Build a place
Square building lots, an angled 3D camera and readable toy models. Roads, walls, fields and waterways form a continuous settlement. The visual direction carries the family resemblance of ComboThrone: Puzzle Quest.
New game concept · Historical realm strategy
ComboThrone:
Civilization
Start in a historic capital. Give your pops lasting livelihoods. Build a kingdom whose farms, workshops and defenses become stronger together.
Concept / awaiting loop validationDaily turnsNative iOS direction
A design proposal, not a playable build. Features, realm rosters and numbers below are hypotheses to test. Updated 20 October 2026.
01 / The promise
A growing realm.
A workforce with roots.
A turn-based historical settlement builder about making a capital productive, resilient and worth expanding. The pleasure comes from seeing a small community become a functioning kingdom, and understanding why its interconnected districts work.
Make it work together
Staff buildings, supply their inputs and research better methods. Combos create food security, skilled production, warning networks and defensible districts. Their value appears in the kingdom’s capabilities.
Grow beyond the capital
Welcome more pops and establish named settlements. Scale through infrastructure, institutions and standing orders. Persistent livelihoods keep growth from becoming a daily staffing chore.
Starting point: a native app shell, realm presentation and existing idle-game foundations live in the ClickerThrone repository. Civilization’s daily economy, research, population growth, combos and combat remain proposed work. Separate lab experiments provide evidence to inspect; they are not an integrated Civilization game.
02 / The hypothesis
Will preparing a living kingdom
make each turn worth taking?
We are testing whether players enjoy choosing between growth, research and preparation, then seeing a deliberately assembled economic or defensive combo solve the problem they anticipated.
- Read the dayInspect food reserves, production shortfalls, vacancies and the approaching season or threat.
- Choose a priorityBuild, improve, research or reinforce. Make a few consequential changes while standing orders continue.
- Preview the resultSee staffing, inputs, consumption, combo contributions and predicted shortages before committing.
- Advance one dayResolve work, deliveries, construction, study, upkeep and any defensive encounter in defined phases.
- Learn and expandRead the outcome, repair weaknesses and invest the surplus. Keep established jobs unless there is a reason to change them.
A meaningful turn, in miniature
A raid is approaching. Your mill and bakery already feed the capital. Researching preservation could free future harvest surplus for a standing garrison, but the same tools could equip a repair crew today. Your tower covers the road; the unfinished gate is still vulnerable.
Staffed farms→Mill + bakery→Food reserve→Supplied garrison→Defended gate
Illustrative dependency chain. A combo adds an explicit benefit beyond ordinary conversion, such as coordinated tower coverage or improved reserve preservation. Values and timing require testing.The loop succeeds when players can explain their plan.
“I chose this research and staffed these buildings because that combination lets my capital survive the next threat.” More output alone does not validate the design.
03 / Systems that serve the loop
Depth in the economy.
Continuity in the workforce.
Pops keep their jobs
Permanent assignments, protected specialists and explicit building crew levels. New arrivals fill approved vacancies or enter a reserve pool. Seasonal plans remember previous roles. District controls manage quotas as the realm grows.
Professional defenders and reserve builders reduce emergency reshuffling. Temporary militia retain their civilian assignments for an orderly return.
Growth has practical requirements
More pops need food, housing, water, services and delivery capacity. Local congestion should have visible infrastructure solutions. Population size does not apply a global threat multiplier.
Hundreds and eventually thousands of pops are scale targets to test. Population groups drive the simulation; representative townspeople provide life on screen.
Resources are stocks and commitments
Begin with food, timber, stone and tools, then introduce intermediates where they create a real decision. Stores, upkeep and repeat production orders reveal whether a district is sustainable.
Track actual output and days of cover. Explain whether a building lacks workers, inputs, delivery capacity or demand. Add transport complexity gradually.
Research creates new capabilities
A visible core tree with optional realm branches. Staffed institutions spend days and materials on preservation, agricultural methods, mills, masonry, specialist training and defensive logistics.
Research competes with immediate production and safety. Institutional improvements also unlock better management as the population expands.
Combos are functioning relationships
A supplied production district, a training-and-workshop partnership, or towers linked by a warning network. Each combo has visible conditions, inputs, limits and a benefit the player can explain.
