One-Party States (as shipped)
Twelve countries run as one-party states: CN, RU, DD, PL, CS, HU, RO, BG, YU, UKR, BLR and BAL. In those countries one party rules by right, other parties are tolerated or banned rather than competitive, and the head of government answers to the party rather than to the electorate. Two scalars drive everything: ruling-party confidence in the leader, and popular legitimacy. When they fall far enough the regime walks a four-stage escalation chain that can end in the country converting to a parliamentary or presidential republic, permanently. Where the command-economy flag is on, the same countries also run a planned economy with administered prices, a wage-fund throttle on cash, and a monetary overhang that feeds back into political legitimacy.
Which countries, and what runs for them#
governmentType: "onePartyState" is set on twelve entries in . The runtime value lives on the countryState collection, not in config, so a country that converts mid-game stops behaving as a one-party state on the very next turn.
Each country registers a per-turn bill-lifecycle binding in : processCNBillLifecycle, processRUBillLifecycle, processDDBillLifecycle, then PL, CS, HU, RO, BG, YU, UKR, BLR, BAL. All of them call processOnePartyBillLifecycleForCountry(countryId, now), which re-reads governmentType from countryState and returns immediately if the country is no longer one-party.
Most of these legislatures are unicameral: a bill passes the single chamber by the cast-votes majority rule and enacts with no assent delay. RU is the exception. Its Supreme Soviet is bicameral (legislature.bicameral plus upperElectionSystem), so a bill must clear both chambers and a second-chamber rejection kills it outright, because a one-party state has no override chamber. The union republics (UKR, BLR, BAL) are unicameral: their Presidium is a standing organ of the same chamber, not a second house.
What a player sees in the executive hub#
Every one-party country renders the shared OnePartyExecutiveHub at /country/[code]/executive. The per-country copy comes from , which carries bespoke surfaces for CN, RU and DD and derives a complete surface from COUNTRY_CONFIGS for the rest, so adding a one-party country needs no new hub.
- CN: Premier, "NPC Delegate", Chinese Communist Party, plus a President plaque auto-synced to the party chair, a Chairman of the NPC Standing Committee plaque (
chair_npcsc) and a CPPCC chair plaque (chair_cppcc). - RU: Premier, "Supreme Soviet Deputy", Communist Party of the Soviet Union, with a Chairman of the Presidium elected by a joint sitting.
- DD: General Secretary, "Volkskammer Deputy", Socialist Unity Party of Germany, with a Chairman of the Council of State synced to the SED chair.
The head-of-state seat is filled per headOfStateSelection: "partyChairSync" reconciles it to the ruling party's chair every turn, "legislatureAppointment" seats it through a chamber vote mirroring the premier flow.
The hub also carries a Ruling-Party Confidence panel and, in RegimeHealthTab.tsx, the regime stage, dwell counters, active decision and convention state.
The two scalars#
Ruling-party confidence is defined in . A newly installed leader starts at INITIAL_CONFIDENCE = 75, the range is clamped to MIN_CONFIDENCE = 0 .. MAX_CONFIDENCE = 95, renewal adds RENEWAL_BUMP = 5, and up to MAX_HISTORY_ENTRIES = 50 history rows are kept. Bands: 80+ secure, 65+ stable, 50+ watchful, 35+ strained, 20+ crisis, below 20 critical.
Popular legitimacy () starts at INITIAL_POPULAR_LEGITIMACY = 75 and clamps to 0..100.
Order per turn, in processCountryBills: run the bill engine, then the popular-legitimacy drift, then the ruling-party confidence drift (which reads the fresh popular value for its coupling bleed), then the regime escalation tick (which reads both post-drift values). The leader-state row is self-healed at the top of the tick via ensureLeaderStateExists, so an admin-installed premier does not silently skip every regime driver.
Purges#
Purges are simulation events, not data deletions. POST /api/admin/country/[code]/ruling-party-purge inserts one row into rulingPartyPurgeEvents with a severity, a reason and an optional target count. The route rejects any country that is not a one-party state.
Confidence deltas, from PURGE_SEVERITY_DELTA in :
| Severity | Confidence delta |
|---|---|
| minor | -2 |
| regional | -4 |
| senior | -7 |
| faction | -10 |
| extreme | -15 |
Purges also carry a kind (discipline, anticorruption, faction, ideological) written by recordPurgeEvent in . The popular driver treats anticorruption as a public benefit rather than repression. Unprocessed purges are folded into both drifts and marked processed: true only once a leader exists to absorb them, so a purge backlog inherited by the first premier is never lost or double-counted.
