Fictional and notional. An invented 2029 crisis. Odds and effects are illustrative game design, not predictions.

Choose your capital

2029. Beijing is turning up the pressure on Taiwan. Lead one of four capitals for up to eight months. Each month you pick a posture, up to four moves and where your forces go, while the other three do the same at once. Every leader has a hidden type the others must read from what it does. First game? Press Learn to play, or keep Staff advice on.

Opponents difficulty
Staff advice

On adds an Ask your staff button each month that fills in a sensible plan for your goals, which you can change before you commit. Recommended if you are new to the game.

How the game works

The basics

  1. Each month, three steps. Pick a posture (how hard you lean), up to four moves (each shows its odds and cost), and orders for your forces. Then End month. All four capitals reveal at once.
  2. Read the others. Every leader is secretly resolute, cautious or opportunist. Costly moves reveal the most, about them and about you.
  3. Mind the ladder. The crisis climbs from gray-zone pressure to war. Every rung raises the risk for everyone, and nuclear use ends the game. Pressing Beijing invites it to answer in kind; it backs down only when the coalition clearly has the upper hand.
  4. Live within your means. Lift, fuel, munitions and readiness limit what you can do; whatever you cannot afford is greyed out.
  5. You are scored on your own goals, the weights you set at the start, and compared with the computer playing your seat.

Stuck? With Staff advice on (set on the start screen), Ask your staff fills in a sensible plan you can change. Everything else (fog of war, elections, allies’ bases, the economic shock, the peace forum, landings) is explained on the move or the map where it comes up, and in full below.