Buildings remain useful independently. Empty support buildings should not become mandatory multiplier decorations. Resolution must be deterministic and prevent circular resource creation.
Defense belongs to the settlement
Threats approach along real terrain and waterways. Walls shape access, towers cover approaches, and gates, supplies and repair routes create vulnerabilities.
Defenders consume food and equipment. A raid can burn an exposed farm or interrupt imports even if the city holds. Tactical phases resolve within a day; each arrow exchange does not advance the calendar.
One turn = one day. Quiet days can advance under standing orders.
Stop at player-selected warnings, research completions, construction decisions and attack deadlines. Test this carefully: a year contains 365 daily turns, so quiet-day advancement must preserve control without demanding 365 staffing reviews. Early growth comes mainly from migration; childbirth does not produce adult workers in a few days.
04 / A historical place to begin
Choose a realm.
Begin at its seat of power.
Realm candidates should differ through terrain, institutions, production options, architecture and threats. Select a period for each scenario before fixing its capital, roster or technology. These are design directions, not researched final rules.
| Realm | Starting seat | Economic and defensive character to test |
|---|---|---|
| Carolingian Franks | Aachen as a royal seat | Estates, institutional learning, inland supply and frontier defense. Confirm the chosen period’s political geography. |
| Medieval England | London for a selected period | Grain and wool choices, markets, walled districts and ranged defenses. Agricultural security competes with trade and military spending. |
| Venetian Republic | Venice | Lagoon access, imported supplies, shipbuilding and harbor defenses. Waterways change the shape of an effective capital. |
| Ming | Nanjing or Beijing, according to date | Irrigation, granaries, administrative institutions and defensive logistics. Capital, construction and technology follow the selected era. |
A city list with a sense of place
Curated realm-specific names: capital first, then a nonrepeating sequence with optional regional weighting and player renaming. Keep historical naming periods and transliteration consistent. Founding a new city should feel like the realm extending its reach.
Crises with historical context
Distinguish dated, sourced events from plausible generated floods, harvest failures, raids and blockades. Documented English harvest failures in 1315–17 offer one scenario candidate. Exact event dates appear only when supported by evidence; warnings reflect what the inhabitants could reasonably know.
Foundation experiment / Before production
Can one foundation
serve two different games?
Test a ComboThrone monorepo and common isometric presentation before substantial Civilization implementation. This is a planned, reversible experiment, not a committed migration. Paper rules and small gameplay models can continue while the foundation is evaluated.
The hypothesis
Shared assets, camera controls and placement rendering should improve Puzzle Quest and Civilization without coupling their rules, changing existing saves or constraining a growing settlement. Repository consolidation and runtime sharing are separate decisions.
| Step | Work to test | Evidence required |
|---|---|---|
| 1 · Capture the baseline | Identify Puzzle Quest’s active target and asset pipeline. Record source revisions, app identity, saves, release configuration, gameplay traces, captures, build commands and target-device performance. | Reproducible gameplay and visual references, usable save fixtures and an intact original release path. |
| 2 · Stage a trial monorepo | Use an isolated branch or candidate repository with separate game apps, shared packages, assets and asset tools. Preserve provenance/history. Keep unrelated ClickerThrone and Throne Supremacy work in their existing home. | Puzzle Quest builds and behaves unchanged after relocation, including reading existing saves. Original repositories remain canonical; no new World app is needed for this test. |
| 3 · Extract a small common foundation | Trial model sources, materials, model/card exports, camera controls, picking, ghosts, overlays, loading and caching. Add neutral spatial primitives only where both consumers need them. | Economy, puzzle rules, staffing, research, combat, progression and save schemas stay game-owned. Civilization does not inherit a compact-board limit. |
| 4 · Prove a second consumer | Render a disposable Civilization district with staffed-building markers, roads, walls, towers and representative pops through the trial foundation. | A shared model or camera improvement benefits both games. A Civilization staffing-rule change leaves Puzzle Quest gameplay unchanged. This scene is presentation evidence, not a completed game. |
| 5 · Measure compatibility and scale | Repeat Puzzle Quest’s baseline checks. Test expanding Civilization maps and 100/1,000-pop scenarios using population groups and representative agents. Rehearse independent builds, releases and rollback. | Record memory, loading, day resolution, rendering and input responsiveness on named devices. Compare against baselines and budgets set before the review. |
| 6 · Decide, then continue the slices | Review regressions, maintenance costs, package boundaries and scale evidence. Choose a monorepo, separate repos with versioned sharing, or independent foundations. | Only a successful review makes the trial canonical. Continue staffing, resources, research/combos and defense on the selected base. |
Pass gate: useful sharing, independent games
- Puzzle Quest retains its identity, saves, controls, gameplay outcomes and release path. Any visual change gets its own review.
