation: genetic heritage, cultural traditions, previous experiences, new information, and analyses and synthesis across them. For a team that translates to:

Decide

A decision is the hypothesis that this action, now, will move things the way you expect. Write down the expectation so the next Observe step can check it. Match the effort to reversibility (two-way doors fast, one-way doors carefully; see decision-making).

Act

Action is the only step that changes the world and the only one that generates new evidence. An action without a defined check is not a test, just a guess you cannot learn from.

Orientation, mental models and Destruction and Creation

The 1976 essay is Boyd's theory of how orientation changes. Paraphrased:

He leans on Gödel's incompleteness theorems, Heisenberg's uncertainty principle and the second law of thermodynamics to argue that a closed system of ideas, "talking to itself", generates rising confusion and disorder; the cure is to go outside it and build a broader concept. The analogies are loose (Gödel and Heisenberg say nothing about organizations), but the practical point survives without them: mental models decay unless you deliberately break and rebuild them.

Boyd's illustration in The Strategic Game of ? and ? (1987): take the skis from a ski slope, the outboard motor from a motorboat, the handlebars from a bicycle and the rubber treads from a toy tank, discard the rest, and recombine the parts into a snowmobile. His summary in Revelation (1987): winners can build snowmobiles when facing uncertainty and unpredictable change, and losers cannot.

For practical mental-model work, see general thinking models and cognitive biases.

Tempo and getting inside the opponent's loop

Patterns of Conflict (1986) states the core claim: "Idea of fast transients suggests that, in order to win, we should operate at a faster tempo or rhythm than our adversaries—or, better yet, get inside adversary's observation-orientation-decision-action time cycle or loop."

Elsewhere in Patterns Boyd describes blitz and guerrilla forces operating "in a directed, yet more indistinct, more irregular, and quicker manner" than their adversaries: quickness is one item on the list, not the whole list. Being inside someone's loop is about relative orientation, not speed alone.

The F-86 vs MiG-15 story, and what is wrong with it

Boyd's founding example was Korea, 1951–53. The MiG-15 out-climbed the F-86 Sabre and hit harder, yet Sabres won most fights. His explanation (as recounted in Ian Brown's history and Boyd's biographies): the F-86's bubble canopy gave the pilot a better view (faster Observe), and its fully hydraulic flight controls let the pilot roll from one maneuver into the next faster (fast transients). Each exchange, the Sabre pilot re-oriented and acted sooner, and the MiG pilot fell further behind.

The evidence is weaker than the story:

The Sabre probably did win, just not by 10:1 and not for one reason. Use the story as an illustration of fast transients, not as proof that the loop wins wars (see Hankins, Flying Camelot, 2021, and Robinson's summary in the references).

Speed is not the point

Chet Richards, who worked with Boyd, argues that "faster through the loop" was never Boyd's claim. A deceiver wants the opponent to decide fast on a false picture, and a martial artist may look slower than a flailing opponent while being inside their loop. Richards on Honda's motorcycle war with Yamaha (from Stalk & Hout, Competing Against Time, 1990): Honda launched 113 new models to Yamaha's 37 in 18 months, but it won because each launch taught it more about what customers wanted. His conclusion: "Just going through some loop without learning anything is a waste of time, money, and energy, and going faster just wastes more, more quickly."

Workable rule: first get the orientation advantage (better picture, surprise, ambiguity); then use tempo to exploit it. Tempo without orientation is thrashing.

Maneuver warfare and the Marine Corps

Maneuver warfare in Warfighting's sense aims to shatter the enemy's cohesion with rapid, focused and unexpected actions, rather than to wear it down by attrition. Its vocabulary (tempo, focus of effort, commander's intent, decentralized command, surfaces and gaps) maps directly onto business competition and is the bridge from OODA to mission command. Source for this section: Ian T. Brown, A New Conception of War (Marine Corps University Press, 2018).

Business applications with worked examples

The loop is most useful where there is an adaptive counterpart (a competitor, an attacker, a failing system, a buyer) and conditions change faster than annual plans. The scenarios below are illustrative.

Competitive response

A rival cuts the price of its mid-tier plan by 30%.

The strong loop is not faster to act, but its orientation is better, and it keeps the option to change course.

Incident response

A production API starts returning 5% errors at 14:02.

OBSERVE  14:02 alert; error rate by endpoint, region, version;
         deploy log shows release 4.12 at 13:55
ORIENT   errors only on /checkout, only on 4.12 pods;
         pattern matches the new payment client (runbook #7)
DECIDE   hypothesis: 4.12 payment client; roll back 4.12
ACT      14:09 rollback started; watch error rate + latency
FEEDBACK 14:14 errors back to baseline -> hypothesis holds;
         open postmortem, keep 4.12 blocked

Implicit guidance and control is what the runbook and practice buy you: an experienced on-call engineer goes from "errors on new pods right after a deploy" straight to "roll back" without a meeting. Incident command structures exist to protect that loop from too many observers and too many deciders.

Product iteration

A team ships a new onboarding flow every week. Observe: activation rate, session recordings, support tickets. Orient: which step loses people, and why (confusing copy, missing data, wrong user). Decide: one change, written as a hypothesis ("removing the workspace step raises day-1 activation from 41% to 45%"). Act: ship to 50% of new sign-ups. Feedback: read the result the following week. The weak version ships something every week and never writes down the expected result, so the team cycles fast and learns nothing.

