tessera
FACILITATOR GUIDE

How to run a Tessera workshop.

Tessera is most useful when you treat it as a 30-minute scaffold for a 60-minute conversation. The game produces an artifact; the value lives in what you talk about after.

Before the workshop

Two ways to set up a game: build the room live (default — share the code, players self-join, you allocate pairs from the lobby) or upload a pre-built CSV (already-decided pairs, you get back per-person join URLs to paste into a calendar invite). The CSV path is the right call when the workshop is scheduled and you want every link in your invite ready to go.

  • Pick a complexityfrom 1–8. Start at 3 if your team hasn't played before; bump to 5–6 once they get the rhythm.
  • Pick a round count (1–5). Multi-round is the easiest way to teach the rules in round 1 and crank the difficulty for round 2.
  • Decide on briefs per side. Library is the safe default. Pick Custom when you want to point the lesson at a specific theme. AI-generated is unpredictable in a good way.
  • Pick remote or in-person at the top of the host form. Remote shows fields for a workshop video call link and (optional) per-pair breakout rooms — Google Meet or Jitsi. In-person hides all the call/whiteboard fields; everyone is already together.
  • Per-pair breakouts are most useful for 3+ pairs. Jitsi is the friction-free option (no accounts, no sign-in) and works on every deployment. Google Meet requires the GM to sign in once with Google, and asks players for an email at join time so they can be added as Calendar event attendees and bypass the Meet knock screen.
  • Tessera doesn't carry voice. Pairs talk through whichever call link you set up — the workshop main room for everyone, plus the pair's breakout link if you minted them.
  • Save your host recovery URL.The create modal shows it once. If you close the tab without saving, it's the only way back to the dashboard. Players see the same kind of URL when they join.

Pre-built games (CSV upload)

On the home page, click ⬆ upload pre-built game (CSV) next to the create button. The modal explains the shape of the CSV and links to a downloadable template — four columns: name, email, team_name, role. Each team needs one builder and one guider; any extra rows in the same team become observers on that pair.

After you upload, Tessera creates the game, the participants, and the pairs in one shot, then hands back the same CSV with an extra join_urlcolumn — one unique recovery URL per row. Paste those into your calendar invite or email and every participant lands directly on /play with their seat reclaimed when they click. Tessera doesn't persist the CSV server-side; once the success modal closes the populated version is gone (the URLs themselves still work for the life of the game).

At kickoff

Pre-round, the master view is a single column with three numbered cards — 1. Invite players, 2. Pairs & observers, and 3. Game settings. Walk down the cards in order; the Super powers rail and focused-pair canvas only appear once the round starts.

  • Share the game code. Players join from the home page; let them choose a display name they're happy seeing in their team's history.
  • Decide whether to let players pick their own role. Builders typically enjoy the hands-on side; guiders are best for verbal communicators. Observers are great for skeptics who want to watch first.
  • Use auto-allocateif you don't care about pairings. Use manual pair when you want to mix functions deliberately (engineer ↔ designer, junior ↔ senior).
  • Need to flip everyone's role for round 2? Pre-round, the pair sidebar shows a ⇄ swap all pill — one click flips builder ↔ guider for every fully-paired team. Per-pair ⇄ swap pills are also there if you want to flip just one team.
  • For larger rooms, click on the pair sidebar to open the fullscreen roster view: a participant table with search by name, team, or partner. The sidebar is for live monitoring; the roster view is for setup wrangling.
  • Encourage pairs to give themselves a name once they read their briefs — “The Pelicans” beats “Sam ↔ Jules” on the leaderboard. Anyone in the pair (and the GM) can rename.

While the round is running

Watch the dashboard, not the participants. Each pair row shows a live progress bar + completion percentage; pairs that solve the puzzle flip green. The right pair to nudge is the one with the lowest accuracy after 60 seconds — they're the ones who've been talking past each other.

  • Frustrated pair? 🔮 Prototype unlock (3–15 s glimpse) gives the builder a brief, deliberately-degraded peek at the goal. Testing is on by default — when the builder hits Test solution, the guider sees the same correctness halos mirrored on the goal canvas, so directions that landed read immediately.
  • Pair too confident?⏱ Time pressure, ✎ Requirement change, or ▲ Make it harder punctures the assumption that they're close to done.
  • Pair gone silent?📖 Reveal briefs unblocks them — but you're trading away the lesson, so save it for the last few minutes. ✦ Change guider brief / Change builder brief is gentler — fresh constraint, same conversation.
  • Pair too far behind? ▼ Make it easier re-rolls the goal at −1 complexity. Use sparingly — keeps the room moving without making the puzzle trivial.
  • Need more time? The +30s / +1m / +2m chips next to the timer extend the round; combine with Time pressure on a confident pair to keep the field even.

The Scoring tile (top of the rail) lets you tune points per correct piece (1–100, default 10) and toggle a per-wrong penalty (−1 per misplaced piece). Mid-round changes pop a confirm modal because they retroactively recompute every score.

If you enabled per-pair breakouts at game-create, each pair row on the dashboard surfaces the pair's call link with a copy button so you can drop it into chat or open it to drop into a room mid-round. Players see the link in their top bar as Breakout room alongside the workshop's Main room; the breakout takes precedence as the primary CTA on their lobby + play screens.

Debrief prompts

The game ends; the workshop begins. Tessera's game-end view ships with the prompts below already on the screen so you don't have to remember them. Try these in order:

  1. “What was your brief?” — read them out loud. The revelation is half the lesson.
  2. “When did you realise the brief was getting in the way?”
  3. “What did you stop asking each other?” — the silences are the data.
  4. “Which of these constraints actually exist on your team today?” — the hardest question; sit with it.

Variations to try

  • Cross-functional pairs.Pair an engineer with a designer; use a custom brief like “refer to everything in the third person” to surface how each role frames problems.
  • Same-discipline pairs.Pair two engineers; use a “three words or fewer per turn” brief to surface how much shared context they really lean on.
  • Big group, multiple rounds. Run round 1 at complexity 3 with briefs off — let everyone get the rules. Round 2 at complexity 5 with briefs on. Round 3 with the same briefs but fire ▲ Make it harder mid-round to push the strongest pair.
  • Observer-led debrief. Assign the most senior person on the call as an observer. They see both canvases and the briefs; when the round ends, ask them to narrate what they saw before you reveal the briefs to the rest of the room.