- Both apps use a real shared asset path and presentation component without importing each other’s rules.
- Target-device budgets hold at the tested scale; shared abstractions do not force gameplay compromises.
- Each app can use a known-good package version or revision and release without forcing the other app’s release.
- A rollback rehearsal succeeds. Review the experiment before broader extraction.
Safe fallback: separate remains viable
Keep original repositories, remote branches, release configuration and player saves untouched during the trial. Record known-good revisions and isolate trial save data. No repository deletion, canonical rename or production-save migration is part of the experiment.
If consolidation fails, resume Puzzle Quest from its baseline and continue Civilization in its current project or a separately chosen repository. Port back only reviewed work that stands on its own.
If sharing succeeds but co-location does not, publish a versioned package or asset bundle with pinned consumers and no sibling-path dependencies. If shared rendering also fails, keep independent renderers and use provenance-tracked asset exports where useful.
Stop or narrow the experiment if it causes gameplay/save regressions, unacceptable scale costs, excessive abstraction or coordinated-release requirements. A failed trial must not block Civilization’s gameplay prototypes or disrupt Puzzle Quest’s release.
05 / Prove the game in slices
Each slice answers
one question about the loop.
Start with one capital and one researched era. Each slice carries forward only what the next test needs. Prototype sizes and success thresholds below are proposed targets; no results have been collected. The foundation experiment informs production architecture; early gameplay models need not wait for a repository decision.
00 · Model the decision Can growth, study and safety compete?
Build: a paper or headless model with a handful of buildings, persistent workers, food, tools and one visible deadline. Compare at least two viable plans on the same starting state.
Player decision: invest scarce labor and materials in expansion, research or preparation.
Gate: reject a dominant “always build this” strategy and avoid mandatory daily job cycling. The player can describe the opportunity cost before adding presentation.
01 · Settle and staff Does an ordinary day feel legible?
Build: miniature square lots, homes, a food source, storage, a workshop, roughly 12 starting pops and one-day advancement. Assignments survive turns and save/resume.
Player decision: which vacancies to fill and which building to establish next.
Gate: target four of five first-time testers correctly explaining a production shortfall and fixing it without help. Previewed results match actual daily outcomes.
02 · Sustain growth Can a stable workforce support more pops?
Build: a grain–mill–bakery chain, housing, reserves, repeat orders and migrating arrivals. Add a forecast seasonal shortfall.
Player decision: grow now, stockpile or add capacity before the shortage.
Gate: at least two viable responses. Target 75% of established assignments remaining unchanged on ordinary non-emergency days; investigate every incentive to repeatedly reshuffle specialists.
03 · Research a useful combo Does a breakthrough change the plan?
Build: a short research branch and one explicit district benefit, such as preservation improving stored-food security. Forecast the contribution and its inputs.
Player decision: sacrifice current output to unlock a capability before a known threat.
Gate: testers deliberately choose different research or layouts to solve the same problem and can identify the combo’s practical benefit. A higher number alone does not count.
04 · Hold the gate Is defense an extension of the economy?
Build: one approach, walls, a gate, two towers, a standing garrison, a repair crew and a small raid. Resolve bounded tactical phases within a day.
Player decision: coverage, equipment, repairs and supplies, with exposed farms at risk.
Gate: show at least two useful defensive plans. Players understand a breach or successful hold. Winning must not require relocating the civilian workforce every combat phase.
05 · Play the complete seasonal arc Does preparation create a reason to return?
Build: connect growth, research and defense across roughly 30–60 daily turns, with a historically plausible seasonal crisis, raid warning, quiet-day stops and a clear scenario objective.
Player decision: commit to a strategy, adapt where needed and live with the consequences.
Gate: players can explain their outcome and voluntarily want another attempt with a different plan. Revisit pacing if too many days offer no meaningful consequence or decision.
