MARS INVADERS Investor overviewBuilt 01 / 12

October 2026 · Mars Invaders

Four hundred waves.
One fleet that remembers.

A real-time space strategy game that runs in a browser tab. Ships that survive keep their names and their damage, and the fleet’s research and credits carry forward across forty operations in four campaigns: from Earth orbit to a citadel beyond the Sun’s light, then back to an alien gate at Venus.

40 / 400operations and waves across
four continuous campaigns

2,963unit slots reserved for the heaviest
operation, inside a 3,072 pool

0installs, launchers or store
clients between link and play

523simulation regression tests,
all passing today

Badge key. Built: implemented in this repository as of 1 October 2026, which is not production readiness. Illustrative: arithmetic on assumptions printed on the same page, not a forecast. Sourced: an external figure linked at the foot of its page. Next: planned, not yet built. The game has no revenue, no store listing and no paying players, and nothing in this deck promises commercial results.

MARS INVADERS Problem and insightBuilt 02 / 12

Depth lives behind an install. Reach lives behind a link.

Today

Campaign-length strategy ships as a client: a store page, an account, a purchase and a large download before the first shot.
Browser games reach players instantly, but most are single-session arcade games that forget the player when the tab closes.
A new strategy title on a storefront is one row in a list, and it has to buy attention before anyone has played a minute.

Mars Invaders puts the deep half on the reachable side

The insight Built

Persistence is the hook. Across 400 waves, survivors, hull damage, cargo, credits, research and station refits carry forward, and losses stay lost.
The game is a static bundle served from an edge CDN. There is no install, launcher or patcher; the link is the demo.

Both lanes describe distribution mechanics, not measured conversion. This game has no funnel data: there is no store page, no telemetry and no player cohort yet. The browser lane is Built: the game loads and plays from a URL today.

MARS INVADERS ProductBuilt 03 / 12

A commander’s job, not a click-fest.

Ten waves per operation, hostiles arriving on separate approaches, and objectives that sit outside the station’s weapon envelope. Every decision is paid for later, because the next operation starts with the same fleet.

01Command the Earth fleetRight-click movement, formations, focus and docking. Only Earth-owned ships accept orders; allied Belt squadrons fight on their own initiative.
02Hold the stationFirst the Earth Nexus, then anchorages founded or inherited across the outer system and at Venus. Waves split across lateral and rear approaches, so the defence cannot face a single lane.
03Work the away objectivesRecover, hold, destroy, escort and found new stations, at real positions beyond the reach of the station’s guns.
04Research and refitReverse-engineer Martian shields from wreckage, then unlock twelve one-off technologies and five-tier upgrade tracks gated by station level. Five refits rebuild the Nexus from a stripped ring.
05Live with the lossesBetween operations you see the survivors by name. Rebuilding costs credits that the next operation will also want.

Operation 5, Behind the Stones: Belt relief ships arriving from behind a wall of asteroids — **Built**Original theater illustrations shipped in the repository

Operation 10, A Sky for Everyone: the HELIOS command lattice over the final assault — **Built**Original theater illustrations shipped in the repository

These are the SVG files the game serves from /assets/campaign/: briefing art for operations 5, 10 and 20, not mock-ups and not combat screenshots. Twenty illustrations cover operations 1–20; operations 21–40 reuse three of them.

MARS INVADERS What exists todayBuilt 04 / 12

The campaign is built. Balance is next.

Everything below is implemented in the committed repository today.

Campaign

Four campaigns, 40 operations

The Broken Orbit (1–10), The Silent Citadel (11–20), The Venus Disturbance (21–30) and A Sky Without Orders (31–40), ten waves each. Operations 10, 20 and 30 end in an intermission carrying one epilogue and the next prologue; final victory follows operation 40.

Combat identity

Four-way faction hostility

Earth, Mars, the Belt and an alien fleet. The aliens fight everyone, the Belt and Mars fight each other, and a projectile or beam keeps its owner after the ship that fired it dies.

Station

Refit levels 0–5

Maximum hull grows 3,600 → 7,200 and Nexus beam turret capacity 0 → 10; separate Nexus gun turret mounts grow 8 → 18, filled up to Gun turret mounts research (3 → 18). The Nexus is one of 14 operational stations, each with its own hull, shield and turrets.

Capacity

Logistics to 20 fleets, 24 with Fold

Fleet cap 3 → 6 → 9 → 12 → 15 → 20, plus four from Fold Logistics; mobile cap 27 → 216 units. Build admission counts queued ships, so the queue cannot oversubscribe a fleet.

Engine budget

3,072-unit entity pool

Before an operation starts, a catalog check reserves the full player fleet, every station and every authored ship across its ten waves. The heaviest case, operation 40, needs 2,963.

