Opening, Exploring, Closing

Definition

Any well-designed game (or meeting, or workshop) unfolds in three acts with different, mutually exclusive purposes: opening is divergent — get the widest possible spread of ideas onto the table with no criticism; exploring is emergent — look for patterns, test combinations, let unexpected things surface; closing is convergent — narrow the field, evaluate critically, and commit to decisions and next steps. The authors' governing rule: "don't open and close at the same time... you can't be creative and critical simultaneously," and "close everything you open," or the group loses momentum and feels it has opened a Pandora's box with no plan to address it.

In the Book

Chapter 1 introduces the three-act shape directly off the "game has a shape... a stubby pencil sharpened at both ends" metaphor: an initial state A, a target state B, and the exploration space between them. Each act gets a distinct keyword — divergent (opening), emergent (exploring), convergent (closing) — and the book compares the acts to a courtroom trial's opening and closing arguments. Chapter 2 elevates opening and closing to essential #1 on its "10 Essentials" list, comparing it to breathing: "it underlies every activity, giving it rhythm and life." The book then shows how games chain into series — three short games in a row, or a variant it calls break out/report back, where a large group diverges into small breakout groups, plays a game, and reconverges to report results, letting a facilitator run multiple parallel goals without losing coherence. A full-day workshop design example shows morning divergence, a split exploration phase around lunch, and afternoon convergence flowing to one outcome — explicit proof that the three-act pattern scales from a five-minute icebreaker to a multi-hour session.

Why It Matters

This gives anyone who runs group sessions a concrete diagnostic for why brainstorming meetings fail: most bad meetings collapse divergent and convergent thinking into the same ten minutes, so ideas get shot down before they're fully formed and nothing ever gets both generated in volume and rigorously filtered. Separating the phases in time — rather than by better facilitation technique within a single phase — is the actionable fix, and it transfers to any domain where idea-generation and idea-evaluation compete for the same room: product strategy offsites, research brainstorms, even individual writing (draft, then edit, never both at once).