Campaign Map

Customize the title and subtitle shown at the top of the map.

Post a message visible to all players on the map. Posting a new one replaces the current announcement.

The active cycle drives the countdown timer — its end date is the deadline players see. Lore is published to the NCOPR website alongside the battles fought during the cycle.

Closes all open events and assigns new ones to random active hexes. Closed events are kept as campaign history, not deleted. Also resets the countdown by sliding the active cycle's window forward. This cannot be undone.

Add a new scenario to the pool available during event creation and monthly rerolls.

Edit resource totals directly for a faction. Values shown are current amounts. Resources cannot go below zero.

Credit all factions with resources generated by their placed buildings. Run once per campaign cycle.

Standalone player poll on upcoming app features — global, not per-faction. Create a feature, then add your own options under it; multiple features can be open at once, each resolved independently. Players rank the options (ranked choice, partial rankings allowed) and can change their ranking until that feature resolves. Resolution is instant-runoff: lowest-ranked options are eliminated round by round until one wins a majority.

Each faction gets a fixed number of actions per cycle (actions_per_month). Queued actions do nothing until Advance Cycle runs them — income is credited first, so a building constructed here does not produce until the following cycle. An action that has become illegal by then is skipped with a stated reason and costs nothing.

Every action attempted, and what became of it. Nothing is deleted here except a queued action you withdraw — a withdrawn action was never taken, so it leaves no row.

Every contested hex with battle reports this cycle. Resolution reads every approved report on the hex in this cycle — an event is not required, and a hex without one resolves the same way. Resolve conflicts before Advance Cycle so territory settles first. The winner is a weighted draw, so a preview is taken once and kept: reopening shows the same result, and only Discard takes a fresh roll.

Accepted results, newest first. These are public — the same breakdown is readable by players and by the website.

Each faction funds one technology at a time (D82), and only that bar advances. RP settles at Advance Cycle, not now. A completed bar clears its own target; leaving a target blank holds the faction's RP unallocated. Bars are permanent — switching defers, it never destroys.

The active passive, once per faction per cycle. The flat and catch-up passives need no control — they are computed at settlement. Iron's Forward Reckoning surveys a hex adjacent to its territory; the other two target a hex the faction controls. Ruin's reading can fail, leaving that hex a hazard that produces nothing for one cycle.

Each faction's leader scores every legal action and proposes up to three. Preview scores without touching anything; Propose writes the chosen actions into the ledger as queued, and Advance Cycle executes them exactly as it executes a GM's own. Re-proposing replaces the leader's previous proposal for this cycle and leaves actions you queued by hand alone.

Each active leader is handed its own top-scoring legal, affordable candidates and returns a proposal in its own voice; the engine decides what is legal, so the model cannot invent an action (D17). Research is a separate button because thirty model calls in one request would time out (D75) — and there the sponsor and the technology are already fixed in the data, so the model only writes the argument. Reactions is pass 2, run separately so a failure in it never discards pass 1. Where a response cannot be used the leader still proposes, on the engine's top pick, flagged below and re-rollable on its own (D18). Nothing here is visible to players until you Publish.

A safety valve, not curation — you are not choosing what the leaders want, only catching a bad generation before players see it. Re-roll one leader, edit a statement, or strike a proposal. Publish opens the ballot from what is staged here, carrying each leader's own words, and posts to Discord. Two leaders proposing the identical action are merged into one consensus entry (D26), or ranked choice would split the vote between them. Open Action Ballots below stays as the no-model path; it builds from engine scores and carries no statements.

Opens one ballot per faction — one proposal per active leader, with two leaders proposing the identical action merged into a single consensus entry (D26). Players rank them on the Council; Advance Cycle elects the top three in order and queues them, and that order is the order they are carried out in. Re-running leaves an already-open ballot alone, because it may already hold votes. Scores stay hidden from players until the ballot closes (D63); a seat the vote cannot fill falls back to engine score, and each row says which it was.

What each leader considered, and why every candidate won or lost. Scores are the ones recorded when the decision was taken, not recomputed — the board moves, and a recomputed score would disagree with the action beside it.

BLD §15's tracking, aggregated over the rows the campaign has already written — nothing here is recomputed, so a figure reads the same in a later cycle as it did when it was taken. GM only: BLD §14.3 says so in its own words, and per-leader frequencies plus score breakdowns are exactly what would let a player reverse the scorer.

The tunable values the campaign engine reads. BLD §1 requires these to be data rather than literals, so every one is a campaign_settings row. Blank a field to restore its default. Two keys (rp_earned_ceiling, rp_fund_research_max) are deliberately unset — there, blank means “no bound configured”, which is not the same as 0.