Design — Politics & the Council

High-level design for the town’s political layer: the council of offices, the powers each seat pulls, how seats are won and lost, and how it all ties back to the family. Answers to ../GAME_DESIGN.md §5 (power), and builds on dynasty.md (offices are seats held by living members) and characters.md (charisma/traits set a seat’s strength). Design doc — the model and its rules, not an implementation plan.

Heavily modelled on The Guild 2’s town politics, trimmed to a single fixed settlement tier (the City) — no settlement-growth ladder. The slate below is The Guild 2’s City office roster, pruned and renamed for our game.

1. What the council is

The town is governed by a council — a small set of offices, each a seat held by one living family member. Holding a seat grants real mechanical power over the whole town, not prestige points: you set the tax everyone pays, command the guards, run the courts, banish people. The council is a weapon, the central prize of the political game (GDD §5).

It is a closed oligarchy. The general populace does not vote — only office-holders elect office-holders (§3). Power therefore cascades from the top: capture the apex and you hand-pick the seats below it. Breaking in from the outside is hard by design, and that difficulty is the political game.

Per GDD §1/§6 there is no political victory — no “you win” for taking the council. The Mayoralty is the summit of one of the three sandbox dominance fronts (economic / political / force), and a potent advantage, but it ends nothing. The only terminal state in the game is the hard loss (dynasty extinction, GDD §6).

2. The offices

A five-seat council, in a strict rank pyramid. Each office carries an application fee (gold paid to run for it — a sink), a salary (paid each period from the town treasury while seated), a set of powers, and an electorate (who votes to fill it). Powers split into active (abilities you trigger, often on a cooldown) and passive (modifiers just for holding the seat).

(Fees/salaries are The Guild 2’s City values, kept as a starting point — subject to balancing, §6.)

A clean shape falls out of the slate:

• Mayor — the impunity & patronage apex. Notably holds no economic lever; its power is being untouchable, appointing the seats below, and corruption.

• Chamberlain — economic warfare: the town’s money dials.

• Magistrate + Procurator + Sheriff — the crime engine: catch → accuse → judge → punish.

Mayor — impunity & patronage (the apex)

The seat everyone climbs toward. Its value is structural — elected by everyone below, it appoints the Chamberlain and Magistrate outright, so capturing it cascades control downward. “A good seat to steal money and mess with characters.”

• Immunity (passive) — cannot be prosecuted in the courts while in office.

• Embezzle public money (active) — siphon the town treasury into your dynasty, with a getting-caught/corruption risk.

• Banish Someone (active) — exile a rival character from the town for the duration of this Mayor’s tenure; the banished return when the Mayor changes (election / death / resignation). Extra-judicial — no trial. Ties the exile to the seat, so it gives rivals a reason to unseat the Mayor (crime.md §4).

• Order Arrest (active, shared with the Sheriff) — send the guards to seize a target character.

Chamberlain — the money office (tier 2)

The economic-warfare seat. (Named Chamberlain, not Treasurer, to avoid colliding with the dynasty Treasurer office in dynasty.md §3 — “Treasurer = your house, Chamberlain = the town”.)

• Set Sales Tax (active / policy) — the town-wide tax rate. This is the lever over the existing daily profit tax (TAX_RATE, §5).

• Set Tariffs / Market Fees (active / policy) — the town’s standing trade duty: a set-and-hold per-good surcharge the seated Chamberlain controls to squeeze rivals’ margins and favour your own goods. Distinct from the Crown tariff — a temporary, externally-imposed duty the Crown applies via a Kingdom Event (crown.md §2): the town’s dial is seat-set and persistent, the Crown’s is event-set and timed. They write the same per-good surcharge hook from two sources (cf. demand pools taking both internal and Crown pushes).

• Targeted Levy / Audit (active) — hit one rival dynasty with a special tax or audit (cooldown). The money office’s equivalent of an arrest.

Magistrate — the judge (tier 2)

The judicial seat, the middle of the legal loop.

• Trial Duties (active) — preside over trials: convict or acquit. Decides the outcomes the Procurator and Sheriff feed in.

• Set Law Severity (active / policy) — town-wide dial for how harsh punishments are; shades every conviction.

• Immunity (passive) — shielded from prosecution while seated (you can’t easily take down the one who runs the court).

• Venerability (passive) — a standing buff while seated (holding the judge’s seat makes you more important in the town’s eyes). A standing modifier, per relationships.md §3 — not a relationship buff, not its own system.

Procurator — accuser & spymaster (tier 3)

The medieval term for a public prosecutor. The “accuse” half of the crime loop, plus intelligence.

• Inquisition Trial (active) — press charges / open a case against a rival → the Magistrate then rules on it. Real crimes, or trumped-up charges at a risk.

• Command Informants (active) — run a spy/informant network: surveillance, uncovering rivals’ crimes. (Exact informant tasks deferred to crime, §6.)

