# Blind Portrait — Facilitator Run Sheet

> Folder: `blind-portrait` · Category: communication · 2 people per pair · 15 min · offline · print props · assessment-type
> License: CC BY 4.0

A back-to-back drawing exercise where one participant describes an image and the other draws it — without either seeing the other's work. The debrief turns the gap between the two pictures into concrete evidence about instruction-giving, listening, and assumption-checking.

---

## Objectives

By the end of the activity, participants will have:

1. Experienced the cost of **one-way instructions** versus checked understanding.
2. Seen their own default role: describer who assumes, or drawer who guesses silently.
3. Practiced one repair move: **asking one clarifying question before starting**.

---

## Setup

**Space:** two chairs per pair, back-to-back, ~30 cm apart. No table between them — the drawer holds the pad on their knees.

**Materials per pair:**

| Item | Source |
|---|---|
| 1 × describer card | `blind-portrait-prop-cards.pdf` (card 1) |
| 1 × drawer card | `blind-portrait-prop-cards.pdf` (card 2) |
| 1 × judge card | `blind-portrait-prop-cards.pdf` (card 3) — cut, one per observer |
| 1 × timing card | `blind-portrait-prop-cards.pdf` (card 4) — for you, the facilitator |
| 1 × mystery image | any line drawing: a bicycle, a lamp, a floor plan. Prepare 2 per pair (round A / round B) |
| Paper + marker per drawer | thick marker, not pencil — mistakes should be visible |

**Grouping rule:** pair people who work together daily for round A; randomize pairs for round B. The contrast is the lesson.

**Roles:** describer (holds image, may not show it), drawer (holds marker, may not ask questions in round A), observer (if you have an odd number — otherwise you float).

---

## Minute-by-minute script

| Time | Phase | Facilitator does | Participants do |
|---|---|---|---|
| 0–2 | Frame | "One of you will describe, one will draw. You may not turn around. The describer may not say what the image **is** — only what it **looks like**." | Take cards, split roles |
| 2–5 | Round A (3 min) | Start timer on the timing card. Say nothing else — silence is part of the design | Describer describes; drawer draws; no questions allowed |
| 5–7 | Reveal | "Turn around. Compare." | Pairs laugh / wince; observers note gaps on judge card |
| 7–10 | Round B (3 min) | Swap roles, new image. Rule change: "the drawer may ask **three** questions total" | Second round runs with checked understanding |
| 10–13 | Reveal + compare | Prompt: "what changed between round A and round B?" | Pairs self-score both rounds on the judge card |
| 13–15 | Bridge to debrief | "Keep both drawings — they are your data for the debrief." | Lay drawings side by side |

Then run the debrief — see `blind-portrait-debrief.md`.

---

## Known failure modes & rescues

1. **The describer narrates the whole story instead of the shapes.** ("It's like a farm…") Rescue: stop the round for 10 seconds, repeat the rule — "describe what you **see**, not what it **means**."
2. **The drawer goes abstract to avoid being wrong.** Rescue: the debrief question "what did you do when you weren't sure?" is built for this. Do not fix it live.
3. **Pairs finish early and disengage.** Rescue: hand them the judge card — score round A vs round B while others finish.

---

## Variations

- **Online:** describer on camera only, drawer on shared whiteboard. Keep the no-questions rule — the constraint is the activity.
- **Large groups (20+):** run in trios (describer / drawer / observer), 2 rounds of 4 minutes, gallery-wall the drawings before debriefing in plenary.
- **Assessment-type:** run the pre-measure *before* pairing up, and the post-measure right after the debrief — see `blind-portrait-assessment.json`. The interesting number is not the score, it's the **delta**.

---

## Props in this folder

- `blind-portrait-prop-cards.pdf` — poker print-ready: describer / drawer / judge / timing cards
- `blind-portrait-guide.md` — this run sheet
- `blind-portrait-debrief.md` — 8 debrief questions, 3 failure modes
- `blind-portrait-assessment.json` — pre/post measure preset
- `blind-portrait-paper-form.pdf` — the A4 answer sheet: the measure as a print-and-fill form
- `blind-portrait-workbook.pdf` — participant workbook: the idea in plain words, guided items, notes and action pages
- `blind-portrait-slide-deck.pdf` — the slide deck for teaching or projecting
- `blind-portrait-slide-deck.html` — the same deck for any browser, speaker notes on N
- `blind-portrait-license.md` — what you may do with all of this, in plain language
- `blind-portrait-activity-poster.pdf` — The whole activity on one wall (A3): meta, run of show, debrief questions, watch-outs.
- `blind-portrait-start-here.md` — one page, thirty seconds, then print and go
- `SKILL.md` — machine-readable manifest for AI agents
- `blind-portrait-run-sheet.pdf` — the print edition: this run sheet + the debrief sheet, bound as one A4 PDF

## Agent version

Every file in this folder has a stable public URL — an AI agent can fetch the props, the debrief and this run sheet with three curl commands (see `SKILL.md`). The paper version itself needs no screen at all: no projector, no accounts, just markers and chairs.