Full rules
  • Each month you pick a posture (stand down to escalate; the United States and China can also make a nuclear signal from Blockade up) and up to four strategic moves across seven lines: diplomatic, information, military, economic, financial, intelligence and law enforcement. Each move asks follow-up questions: which route, which ports, which sector, how much to commit. Some moves unlock as the crisis climbs. Opportunity moves (● on the tab) are on the menu only while a condition holds, such as markets falling or Taiwan divided; the badge on the move names the condition. Once a game moves (dispersing and hardening, calling up the reserves, asset freezes, Japan’s survival declaration, evacuating the southwest islands) have effects that last, so after the month you carry one out it is marked USED and cannot be chosen again. All four capitals reveal at once.
  • Gray zone. The bottom two rungs of the ladder are coercion below military action: coast guard “law enforcement” patrols and quarantines, maritime militia swarms, deniable undersea-cable cuts (which keep costing Taiwan until repaired) and cyber intrusions. They build pressure without firing a shot; only the quarantine and the militia swarm climb the ladder, and only to Coercion. Each has an answer: coast guard escorts and expulsions, cable patrols and repairs, cyber defense teams.
  • Logistics. Four resources limit what you can do. Lift (sealift and airlift) refills each month and pays for moving formations. Fuel is a slowly refilling stock burned by every formation move and every month a formation sits at sea in Contest or Attack. Munitions are a stockpile, slow to rebuild, spent by strikes, blockade enforcement, escorts and combat: China’s is largest, Taiwan’s and Japan’s smallest. Readiness belongs to each formation: it drains while the formation stays forward (faster in Attack and in combat) and recovers in the Rear, and low readiness cuts its strength. Some moves cost resources (shown on each move) and some add them (a U.S. surge adds Lift, Taiwan’s call-up adds Lift and readiness). The logistics table shows what you have, what your choices commit, what is left and what comes back next month; the list below it greys out whatever you can no longer afford. A blockade cuts Taiwan’s fuel refill and stops its munitions resupply.
  • Formations. Each capital has a handful of named formations (real public unit types, notional strengths, no real order of battle) in four sea areas (North, Strait, South, East) and its Rear. Give each one an order a month: move it (1 Lift to a neighbouring area, 2 to or from the Rear, amphibious forces 1½×, plus 1 fuel), or aim a strike force at a sea area. If you play the United States, formations leaving the Rear (the continental United States, Hawaii) take two months to arrive, one with a surge, so decide early; the computer’s United States moves on the old timetable (next month, this month with a surge). Ground forces cannot cross the Strait. Air formations add half their strength to neighbouring areas once the shooting starts; strike forces count only where aimed, and only then. Set each area to Defend, Contest or Attack. Holding areas matters: a blockade needs China to hold the Strait and a flank, escorts need their route. Below Limited strikes nobody fires; after that, areas where both sides contest fight every month.
  • Fog of war. You see your own and your partners’ forces exactly. Rival forces show as ranges (“about 6–9”): exact in a sea area where your own formations are, or everywhere for a month after your surveillance or early-warning move pays off; a close read (narrow range, types known) next to your forces or where a partner’s are; a rough read (wide range, some formations “unidentified”) elsewhere, including the Rear. China’s feints and Taiwan’s decoys shift or widen ranges for a month; surveillance sees through them. Releasing intelligence to the public burns your sources: a month of blurrier sight. The key sits under the map, and the computer capitals see by the same rules.
  • Taiwan’s coast. The island is four landing sectors (northwest, the approach to the capital; central-west; southwest; east) plus an inland reserve. A landing is possible only once the crisis reaches Limited strikes and needs China to hold the sea in front of the chosen coast with an amphibious formation there, plus Lift. China picks a landing emphasis; Taiwan picks where its reserve waits and the United States where its fires go, without seeing China’s choice. Region level only: no beaches, no real force numbers.
  • Politics at home. A notional election calendar (invented for play): Taiwan holds legislative by-elections in June and Japan a snap general election in July. In an election month, home-support costs of postures and escalatory or accommodating moves are 1.5×, and the result can change the government’s hand for the rest of the game. Taiwan’s legislature votes on budget-type moves (emergency arms purchases, the reserve call-up, reinforcing the outlying islands): below 55 home support it slow-walks them (−15), below 40 it refuses, and once the shooting starts it rallies. Japan’s forces cannot fight (from Limited strikes up they hold to Defend, and cannot escort under fire) until Tokyo declares a survival-threatening situation. In Washington, low home support stalls aid and surges, Congress will not back strikes on the mainland below 45, and after two months of fighting the War Powers clock weighs on U.S. strikes. Each constraint shows as a small chip on the move it affects, with the reason.
  • Allies and bases. U.S. strikes flown from Japan (Okinawa) or the Philippines, and escorts by the southern route, need the host’s consent each time. The chip on the move shows the chance: Japan’s basing decision (opening the bases or limiting them) sets the starting point, and the host’s home support, the coalition’s cohesion, the fear of nuclear escalation and whether it has already been struck move it. A refusal means the move is not carried out (and is refunded). Below a cohesion of 45, partners’ aims diverge: consent comes harder and rallying allies gets −10.
  • Economic shock. The shock track is a market and shipping-insurance index. Blockades, quarantines, sanctions, mineral bans and cable cuts push it up, and a blockade, cut cables, sanctions or a shooting war keep pushing it up every month. Each month it erodes every economy and, above 20, costs every capital home support, China included. It eases at Coercion or below.
  • Peace forum. The month after a capital’s offer of talks succeeds, it may (once a game) call a forum to end the conflict, addressed to a rival. The rival weighs four things: its resolve (type and home support), its anger at the caller (recent strikes and losses, mainland strikes most), its opportunity costs (economic damage, the shock, losses, fuel and munitions used, sanctions or blockade), and what it expects from fighting on (the force balance, Taiwan’s position, any landing) against how limited it thinks the caller’s aims are. Accepted: the ladder steps down a rung and a ceasefire holds the next month (escalatory moves cost more; breaking it costs credibility and angers the other side); if it holds, the crisis ends in a settlement. Declined: the caller loses a little credibility. If a computer capital calls a forum with you, you choose, seeing the four factors.
  • Every move shows its odds: the chance of success this month, worked out as a base chance plus the listed factors (follow-ups, who holds the route, fatigue, what others are doing), kept between 5% and 95%. One roll from 0 to 100 decides it: below the odds is a success, within 20 points above is a partial result (half the effect), anything higher a failure. The odds you see assume the others repeat last month; they are worked out again with what everyone actually did when the month resolves. Tap or hover the i beside the % for the breakdown. Military moves need the crisis to have reached a certain rung.
  • The ladder runs from Pressure to Nuclear use. It climbs when escalatory moves succeed, at most two rungs a month, and eases when capitals step back.
  • Hidden types. Resolute leaders pay at home for backing down; cautious ones pay for escalating; opportunists escalate when a rival looks weak. The computer updates its beliefs about you by Bayes' rule from what you do, and costly moves say the most.
  • Nuclear use is never a menu choice. Once the shooting starts, a rising nuclear shadow makes it a risk each month.
  • Starting situations. Gray-zone pressure (the standard start, March), a coast guard quarantine already under way (March), an energy blockade in force with a second U.S. carrier on its way (April, seven months), China’s seizure of the Kinmen island group (April, seven months), or a crisis after a bitterly disputed election in Taipei with a hostile legislature (March). Each sets its own rung, tracks, resources and a few formations; all notional.
  • Hidden traits (optional). Say yes before you start and each computer leader gets one hidden trait, drawn at random with odds that fit the capital: risk-taking (likes bold moves, wants to fight on), face-saving (hates backing down, slow to accept a forum), impatient (discounts later payoffs, keener on a forum as the months pass) or prudent (careful with stocks and escalation, open to a way out). Beijing, which chose to start the crisis, is never prudent and leans face-saving or risk-taking; Taipei and Tokyo lean prudent or face-saving; Washington could be any. Traits change the computer’s choices and add a term to the peace-forum equation; they are revealed only at the end.
  • After each month a one-sentence summary under the country boxes says what happened, using only that month’s results and public tracks. At the end, an after-action replay shows every month’s moves, how each capital’s type was read, any traits, and the computer’s main reason for its choices; a nuclear review shows what pushed the nuclear shadow, month by month, and when it crossed 65.
  • Beijing steps back only against a strong coalition. China’s face-saving pause ends the game as a climb-down only when the ladder is at Blockade or below, Taiwan’s position is 60 or more, and the coalition has the upper hand: at least two of Taiwan’s position at 70 or more, coalition cohesion at 70 or more, China’s home support below 45, and the coalition holding the Strait or the South. Short of that, a pause is only a pause. And until the coalition has the upper hand, Beijing tends to answer escalation in kind: the more escalatory postures and moves Washington, Taipei and Tokyo made last month, the likelier it is to push back.
  • Endings: a settlement after two calm months of talks, or after a peace forum’s ceasefire holds; Beijing stepping back (above); Taiwan forced to terms; nuclear use; or October.

All numbers are illustrative game design. In 1,000 test games from the standard start (gray-zone pressure, no traits) with the computer in every seat on Normal: 11% ended with Taiwan forced to terms, 15% with Beijing stepping back, 7% in a settlement, 11% in nuclear use, and 56% were unresolved in October. 62% saw a blockade or worse and 25% major war or nuclear use. A peace forum was called in 29% of games, and China attempted a landing in 14%. With hidden traits on, Beijing steps back in only 5%. The other starting situations change these (tables in the method notes): a quarantine start sees the most nuclear use (about 14%); blockade and Kinmen starts end unresolved most often (70–76%) and see the least major war (13–18%).