# Blind Portrait — Debrief Sheet

> 8 questions, 3 failure modes, one rule: the drawings are the data.
> Use with `blind-portrait-guide.md`. Debrief runs 10–20 min after the two rounds.

---

## How to run this debrief

- Keep both drawings on the table, side by side, for the whole debrief. Every question points back at them.
- Ask questions **to the pair first**, then to the room. The pair has the freshest evidence.
- Write answers on this sheet or on the back of the drawings — participants keep them.

---

## The 8 questions

**1. Shapes, not stories.**
"Describer: when did you catch yourself explaining what the image *means* instead of what it *looks like*?"
→ Targets: instruction clarity. Most describers do this in the first 30 seconds.

**2. The silent guess.**
"Drawer: what did you do in the moments you weren't sure?"
→ Targets: assumption behavior. Watch for "I just drew something plausible" — that is the finding, not a fault.

**3. The cost of the rule.**
"Round A, no questions allowed: how much of the gap between your two pictures came from the rule, and how much from the description?"
→ Separates constraint effects from skill effects. Don't let them blame the rule entirely.

**4. The three questions.**
"Round B, you could ask three questions. Which question helped most? Which would you ask first if you could restart?"
→ Surfaces the repair move: check understanding *before* starting.

**5. Same pair, different result.**
"When you swapped roles, what did you do differently as describer — knowing what it felt like to draw blind?"
→ Perspective-taking evidence. This is the strongest single indicator in the debrief.

**6. The daily version.**
"Where does this exact pattern — sending instructions into silence — show up in your week?"
→ Transfer question. Expect: email specs, handovers, ticket comments, stand-up updates.

**7. The one repair.**
"If you could install one habit from this exercise into your team's communication, which one?"
→ Commitment question. Make it singular — one habit survives Monday.

**8. The drawing you'd frame.**
"Which of the two drawings is the honest portrait of how your pair communicates today?"
→ Closing question. It ends the debrief on the evidence, not on a platitude.

---

## 3 known failure modes of the debrief itself

1. **Laughter eats the lesson.** Round A drawings are funny; the room will want to stay in the laughter.
   Rescue: laugh with them once, then place the two drawings side by side and ask question 3. The contrast quiets the room.

2. **Blame lands on the describer.** The gap feels like one person's fault.
   Rescue: reframe — "the describer described alone; the drawer drew alone; the *pair* produced the gap." Question 5 is the pivot.

3. **Meta-talk replaces examples.** Participants start speaking in abstractions ("communication is important").
   Rescue: point at the drawings. "Show me where on the paper that happened."

---

## Scoring cues (assessment-type sessions)

If you are running the pre/post measure (`blind-portrait-assessment.json`), note these observables during the rounds — they justify the delta in the report:

| Observable | Points to |
|---|---|
| Describer checks in before starting (round B) | Repair habit forming |
| Drawer asks a question in the first 30 s | Uncertainty tolerance |
| Pair scores round B higher without a "better artist" | Understanding, not skill, improved |

The report your participants receive is a pair-level comparison, never an individual ranking — keep it that way when you present it.