Escalation: four stages#
The state machine is defined in (one document per country, _id === countryId) and the pure math in . Stages: stable, discontent, crisis, internalChallenge, collapse. It is forward-only: raw scalar recovery never regresses a stage, only an explicit decision does, which stops a country from oscillating every turn around a threshold.
STAGE_THRESHOLDS:
- Stage 1 (discontent): popular <= 60, dwell 12 turns.
- Stage 2 (crisis): popular <= 35, sustained dwell 36 turns or cumulative 48 turns inside a 168-turn sliding window.
- Stage 3 (internal challenge): party confidence strictly below 15, dwell 24 turns.
- Stage 4 (collapse): popular strictly below 15, dwell 72 turns, only reachable from
internalChallenge.
Dwell counters move by one per turn in each direction, except the Stage 2 cumulative counter, which erodes by 1 every 4 recovery turns instead of resetting: a regime that escaped Stage 2 once falls back in faster. The Stage 4 counter is frozen entirely while a constitutional convention is in progress. computeNextStage evaluates highest stage first, so a country meeting several entry conditions at once jumps straight to the worst.
Decisions#
Each stage transition offers the leader one decision (). Only one decision is active at a time; further ones sit in a passive queue and promote on resolution. DECISION_TIMEOUT_TURNS = 48: if the leader does nothing, the stage's default option fires automatically.
Stage 1, stage1.addressDiscontent (default ignore):
- Acknowledge and promise reform: -2 confidence, +3 popular, and sets
pendingReformDiscountat 0.5x for the current turn, halving the next reform's intra-party cost. - Crack down: -1 popular, one
minordisciplinepurge, popular decay multiplier 0.5x for 24 turns. - Ignore: popular decay 1.25x for 24 turns, plus 2 extra Stage 1 dwell.
Stage 2, stage2.respondToUnrest (default ignore):
- Open dialogue: -3 confidence, +6 popular, Stage 2 dwell decay doubled for 48 turns.
- Selective concession: legalize a chosen banned party (
payload.bannedPartyId), -5 confidence, +10 popular, and a country-history entry tagged to the beneficiary party. - Martial law: -8 popular plus an
extremedisciplinepurge, in exchange for popular decay 0x (frozen) for 72 turns and a Stage 3 block for the same 72 turns. If the Stage 3 dwell counter is already past threshold when the block lapses, this only bought time. - Ignore: popular decay 1.5x for 48 turns.
Stage 3, stage3.respondToFactionSplit (default ignore): negotiate power-sharing (-4 confidence, +8 popular), purge the defectors, concede leadership (+5 popular), or ignore.
Stage 4, stage4.faceCollapse (default acceptPeacefully):
- Resist: -25 confidence, -20 popular, an
extremefactionpurge, bans every approved party, and delays conversion by 48 turns. If popular legitimacy is still below 15 when the delay ends, the post-conversion legacy seat reservation is halved. - Accept peacefully: no direct cost; marks the regime for conversion on the next tick.
Faction split#
Entering Stage 3 fires fireFactionSplit (). It picks max(3, ceil(15% of ruling-party officials)) defectors by divergence score, spawns a new approved party under the country's factionDefectionName (CN: "Democratic Faction of the CCP"; RU: "Reformist Faction of the CPSU") seeded from the defectors' own axis positions, not the ruling party's, and reassigns each defector's party field. Only character-held seats defect; NPP-held seats have a null characterId and are skipped. The split no-ops when the country has no factionDefectionName, no ruling party, or no officials.
Faction divergence scoring#
Who actually leaves is decided in , two pure (DB-free) kernels:
scoreDivergence(official, line), Manhattan distance between an official's own (economic, social) axis position and the ruling party's line, normalized to 0..1 (axes run roughly -5..+5, so max distance is 20). An official with no recorded axis position scores 0, since absence of a position is not evidence of disloyalty. Treating a null as maximally distant would purge NPC-imported officials first.applyCaucusCohesion(scored, weight = CAUCUS_COHESION_WEIGHT), blends each official's individual score toward their caucus's mean score atCAUCUS_COHESION_WEIGHT = 0.6, so a faction leaves together rather than as scattered individuals who happen to cross the divergence line alone. Officials in no caucus keep their unmodified individual score.
The max(3, ceil(15%)) defector count is then taken by highest cohesion-adjusted divergence, ties broken by input order. factionCentreOfGravity averages the defectors' own axis positions (falling back to the ruling party's line only when no defector has a recorded position) to seed the new party's ideology, so the breakaway faction is ideologically distinct from the party it left rather than a renamed copy of it.
