# Common Ground — Debrief Sheet

> 8 questions, 3 failure modes, one rule: verified ground is the only ground.
> Use with `common-ground-guide.md`. Debrief runs 10 min after the read-out.

---

## The 8 questions

**1. The three-pointer.**
"Read your group's best 3-point ground again. What question uncovered it?"
→ The question is the skill; the fact is just the receipt.

**2. The easy camp.**
"Which 1-point grounds did your group write before it started hunting? What finally moved you off them?"
→ Targets: the cost of settling for the first answer — a daily-work habit.

**3. The verification moment.**
"When did verifying almost break a ground? Who almost nodded falsely, and what stopped them?"
→ Targets: psychological safety. Handle gently; the honesty is the win.

**4. The quietest member.**
"What did the quietest member contribute that the group almost missed?"
→ Almost always the best fact in the game came from the person who spoke least.

**5. The surprise ranking.**
"Which ground from *another* group surprised you most? Why that one?"
→ Cross-group learning; also tells you what this room values.

**6. The daily version.**
"Where does your week reward the first easy answer instead of the second question?"
→ Transfer question. Expect: stand-ups, requirements interviews, support tickets.

**7. The one question.**
"What is one question from this game you will ask a colleague this week?"
→ Commitment question. One question, named, with a deadline.

**8. The group identity.**
"Your recorder card has ten lines and your names on it. What is this group now — one sentence?"
→ Closing question. Groups keep these cards; the sentence becomes the group's name for itself.

---

## 3 known failure modes of the debrief itself

1. **Read-out becomes a talent show.** Groups perform their funniest facts; nobody reflects.
   Rescue: question 1 first — always the question behind the fact.

2. **A false verification gets exposed.** Someone admits they nodded along to a 3-pointer that wasn't true.
   Rescue: praise the confession louder than the fact — "verification just worked, in front of us."

3. **The debrief skips the quiet.** Loudest groups dominate again.
   Rescue: question 4 goes to a named quiet member, with a five-second silence buffer before they answer.

---

If you are running the pre/post measure (`common-ground-assessment.json`), note these observables during the hunt:

| Observable | Points to |
|---|---|
| Group asks a second question after an easy "yes" | Curiosity habit |
| Quietest member's line is written down | Information flow |
| A ground dies at verification and the group thanks the verifier | Safety culture |

The report your participants receive is a self-comparison, never a group ranking — keep it that way.