Sheriff — law muscle (tier 4)

The cheapest entry seat, yet wields the town guards — a lot of muscle for a low office. The catch → hold → question end of the crime loop.

• Order Arrest (active, shared with the Mayor) — seize a target character.

• Command City Guards (active) — direct the town watch (patrols, defence, raids). Sets guard tasks; which tasks is deferred to crime (§6). Distinct from the dynasty Marshal (dynasty.md §3), which buffs your own private guards — the Sheriff points the town’s.

• Interrogate (active) — extract a confession / information from a prisoner (rolled in from The Guild 2’s Jailer).

The crime engine

Three seats chain into the central rival-fighting tool. The machinery they pull — how a charge is filed, how the Magistrate’s verdict and its corruption-backfire work, what the guards’ and informants’ tasks are, and the punishments — is designed in crime.md (the courts loop, §3b/§4); the seats and their powers below are the politics side of it:

  Sheriff            Procurator         Magistrate          Sheriff
  (catch /     ──►   (accuse:     ──►   (judge:      ──►    (enforce /
   interrogate)       Inquisition        convict /           the guards)
                      Trial)             acquit)

A dynasty that captures all three can weaponise the law against rivals — arrest, charge, convict, punish — the nastiest play in the game. The Mayor’s and Magistrate’s immunity is the counter: the people who run the courts can’t be dragged into them.

Extensibility — the slate is open-ended

This is the starting set of offices, not a closed list. As other systems are designed — crime, guilds, relationships, economy — each may add new council offices and levers into this same pyramid. (The Guild 2 had a Seneschal, Inquisitor-style social/dungeon seats we dropped for now; the social/diplomatic seat stays parked — relationships.md §6 keeps it with the future diplomacy layer, since a social seat needs levers that layer would supply.) Design new systems with “could this want a council seat?” in mind.

3. Elections

The electorate — council-only, cascading

Only office-holders vote. For each seat, the electorate is the set of seats listed in §2’s table:

• The Mayor is elected by all four other office-holders.
• Every lower seat is elected by the holders above it.

So the sitting council perpetuates itself, and control flows from the top down — hold the Mayoralty and you choose the Chamberlain and Magistrate, who in turn help choose the Procurator, and so on.

Breaking in — the entry path

If sitting office-holders elect everyone, how does a dynasty holding zero seats ever get in? Anyone may run for an open seat by paying its application fee; the sitting electors then vote among the candidates. The way in is the bottom rung: run for the Sheriff (elected by just the Mayor + Procurator — only two people to win over). The loop is:

money → candidacy + bribes → a low seat → votes in lower elections → climb.

A wealthy outsider can effectively buy their way onto the lowest rung by bribing its two electors, then start climbing.

How a vote resolves

An elector’s vote is swayed by, for each candidate:

• Charisma — the candidate member’s attribute (characters.md).
• Wealth / campaign spend — gold poured into the campaign.

• Bribes — direct payment to a specific elector; chance-based, can backfire (refusal + reputation hit).

• Relationship + standing — the elector’s relationship toward the candidate’s dynasty (warmed by kinship from marriage, soured by past slights) plus the candidate house’s standing. Both are the derived numbers of relationships.md (§5 vote weighting) — no longer a flat stub.

Cadence — terms + by-elections

• Terms — every seat is re-contested on a periodic timer, so no seat is held forever uncontested and the oligarchy can’t fully freeze.

• By-elections — a seat that empties mid-term (death, conviction, or being vacated by climbing) is contested immediately, not left open until the next term.

Because family members age and die (dynasty.md §4), mid-term vacancies arise naturally.

Climbing vacates

A member holds one council seat at a time (§5). Winning a higher seat means vacating the old one — which itself triggers a by-election. So a single successful climb can cascade into several downstream contests: churn the rest of the council competes over.

4. Losing a seat — forced removal

A seat opens on term expiry, the holder’s death, voluntary resignation, or forced removal:

• Conviction strips the office. A seat-holder without immunity (Chamberlain, Procurator, Sheriff) who is convicted/jailed via the crime loop loses their seat → by-election. The courts are thus a real political weapon: control the Magistrate/Procurator and you can legally unseat rivals.

• Immunity blocks it. The Mayor and Magistrate cannot be prosecuted, so the courts cannot touch them. This is a deliberate gap: nothing inside the council’s own machinery can remove an immune seat-holder.

The guild check (parked for guilds.md)

That immunity gap is the intended niche for the Guilds pillar (GDD §5). A guild-driven recall — a guild master pushing a council seat back to election, but only by rallying other guild masters to vote it through — becomes the check on the immune Mayor/Magistrate that the courts cannot reach. Two power structures checking each other, with immunity as the reason the guilds matter politically. This is the core Guild-vs-Council powerplay; its mechanics are deferred to guilds.md (not yet written).