Liberalization reforms#
Five actions in are available to the ruling-party leader at any time while the country is still one-party. Each trades intra-party confidence for public standing, and most leave a per-turn popular boost with a hard expiry. Cooldowns live on countryState ().
| Action | Confidence | Popular | Per-turn boost | Cooldown |
|---|---|---|---|---|
| Legalize a banned party | -6 | +8 | +0.2 for 120 t | 168 turns, per party id |
| Reduce vote multipliers | -4 | +4 | +0.15 for 96 t | 240 turns |
| Honest by-election | -3 | +5 | none | 96 turns |
| Anti-corruption purge | -2 | +3 | none | 72 turns |
| Constitutional amendment | -8 | +12 | +0.3 for 240 t | one time only |
The anti-corruption purge records a minor anticorruption purge event rather than deleting anything. The constitutional amendment is permanent in a second way: it writes renewalBumpOverride: 2, so every future leader renewal in that country grants +2 confidence instead of +5. Stage 1's acknowledge option halves the next action's confidence cost, and the discount is consumed on use.
Elections, vote multipliers and NPPs#
Elections still run in one-party states, but they are weighted. resolveRegimeMultiplier in reads countryState.opsVoteMultipliers, defaulting to DEFAULT_OPS_VOTE_MULTIPLIERS: ruling 3.0, approved 0.375, independent 0.0, banned 0.0. Independents (no party document) are weighted like banned parties.
The same module holds the hard gates:
canFormGovernment: ruling party only.canTriggerNoConfidence: ruling party only.canCollapseGovernment: always false, so the generic coalition-loss collapse path never fires.canFieldExecutiveCandidate: premier and president races are ruling-party only; other offices are open to anything not banned.canFieldLegislativeCandidate: ruling or approved parties only, which blocks independents at filing.canInviteToCoalition: ruling party only.
Reduce vote multipliers steps the ruling weight down one rung of RULING_MULTIPLIER_LADDER in : 3.0, 2.0, 1.5, 1.0, with 1.0 as the floor. Approved, independent and banned weights are untouched: the reform loosens ruling privilege, it does not enfranchise banned parties. Honest by-election writes a one-shot pendingHonestByElection: { atMultiplier: 1.0 }, which reads, applies as a flat override on every regime multiplier for that election, and then clears.
NPPs (the game's AI-run politicians) file, hold seats and are enriched with their own policy positions exactly as elsewhere. Two one-party specifics: an NPP-held seat carries a null characterId and therefore cannot defect in a faction split, and a banned party's NPP seat-holders have currentOffice cleared like any other holder. Most Eastern-bloc satellites are largely NPP-governed, so the NPP brain is what actually operates the planned economy there, using NPP_DEFAULT_REFORMISM = 0 and NPP_DEFAULT_INTERNAL_REPRESSION = 0.5 when no player directive exists.
Banning and unbanning parties#
processBanPartyEffects () makes a party inert: it deletes every electedOfficials document for that party, clears currentOffice on the holders, clears autoRunForReelection so the auto-re-entry pass does not silently re-file them next turn, and flips the party document to regimeStatus: "banned" with bannedAt, bannedReason and bannedAtTurn. There is no by-election machinery in AHD, so vacated seats stay empty until the next regular cycle for that office. Unbanning flips the status back to approved and clears the audit fields; it does not restore any seats.
Constitutional convention: the voluntary exit#
runs three phases.
announced: player-initiated. SetsconventionInProgress: true(freezing the Stage 4 dwell counter), parks a draft deadlineDRAFT_PHASE_TURNS = 48turns out, and applies +15 popular / -10 confidence. Blocked if the country is already incollapse(that would race the forced path) or if a convention is already running. Defaults come from config:legacyReservationDefault(20 for CN and RU) andelectionDelayDefault.draft: locks in atargetSystemfromcollapseTargetAllowlist(CN and RU:parliamentaryRepublicorpresidential), alegacyReservationin 0..35, and anelectionDelayTurnsof 12, 24 or 48. If the deadline passes, the tick advances to ratification anyway.ratification: atdraftDeadlineTurn + electionDelayTurnsthe tick fires the conversion withpath: "voluntary".
Conversion: one way out, and it is one way#
triggerSystemConversion in does four things: sets the new governmentType, nulls opsVoteMultipliers and sets hasLeaderConfidenceModel: false (which is what switches off every intra-party and popular driver); clears regimeStatus on every party in the country (clearAllRegimeStatusForCountry in , which sets the field to null rather than unsetting it, so legacy readers still find a sentinel instead of undefined); writes a regime_escalation history event with subtype: "conversion"; and calls bootstrapNewSystem to park the snap-election marker. No rows are deleted, so the history survives.