Persistence

Versioned checkpoints

Browser saves, downloads and imports share one serialised format. Saves from version 5 onward migrate forward. Loads validate before changing any state, and a corrupt file is rejected.

Art

20 theater illustrations, 71 portraits

Hand-authored SVG briefing art in one etched-space style for operations 1–20, reused in campaigns III and IV, plus 71 portraits for the transmission system.

Pipelines

Scripted asset production

371 GLB models including LOD variants, 346 KTX2 textures and 506 audio files, with committed import and generation scripts for models, LODs, textures and sound.

Tooling

Eleven review pages and a builder

ships, turrets, effects, explosions, missiles, stations, asteroids, nexus, avatars, sounds, planets at the developer origin root render the real libraries outside the game, so art and audio can be checked without playing an operation.

BuiltHistorical verification. On the recorded committed tree, TypeScript checking and the production build passed, and all 523 simulation regression tests passed under bun test src/simulation. Campaign lifecycle, checkpoint and capacity evidence came from accelerated runs and catalog checks. No operation had been played through and signed off at normal speed, so combat balance was unverified.

Model, texture and audio counts come from the committed public/ tree on 1 October 2026. “Built” describes the repository, not production readiness: there is no launch build, no store listing and no public launch.

MARS INVADERS The campaignBuilt 05 / 12

Forty operations, one continuous fleet.

A war over water and transit rights becomes a voyage out of the solar system, then a return to an alien gate at Venus. Every operation has its own briefing, objectives, debrief and in-battle transmissions tied to real wave events.

Portrait of Commander Imani Vale — **Imani Vale**Earth · CommanderHolds the Nexus and keeps a civilian corridor open at tactical cost.

Portrait of Captain Sera Venn — **Sera Venn**Belt · Salvage captainSpeaks for a fractured Belt and becomes an ally only through mutual survival.

Portrait of Marshal Ares Kade — **Ares Kade**Mars · MarshalTurns a blockade into an extermination campaign, then surrenders to the Nexus.

Portrait of Fleet Commander Ixael — **Ixael**Alien · Fleet commanderThe living commander aboard the alien carrier, increasingly overridden by Control.

Portrait of Mothership Control — **Mothership Control**Alien · Relay intelligenceThe voice every alien hull repeats; a local copy outlives the closed gate.

Titles, locations and speakers come from src/simulation/campaign.ts. Node colour marks the hostile faction: red for Mars and its Remnant, amber where Belt or Titan Rim crews are hostile (operations 2, 3 and 12), violet for the alien fleet. Solid nodes are campaign finales. Portraits are the speakers’ own files in /assets/avatars/.

MARS INVADERS TechnologyBuilt 06 / 12

A full real-time strategy loop inside a browser tab.

Fixed pools, a headless-testable simulation and a static edge-delivered bundle. Nothing here needs a server at play time.

TypeScript, Three.js and WebGLThe renderer runs on three@0.186, and the build is Vite 8 targeting ES2022. Bun runs the tests and the asset pipeline scripts.
Simulation separated from renderingsrc/simulation is a pure module with no renderer import, which is why the campaign can be exercised headless under Bun instead of only by playing it.
Fixed budgets, no per-frame allocation3,072 units, 2,048 projectiles, 256 beams, 384 effects. Targeting reuses pooled arrays instead of allocating temporaries each frame.
Chunked delivery from CloudflareThree’s core and renderer ship as separate chunks, and KTX2 planet textures and LOD ship models keep the first payload down. Each frontend has its own static build and Cloudflare Worker target; the engine remains a shared library.
Checkpoints are a versioned contractPayloads carry hull kinds and loadout designs, construction jobs, per-station levels and turrets, a technology table and typed research jobs. Saves from version 5 onward migrate forward, and anything older is rejected. Validation runs before any state changes.

Budgeted against the 3,072-unit pool

2,963units reserved for the heaviest operation

310hostiles in the largest authored wave

2,048projectile slots, pre-allocated

The 2,963 figure is a catalog preflight bound from docs/campaign-plan.md, not a measurement. It counts every authored hull and a full player fleet as if all were alive at once, and it predates the Hive drone corvette, whose swarms it does not include. It makes no claim about GPU load, sustained combat or frame rate on any device. The earlier 557-ship pass ran against a 640-unit pool and is superseded.

MARS INVADERS Why browser, why nowSourced 07 / 12

The distribution is the feature.

The browser is no longer the compromise platform. What it still lacks is depth, and that is the part already built here.

01The link is the demo BuiltThe whole game is static files on an edge CDN. A player can be giving fleet orders while a client download would still be queueing, and the same URL works on a laptop handed across a desk.