5. Seats as positions — the vessel & succession tie-in

A council seat is a position held by one living family member — the same vessel pattern as the dynasty offices and building managers, now unified as the third class of position in positions.md §5 (the authoritative model; council seats differ only in being won from the town, not appointed by the dynasty). Consequences:

• The holder’s traits set the office’s strength. A charismatic Mayor sways electors and wins votes more easily; a sharp Magistrate judges more effectively. Seating who matters, not just that the seat is filled.

• One member = one council seat. A character can’t be Mayor and Magistrate; each seat needs its own body. Climbing to a new seat vacates the old (§3).

• A dynasty holds many seats via different members. This is the “capture the council” fantasy — hold the Mayor with one member, install relatives as Chamberlain and Magistrate. It makes family size the bottleneck on political power: you cannot dominate the council without the bodies to fill it (reinforcing dynasty.md: family is the hard limit).

• Council seats are exempt from the overload rule. Overload is a dynasty-internal pressure, pooling only across cabinet offices + building managers (positions.md §1); council seats — an external, contested class — do not count toward it. A member who is both dynasty Elder and city Mayor carries only the Elder seat’s overload. Winning the council is a reward, not a self-inflicted penalty.

• Candidacy = field a member + pay the fee. To run, a dynasty nominates one of its family members for an open seat and pays the application fee; the electors then vote.

• Death or conviction → by-election (§3, §4), including the Mayor — the council elects a replacement from the candidates.

6. The money loop — fees, salaries & the town treasury

The council’s gold flows through a real town treasury (a city-held balance):

   taxes (Chamberlain's rate, the daily profit levy)  ┐
   application fees (candidates paying to run)         ├─►  TOWN TREASURY  ─┐
   targeted levies / audits (Chamberlain)              ┘                    │
                                                                            │
         ┌──────────────────────────────────────────────────────────────────┘
         ▼
   office salaries (paid each period)   ·   city-owned workshop funding
   the Mayor's embezzle pool            ·

This closes a satisfying loop: dynasties pay tax → treasury → council salaries + the Mayor’s skim. It makes the Chamberlain’s tax dial self-interested (a fat treasury funds fatter salaries and a juicier embezzle) and gives the Mayor’s Embezzle something concrete to drain.

The treasury is a real mechanic, but its balance is an open tuning question (§7). If a genuinely closed loop proves too hard to keep solvent — treasury runs dry, salaries can’t be paid — we reserve the right to subsidize it (inject gold to keep it afloat). Real first; cheat only if balancing forces it.

7. ⚠ Tensions with current code

• No council, no offices. CityMemberComponent.role has a Mayor enum value (and Guard), but there is no system behind any office — no Chamberlain / Magistrate / Procurator / Sheriff roles, no powers, no election system, no terms. The entire council is greenfield.

• Tax is a hardcoded constant. CitySystem::TAX_RATE = 0.10f is fixed; the design needs it to become the Chamberlain-controlled dial.

• No town treasury balance. The daily tax (LedgerComponent / CitySystem) deducts gold from dynasties but does not accumulate it into a city treasury — it effectively vanishes as a sink. Salaries, embezzlement, and the §6 money loop require a real city gold balance (a CityComponent inventory/gold field) that does not exist yet.

• No courts / crime system. The Magistrate, Procurator, and Sheriff’s powers (trials, charges, arrests, interrogation, conviction-strips-office) all depend on a crime/justice loop that is deferred (GDD §5, future crime.md).

• No relationship system (now designed).* Venerability, bribery odds, and the relationship input to votes are now specified in relationships.md (two derived numbers); still greenfield in code, but no longer a design stub.

• Offices as vessels. Council seats being positions seated in living members — with trait-driven strength and the overload rule — is the same greenfield gap as the dynasty positions (dynasty.md §6); they share a solution.

• Guards exist, tasks don’t. CityRole::Guard exists, but “Command City Guards” has no defined task set yet (deferred to crime).

8. Open questions (next pass)

• Treasury balancing (§6) — can a real closed treasury stay solvent, or does it need subsidizing? The headline tuning risk.

• Guard & informant tasks (§2) — the courts loop they feed is now designed in crime.md (§3b); the exact guard/informant task list and what interrogation extracts remain balancing/content there.

• Guild-recall mechanics (§4) — the guild check on immune seats. Deferred to guilds.md.

• ~~Bribery & relationship (§3)~~ — settled in relationships.md: a bribe rides the relationship number (backfire = corruption isProof), formula = traits + amount + standing; votes are weighted by relationship + standing. Exact curves remain balancing there.

• Term length & election cadence numbers (§3) — how long a term runs.

• Fee / salary tuning (§2) — The Guild 2 values are a starting point only.
• Overload curve (§5) — shared open question with dynasty.md §7.

• Corruption mechanics (§2) — the embezzle / levy / audit getting-caught risk and its consequences.

• ~~A returning social seat (§2)~~ — decided: no, parked. relationships.md §6 holds the Seneschal back — a social seat needs levers, and its natural levers are the formal diplomacy (favours/pacts) that pillar deliberately parked. It returns only if that future diplomacy layer is built.