What pacing taught us

The first version of the Vellestria roadmap we drew on a kitchen wall had a box for every chapter, a coloured sticker for every boss puzzle, a smaller sticker for every fragment, a tiny dot for every secret. We stared at it for a week. Then we noticed the wall was telling us something: the boxes were not the same shape, the bosses were not evenly spaced, the fragments came in clusters and then went quiet, the secrets were almost invisible. This is the dispatch about pacing, the thing we got wrong twice on paper and corrected slowly in the build.

1. The chapter as an act

Each chapter in Vellestria has the same internal shape, and we did not realise we were following a shape until we had finished four of them. The shape is: rising action through the lieu, one or two puzzles that turn the screw, a boss puzzle that synthesises what the chapter has taught you, an emotional beat in which the pressure releases into character, and a cliffhanger that does not resolve so much as it points.

Chapter 2 is the cleanest example we can publish without spoiling the rest. The rising action is the workshop, methodically ransacked, the third caliper in a closed drawer upside down. The screw turns through the riddle about the friend who no longer leaves the port, the deduction of Mateu Solera by a child memory of brown sparrows and oranges, and the visit. The boss puzzle, The Old Town Final Trial, gathers the chapter's vocabulary into one figure. The emotional beat is the sextant unfolding at Mateu's table, Tomás saying quietly that this place does not exist on any Council chart. The cliffhanger is the chart itself: a peninsula the official maps do not contain.

Every chapter from 1 to 7 has this shape. The boss is the synthesis. The emotional beat is the cost of synthesising. The cliffhanger is the next chapter pointing at the player from inside this one. We did not invent this; the classic adventures of the 2000s ended every chapter on a beat, not a resolution, and we have been living in that lesson.

2. Why Chapter 8 has no boss in the usual sense

Chapter 8 is the departure act, and it does not have a boss puzzle in the Chapter 1-7 sense. There is no synthesising trial in the harbour, no chapter-end riddle Cora has to solve before she boards the schooner. The pacing of Chapter 8 is the pacing of an exhale.

What sits in the boss-puzzle slot, at the close of the saga, is the secret-puzzle finale The Sundered Coordinate. It does what a boss does — synthesises what came before — but as a chosen act, not a required one. Players who reach the harbour can leave without solving it; players who solve it get the cleanest cryptogram in the published set and a private goodbye to the first nation. The slot is the same; the contract with the player is different. Chapter 8 asks rather than demands, because it is the chapter where Cora is asking the questions of the world rather than the world asking them of her.

3. How fragments are paced

Three fragments wait inside Vellestria's eight chapters: the sextant at Mateu's table in Chapter 2, the telescope's hollow base at the Observatory in Chapter 4, the panel behind the Restricted Stacks in Chapter 6. Not one per chapter, not a chain. The first version of the spreadsheet had a fragment in every chapter, dutifully, and the playtest read of that version was: it stops feeling like a discovery and starts feeling like a quota.

The pattern we settled on is front-load, then space out. Chapter 2 teaches the player that fragments exist — the sextant is found early enough that the rest of the saga can refer back to it without explanation. Chapter 4 confirms the pattern and raises the stakes: the second fragment shows that whatever Theodor was assembling was larger than Vellestria. Chapter 6 deepens it: the third fragment ties the work to the family.

After Chapter 6 there are no more fragments in Vellestria. More come in later nations. The silence between Chapter 6 and the next fragment is the part of the pacing that took the longest to defend, because every instinct in adventure design says fill the gap. We left it. The gap is where the player carries what they have found.

4. Why secret puzzles are rewards, not gates

Vellestria contains exactly eight secret puzzles, hidden in lieu hotspots placed deliberately enough that a careful player will find some and a careless one none. They are not on the critical path. A player who finds zero secrets can still finish the saga; a player who finds all eight gets a private satisfaction and a couple of Carnet entries the rest do not.

The reason this is a rule and not a default is that we spent a month, early on, trying to gate progression on secrets. Beautiful in theory: a secret unlocks context for the next chapter. Crushing in practice. Playtesters who missed the secret felt the next chapter was thin; playtesters who found it felt the secret was a tax rather than a gift. We pulled the gating out in an afternoon. Discovery rewards, not gatekeepers. Eight per nation, no exceptions.

5. The metronome we tried not to become

The most tempting mistake in pacing puzzle games is the metronome. Every thirty minutes a hint, every forty-five a reveal, every chapter a fragment, every lieu a secret. The numbers can be tuned. The result feels like a stopwatch.

We pace by the shape of the act, not the clock. Some chapters of Vellestria take a player two hours; some take five. Some hotspots reveal a secret in the first five minutes; some take three return visits, after the player has read something in the Carnet that recontextualises the room. We do not try to keep the player on a beat. We try to keep the chapter on a shape. When the shape is right, the beat takes care of itself.

The wall in our flat now has the same layout, redrawn, with bosses paced to act rather than time and fragments clustered rather than chained. The secrets are still tiny dots. Four nations remain. We expect the shape to deform a little for each of them; that, too, is part of the pacing, because a saga that maintains a perfectly identical shape across five acts is not pacing — it is wallpaper.