Sales

In a discovery call the seller's loop runs in seconds: observe what the buyer says and avoids, orient against what similar buyers turned out to need, decide the next question or proof point, act, and watch the reaction. Good sales training builds implicit guidance (experienced reps do not consult a playbook mid-call). Across a deal, the loop is slower: every meeting is an Act that should produce an observation (who else is involved, what the real deadline is, what the alternative is).

OODA vs PDCA vs Build–Measure–Learn

They are complementary: PDCA for the machine, Build–Measure–Learn for the product, OODA for the fight. See idea to MVP for the product loop.

How to improve your loop

Improving the loop is mostly about orientation and friction, not about working faster.

Loop health checklist

[ ] We can name the 3-5 signals that tell us the situation
    changed, and someone watches each one.
[ ] Frontline observations reach decision-makers in hours,
    not at the monthly review.
[ ] Our key assumptions are written down, dated, and were
    last challenged within a quarter.
[ ] Every significant decision records the expected result
    and the date we will check it.
[ ] Routine situations have pre-agreed responses; nobody
    waits for approval to execute them.
[ ] Most decisions are made by the people closest to the
    problem, within a written intent.
[ ] We run a review after incidents, launches and losses,
    and the actions from it get done.
[ ] We know roughly how long our loop takes (signal ->
    decision -> action -> visible result) and it is
    shorter than the situation's rate of change.

Templates

OODA situation card

SITUATION: ____________________   DATE: ________
OBSERVE
  What changed (facts, sources, times):
  What we are not seeing / cannot measure:
ORIENT
  Our current explanation:
  Alternative explanations (at least 2):
  Assumptions this rests on (oldest first):
  What would change our mind:
DECIDE
  Options considered:
  Chosen action (as a hypothesis):
    "If we ___, we expect ___ by ___."
  Reversible? Y/N   Owner: ______   Deadline: ______
ACT
  First step, who, when:
  Signal we will watch and when we check it:
FEEDBACK (fill in later)
  What happened vs expected:
  What we now believe that we did not before:

Orientation review (quarterly, 60 minutes)

1. List the 10 beliefs our strategy depends on
   (customers, competitors, technology, us).
2. For each: evidence for, evidence against, date last
   tested.
3. Pick the 3 weakest. For each, design the cheapest
   test that could prove it wrong this quarter.
4. Ask: what would a smart competitor conclude from the
   same data? What would a new hire find odd about us?
5. Record changes to our picture; share them with every
   team that makes decisions on these beliefs.

Criticisms and misuses

The loop is at its best as a shared vocabulary ("our orientation is stale", "that's implicit guidance, no meeting needed", "we acted, but where's the feedback?") and as a reminder that the scarce resource is an accurate picture of a changing situation.

References

• Colonel John Boyd digital archive (opens in a new tab): scans and editions of Boyd's essay and briefings (Destruction and Creation, Patterns of Conflict, Organic Design, The Strategic Game, Conceptual Spiral, Essence), plus related papers
• John R. Boyd, The Essence of Winning and Losing (1995–96; Richards and Spinney edition, 2012) (opens in a new tab): the only OODA sketch Boyd drew, with his key statements
• John R. Boyd, Patterns of Conflict (December 1986) and Organic Design for Command and Control (May 1987), available from the archive above: "getting inside" the loop, orientation as Schwerpunkt, implicit guidance
• John R. Boyd, A Discourse on Winning and Losing, ed. Grant T. Hammond (Air University Press, 2018): all the briefings in one volume, including the snowmobile slides
• John R. Boyd, Destruction and Creation (3 September 1976), available from the archive above: destructive deduction and creative induction
• Frans P. B. Osinga, Science, Strategy and War: The Strategic Theory of John Boyd (Routledge, 2007): the standard scholarly study; source of the Storr critique
• Robert Coram, Boyd: The Fighter Pilot Who Changed the Art of War (Little, Brown, 2002): the biography
• Grant T. Hammond, The Mind of War: John Boyd and American Security (Smithsonian Institution Press, 2001): a second biography and intellectual history
• Chet Richards, Certain to Win: The Strategy of John Boyd, Applied to Business (Xlibris, 2004): Boyd for companies, by a collaborator
• Chet Richards, "Boyd's OODA Loop", Necesse 5(1), 2020 (opens in a new tab): the full sketch explained; why speed is not the point; the Honda–Yamaha example
• Ian T. Brown, A New Conception of War: John Boyd, the U.S. Marines, and Maneuver Warfare (Marine Corps University Press, 2018) (opens in a new tab): Boyd's route into FMFM 1 and MCDP 1; the F-86 canopy and hydraulic-controls explanation
• US Marine Corps, FMFM 1 Warfighting (1989) and MCDP 1 Warfighting (1997): maneuver warfare doctrine
• Wikipedia: North American F-86 Sabre, Korean War kill claims (opens in a new tab): Sabre Measure Charlie, RAND and Dorr et al. revisions of the 10:1 claim
• Stephen Robinson, "The Korean War and the OODA Loop: What Happened to the Kill Ratio?" (2025) (opens in a new tab): summary of the evidence against the 10:1 figure and the alternative explanations
• Michael W. Hankins, Flying Camelot: The F-15, the F-16, and the Weaponization of Fighter Pilot Nostalgia (Cornell University Press, 2021): critical history of Boyd and the Fighter Mafia
• George Stalk Jr. and Thomas M. Hout, Competing Against Time (Free Press, 1990): time-based competition; the Honda–Yamaha war
• Eric Ries, The Lean Startup (Crown Business, 2011): Build–Measure–Learn