06 · Stress the growing realm Does scale add strategy rather than chores?
Build: test population groups at 100 and 1,000 pops, district quotas, protected jobs, reserve labor and save/resume. Use representative visual agents.
Player decision: allocate capacity and infrastructure at district scale.
Gate: record meaningful decisions, corrective staffing actions, day-resolution time, memory and rendering cost on target devices. Establish budgets from measurements. Growth must not require population-proportional manual job changes.
The production gate
Advance only when the connected slice shows understandable choices, distinct viable plans, useful combos, manageable staffing and interest in another run. If it fails, revise the economy or pacing and retest before expanding the realm roster.
06 / If the loop works
From a capital district
to a complete historical realm.
This is a sequence of evidence gates, not a promised release calendar. Production scope follows the successful prototype; existing sibling gameplay is not a shortcut around validation.
Establish Civilization’s own foundation
Apply the recorded outcome of the monorepo experiment: adopt the proven common presentation or take the separate-repository fallback. Give Civilization a dedicated deterministic simulation, versioned save format, seeded scenarios and a documented daily resolution order. Separate its rules from the inherited idle economy.
Exit: documented architecture decision, reproducible outcomes, safe saves and correct previews; no accidental changes to sibling games. A rejected monorepo must not block implementation.
Finish one capital campaign
Complete one researched realm/era, its building set, core research tree, population needs, economic combos, seasonal calendar and bounded raids. Add a learn-by-doing opening, clear objectives, difficulty choices and readable outcome reports.
Exit: a complete start-to-finish campaign with multiple viable approaches and no required daily staffing churn.
Make expansion sustainable
Add named satellite settlements, regional specialization, supply connections, administrative capacity and district management. Introduce professional defense and logistics where they create useful trade-offs. Validate that distant settlements remain understandable.
Exit: a larger realm stays manageable; transport and services explain its limits, with no arbitrary population penalty.
Prove a second realm is genuinely different
Choose a contrasting setting, such as a lagoon trade capital after an inland capital. Research its institutions, architecture, names, technologies and threats. Build distinct economic and defensive possibilities before extending the roster to Franks, English, Venice and Ming.
Exit: players change their plans because of the realm, rather than repeat the same build order under a new skin.
Qualify the native experience
Finish toy-model silhouettes, production and defense feedback, large-settlement camera controls, accessibility, localization, audio and haptics. Verify touch targets, text scaling, reduced motion, long campaigns, background/resume, migrations and performance on physical devices.
Exit: tested complete campaigns and reliable recovery across supported devices; internal distribution evidence recorded.
Expand testing, then prepare release
Run internal and external playtests, resolve balance exploits and late-game management fatigue, and finalize campaign scope. Prepare store materials, privacy disclosures and support. Expand historical scenarios and realm content only after the core campaign holds up.
Exit: an approved release candidate supported by gameplay, reliability and usability evidence.
07 / Research & decisions still open
A grounded setting.
An honest scope.
Questions to answer before production
- Which realm, starting year and capital give the smallest convincing first campaign?
- How many productive daily decisions fit a session, and when should quiet-day advancement stop?
- How much tactical control does defense need before it competes with settlement management?
- Which staffing and transport details create strategy, and which should become standing orders?
- How much history can daily turns cover? Multi-century progression needs a separate pacing design.
- Which research and construction durations can be compressed without confusing the calendar?
Research commitments
Build a scenario dossier for each realm: period, seat of power, settlement names, architecture, production, institutions and dated threats. Keep sources and uncertainty with the content.
Separate historical events from plausible generated incidents. Use local observations for forecasts. Realm traits represent terrain and institutions, with player choices shaping outcomes.
Prototype evidence, target-device measurements and player observation determine scope. Dates below describe historical sources, not implemented game content.
Historical starting references
- UNESCO · Aachen Cathedral and Charlemagne’s royal residence — context for a Carolingian starting seat.
- UNESCO · Venice and its Lagoon — settlement geography and maritime development.
- The Palace Museum · Palace construction and the move from Nanjing to Beijing — capital choice must follow the Ming scenario’s date.
- English Heritage · Hound Tor and the harvest failures of 1315–17 — a sourced agricultural crisis candidate.
- Historic England · Berwick’s medieval defenses — reference for walls, towers, repairs and changing defensive needs.