Tool 18 of 32 · about an hour

Room Bias Scan

Before the call, each person names the distortion most likely pulling on them today. Naming reduces grip.

Fifth check in the Depth working session, round the table, each person naming the distortion pulling on them today, before the depth verdict is final.

What it is for

To name what is distorting judgment in the room, right now, so a perfect set of readings does not still produce the wrong call.

The failure it catches

The readings can be perfect and the call still wrong, because the people making it are compromised in predictable ways. It is also the guard against cycling through the other four checks until one supports a feeling.

Why it works

Naming the distortion most likely pulling on you today, yourself, not your neighbour, reduces its grip. Diagnosing other people's biases is how the scan becomes a weapon and goes blind.

How to run it

  1. Before the call, name the seven distortions: sunk cost, we have already done so much; urgency, everything feels like a fire so the patch always wins; ego preservation, admitting the deep fault feels like admitting failure; narrative, we are so close and the story must stay linear; recency, the latest bug feels like the real one; social pressure, nobody wants to be the blocker; ambiguity aversion, the symptom fix feels clear and the structural fix feels scary.
  2. Run it round the table. Each person, including whoever is running the session, names the one most likely pulling on them today.
  3. If the round produces only nothing, really, whoever is running the session goes first with a real one, and the round runs again.
  4. Nobody is pushed to name a bias they will not name. A refused round is logged, and the Diagnosis Bias Scan carries more of the weight on paper afterward.

How to read the result

Naming reduces grip. If someone has cited only the check that agrees with their feeling, this is the check that catches it. All five checks still run, in order. The depth verdict then goes on one page: what we fix, to which layer, the stop line, the honest cost including re-fixing symptoms on top of the root, and what we are explicitly not doing, priced.

How to run it well
  • Each person names the distortion pulling on them, not on their neighbour. The round-the-table format is the defense.
  • If you are running the session, go first with a real one when the round is fake.
Where it breaks
  • Diagnosing other people's biases instead of your own.
  • Check-shopping: cycling through the other four checks until one supports the feeling. All five, every time.
  • Narrative lock-in: so close, push through, while every check reads red.
What would prove it wrong

If named biases show no correlation with which way people argue the depth call, the scan is theatre.

What we watch

Whether the named distortion matches how that person argues the depth call, or whether the round is only a ritual before the same argument.

Ask yourself
  • Which of the seven distortions is most likely pulling on you today, not on anyone else?
  • Have you cited only the depth check that agrees with how you already felt?
  • Is we are so close doing work that the artifacts are not doing?
  • Which layer would be uncomfortable to name out loud, and is that discomfort where the leverage is?

Fed by: Convergence Test, Stability Test

Feeds: Diagnosis Bias Scan, Decision Contract

Get the toolkit