Wall the smallest cluster that already scores, and only expand an enclosure you can afford to finish. The wall thread’s practical sequence: choose the inside first, count the perimeter against the wall sets and blueprints you actually hold, close the loop, and only then evaluate what the enclosure earns. An unfinished ring is the documented failure — not a smaller bonus, a wasted one.
Wall a cluster that already scores
The enclosure is the multiplier; the cluster is the number it multiplies. Players in the wall thread advise starting with a compact Civic-style scoring group — a few tiles that already earn — rather than fencing a field you hope to fill later. One reply also notes the build itself can pay along the way when other create- or tag-based benefits trigger from placing wall pieces, which is the only time a half-built ring is not dead turns.
Closed ring versus one gap · the boundary is the enclosure
Closed · continuous boundary
Gap · one missing segment breaks it
inside · scoring clusterwall boundarymissing segment
Count the perimeter before you commit
Perimeter grows faster than intuition suggests, and wall supply is finite: the thread’s failure case is a player who received only two wall sets for a design that needed three and could not close the ring. Before the first tile, count the ring your cluster needs and compare it against the wall pieces in hand, in the shop, or substitutable by blueprint — one player leans on blueprints precisely to avoid spending turns on wall pieces, and another patched a shortfall with a Printing Press and an Imprint Charm. If the count does not close, shrink the inside, not your standards.
Close it, then judge it
Only a finished enclosure can be evaluated: the documented payoff logic rewards completion, and players report judging the bonus only after the last gap closed. Resist the urge to pre-expand because the ring “almost works” — filling empty space inside an existing ring adds no perimeter, but enlarging the outline itself may demand wall supply you do not have, so count the new ring before expanding. Once closed, check what the enclosure actually changed for the cluster’s scoring, and decide the next ring with fresh numbers rather than sunk cost.
Commit or stop · supply decides
When to stop building walls
- Supply ran out: no sets, no blueprints, nothing in the shop. A paused ring is a loss, so stop before starting it — the milestone clock keeps running either way.
- The ring costs more than the cluster earns: recompute with the real payoff after your first closure; walls are a means, not a deck identity.
- A better engine exists for the same turns: compare against simply extending a working chain from the guild pair plans.
- You are improvising edge rules: the sources do not settle coast edges, diagonal joins or mixed wall types. Plan rings you can close with ordinary straight runs instead of betting on unsettled rules.
Timing interacts with milestones: an enclosure that closes with weeks to spare keeps paying, while one finished at the deadline bought a bonus with no time to use it — the end-early decision covers that trade.
Cases and counts come from the walls discussion (Aug 31–Sep 12, 2026, several authors), read Oct 5, 2026; an older pre-release thread documents which edge rules stayed unanswered. Player-reported planning material — no fixed ring sizes, no current-client test.