The forced path, checkForcedConversion, fires on either of two conditions: Stage 4's accept-peacefully set conversionPendingAtTurn and that turn has arrived, or the country has sat in collapse with no convention and any stage4Delay window has elapsed. Forced constants: FORCED_LEGACY_RESERVATION = 5 seats (halved to 3 when the resist delay expired with popular still below 15), FORCED_VOTE_SHARE_PENALTY = -0.2 on the former ruling party's first post-conversion election, and FORCED_ELECTION_DELAY_TURNS = 12. Both paths are idempotent: the pending marker and the delay marker are cleared once consumed.
bootstrapNewSystem writes pendingPostConversionElection on countryState with the target turn, the legacy reservation, the former ruling party id (captured before the flip nulls it), the forced penalty and the path. The election engine consumes it.
Command economy#
Everything below is gated behind the commandEconomyEnabled GameConfig flag. Fresh bootstrap ties the flag to the selected era: Eastern Bloc presets enable it and other presets disable it. The static seed default is fail-safe enabled, but bootstrap makes the final preset-specific write before budget seeding.
The regime is a dial, not a switch. marketizationLevel runs 0 (fully command) to MARKET_LEVEL = 100, with bands at COMMAND_CEILING = 30 and DUAL_TRACK_CEILING = 70 (). MARKETIZATION_SCHEDULE gives each country a trajectory: RU, UKR, BLR and BAL sit at level 10 through 1991; CN moves 10 (through 1978) to 50 (through 1992) to 85 (through 2018) and then to market; DD is at 10 through 1990, PL at 12 through 1989, HU at 15 through 1989.
Offices#
Three player seats operate the planned economy, mapped in : RU and DD use chairman_of_gosplan plus gosbank_liaison; CN uses vice_premier plus pboc_governor. SOE directors are appointed per enterprise onto SoeState.directorId, not to a fixed cabinet seat. Authorization is in : the Gosbank chair or the head of government may drive state credit; the Gosplan planner or the head of government may drive the plan and appoint SOE directors; only the SOE's own director drives that SOE; and internal repression is set by the head of government or the Gosbank chair, never by the planner alone (it is a security call, not a planning one). Lever ranges are clamped server side: credit aggressiveness 0..1, budget softness 0..1, labour-versus-quality 0..1, internal repression 0..1.
Administered prices#
holds consumer prices at the era base price marked up by a turnover tax, DEFAULT_TURNOVER_MARKUP = 0.12, with no supply-and-demand response at all. That is the point: the unmet demand at the held price is the shortage, reported by demandSupplyGapPct. The turnover wedge times units actually supplied is the state's main indirect revenue. In a dual-track country dualTrackPrice blends administered and market prices by plannedShare. This branch runs only on the country-scoped national-price leg, never on the shared global leg, because an administered price leaking into the world price is the naira 100x incident class of bug.
Two-circuit money#
models the Soviet split between enterprise non-cash money and household cash at the macro level: a wage-fund ceiling. wageFundConstrainedGrowth pulls nominal wage growth down toward realGoodsGrowth + WAGE_FUND_SLACK_PP (2 percentage points), by the planned share, and never raises it. A tight wage fund is the primary brake on forced saving.
Overhang, shortage and the second economy#
accumulates the monetary overhang: the excess of nominal income over real goods, scaled by plannedShare, divided by TURNS_PER_YEAR (48) because the growth inputs are annual, plus any monetized directed credit (CREDIT_OVERHANG_SCALE = 1.0), less what the second economy absorbs. It persists at OVERHANG_DECAY = 0.99 per turn (roughly a two-year horizon) and is capped at OVERHANG_CAP = 100. The black-market premium is capped at MAX_BLACK_MARKET_PREMIUM = 2.0 (plus 200 percent over official) and the second economy at MAX_SECOND_ECONOMY_SHARE = 0.6 of activity, tracking its target at 0.25 per turn. The readouts land on FederalBudget.economicFactors as monetaryOverhang, shortageIndex, blackMarketPremium and secondEconomyShare, and the dashboard presenter () turns them into the marketization gauge alongside SOE plan fulfillment and the government's policy stance.
Politics and economics meet through the drift collectors in : shortages and economic performance feed popular legitimacy, repression intensity feeds it the other way, and the resulting number is what walks the regime up the escalation chain.