LLenny's Podcast
← All frameworks
StrategyRory Sutherland (Vice Chairman of Ogilvy UK, author)

The Double Lock (Sober and Drunk)

Debate every big decision twice — once rationally, once psychologically — and only act if both agree.

Difficulty
Easy
Time to result
~weeks to results
Steps
5
Confidence
90%

Corporate decision-making has a structural asymmetry: creative people must submit their ideas to rational people for approval, but rational people never submit their answers to creative people for a better one. Sutherland's fix, borrowed from Herodotus' account of the ancient Persians, is a deliberate two-stage lock — run every consequential decision through both a logical and a psychological pass, and act only if it survives both.

Origin

Herodotus records that the ancient Persians debated everything twice, once sober and once drunk, and acted only if the decision held in both states. Sutherland proposes it as a template for a 'double lock' on business decisions.

Core principles

  • 01There is a fundamental asymmetry: creatives present to rationalists for approval, never the reverse.
  • 02The two passes may be rational-versus-emotional appeal, or simply two different generative states — either way, the second pass is a genuine veto and a genuine idea source.
  • 03Real business decisions violate every condition of the school-exam model: multiple right answers exist, information is incomplete, and some data in front of you is irrelevant.
  • 04The opposite of a good idea can be another good idea — high-touch and no-touch hotel check-in are both excellent and are exact opposites.
  • 05Psychology is a branch of complexity theory: small inputs can have huge nonlinear effects, so proportionality reasoning misleads.

How to run it

  1. 1

    Run the rational pass first

    Produce the logical, defensible, spreadsheet-supported answer to the question, exactly as you would today. This is the answer your organisation is already good at generating.

  2. 2

    Schedule a separate psychological pass

    In a distinct session — different time, different state, ideally different people — ask the psychological question: does this appeal emotionally, does it feel right, and is there a weirder, neater, more cunning idea that the rational pass could not see?

    Pro tip Make it a standing step, not an optional one. The whole point is that engineers and accountants never voluntarily take their answer to the wacky people first.

    Watch out Do not merge it into the rational meeting. In one room, the rational answer always wins by default.

  3. 3

    Generate the opposite of the good idea

    Deliberately construct the inverse of your leading option and evaluate it on its merits — in psychology, the opposite of a good idea can be another good idea.

    Pro tip Sutherland's example: the two great ways to check into a hotel are insanely high-touch and completely no-touch. Both are excellent; both are distinctive; they are opposites.

  4. 4

    Require both locks to open

    Proceed only where the rational and the psychological pass agree. Where they conflict, treat it as a signal that the problem is not yet understood rather than as a tie to be broken by seniority.

  5. 5

    Discard the exam-question assumptions

    Explicitly check the four assumptions imported from school maths: a single right answer, all information given being relevant, no missing information, and proportional linear effects. Real decisions meet none of them.

    Pro tip Practise on the ship-and-goats problem — most bright students will not answer 'it is impossible to tell' because they assume the data given must produce a single answer.

    Watch out We select and promote people on their ability to perform these artificial, exam-shaped activities, which is why the bias is baked into your senior team.

In the wild

The Persians' twice-debated decision

Herodotus reports that the ancient Persians deliberated every matter twice — once sober and once drunk — and only proceeded if the same conclusion held in both states.

Sutherland reads this as a double lock on decision-making: a check that the choice satisfies both the rational and the emotional/psychological faculty, rather than only the one your culture rewards.

The captain's age problem

In a famous experiment, children are told a ship takes on 27 goats, three sheep get off, 25 cows are added, and then asked how old the captain is. Pretty intelligent kids will not answer 'it is impossible to tell' — they assume the numbers in the question must yield the answer.

Sutherland uses it to show how thoroughly schooling habituates us to assume single right answers and fully sufficient information — assumptions that never hold in real business decisions.

Common mistakes

Letting rationalists be the only gate

Creative ideas must pass a rational review, but rational answers are never taken to creative people to see if a neater, more cunning idea exists. The asymmetry silently caps the quality of every decision.

Assuming a single right answer

Real decisions have multiple right answers, incomplete information, and irrelevant data mixed in. Applying exam logic to human behaviour produces confident, wrong answers.

Assuming proportionality

Psychology is a branch of complexity theory. Butterfly effects mean the scale of the input frequently has no relationship to the scale of the effect.

Is it for you?

Best for

Leaders and product teams making consequential, human-behaviour-dependent decisions inside organisations where the rational case always wins by default.

Not ideal for

Genuinely deterministic, closed-system decisions (engineering tolerances, compliance thresholds) where there really is a single right answer.

From the transcript

the ancient Persians when they had to deliberate they debated everything twice once while sober and once while drunk and only if they agreed in…

56:30

there's this fundamental asymmetry which is creative people have to present their ideas to rational people for approval

57:00

what I say is the opposite of a good idea can be another good idea

16:30

they'd assume that the information in the question because it's in the question had to be of use in formulating the answer

20:00

From the episode

What most people miss about marketing

Rory Sutherland (Vice Chairman of Ogilvy UK, author)