02The browser runtime caught up SourcedWebGPU now ships in all three major engines. Apple’s Safari 26 announcement names Three.js, the renderer this game uses, as working well on it, and calls WebGPU the successor to WebGL.

03Web portals are a real acquisition channel SourcedPortals report audiences in the tens and hundreds of millions of players, and publish developer terms in which the portal carries user acquisition rather than the studio. See page 08.

04Nothing sits between the player and the game IllustrativeOn the open web there is no storefront approval, no platform build certification and no client update channel. A payment processor still takes a fee; the deck assumes 15% of gross on page 10.

05Desktop stays open as a second act NextThe same bundle can later be wrapped for Steam and itch.io, using storefronts for discovery once the web build has proven the game. Not built, and not scheduled before Q3 2027.

Browser-engine facts come from the vendors’ release posts linked above, accessed 15 September 2026. Point 04 describes structure, not a measured saving; the 15% fee on page 10 is an assumption. No portal has been approached and no distribution agreement exists.

MARS INVADERS MarketSourced 08 / 12

A large audience on the reachable platform.

Five external figures, each linked. None of them is a forecast for this game; they describe the size and shape of the surface it would launch onto.

$43BSourced

PC games spending in 2025 within a $197B global market, and at +10.4% the fastest-growing platform in that forecast.

Newzoo global games market forecast, 2025, via Outlook Respawn. Secondary report; Newzoo’s own page is not publicly readable.

625MSourced

Players reached by the browser portal Poki in 2025, with roughly 100M monthly actives and about 1B gameplays per month.

Poki company announcement, 2026. accessnewswire.com. Company-reported figure.

50M+Sourced

Monthly players on CrazyGames, whose published developer terms state the portal carries user acquisition at no cost to the developer.

CrazyGames developer documentation, 2026. docs.crazygames.com. Platform-reported figure.

13.97%Sourced

Share of all-time Steam revenue attributed to strategy, with real-time strategy alone at $1.84B lifetime gross. The genre has a paying audience.

GameDiscoverCo genre analysis, 2025. newsletter.gamediscover.co. Covers only titles above $1M lifetime gross; all-time, not annual.

73%Sourced

Of surveyed industry professionals place PC in their top three next-generation platforms. In the same survey, 28% report being laid off in the past two years.

GDC State of the Game Industry 2026, n > 2,300. gdconf.com. Self-selected developer survey.

0Built

Players Mars Invaders has today. There is no store page, no portal listing, no telemetry and no marketing spend, so every conversion number in this deck is an assumption, not a measurement.

Repository state, 1 October 2026. The counterweight to the five figures on this page.

All sources accessed 15 September 2026. Platform and portal audience totals are self-reported and are not comparable with one another; registered users, monthly actives and lifetime players count different things. The Steam genre split covers only titles above $1M lifetime gross and is all-time, not annual. None of these figures is addressable market for this game: no share of any audience above is assumed anywhere in this deck.

MARS INVADERS Business modelIllustrative 09 / 12

Play three operations free. Buy the campaign once.

One purchase, no subscription, no advertising and no loot boxes. Every price below is a working assumption chosen for this deck, not a tested or decided price.

Tier 01 · Acquisition

Free web demo

$0Assumption: operations 1–3

Operations 1–3 in the browser with no account: first contact, the water ledger and the Martian trap. The checkpoint at the end of operation 3 is what the unlock continues from.

Tier 02 · Primary

Campaign I unlock

$14.99Assumed price, one-time

The Broken Orbit in full: operations 1–10, all station refits, all research tiers, export and import of checkpoints. One payment, permanent.

Tier 03 · Expansion

Campaign expansions

$9.99Assumed price, one-time, each

The Silent Citadel (11–20), The Venus Disturbance (21–30) and A Sky Without Orders (31–40), each sold to players who finished the campaign before it, so every expansion sells to a known audience.

Tier 04 · Optional

Fleet liveries

$3.99Assumed price, cosmetic only

Cosmetic hull liveries for the persistent fleet. No stat effect, no gameplay unlock, no randomised purchase. Excluded from every revenue figure in this deck.

Assumptions, not decisions

No price test has been run, no payment processor is integrated, no store or portal account exists, and no player has been asked to pay. The four tiers propose one way to package the built content. Treat every price on this page as a variable to set after the closed beta on page 11.

The content boundaries are real. Campaigns span operations 1–10, 11–20, 21–30 and 31–40 in src/simulation/campaign.ts, and a checkpoint can be saved after any operation before the last, including operation 3. The prices, the free/paid split and the cosmetic tier are Illustrative and carry no revenue promise.

MARS INVADERS Illustrative economicsIllustrative 10 / 12

Base case, every input visible

60,000monthly players who start the free demoAssumption

3.0%of those who buy the Campaign I unlockAssumption

1,800buyers per monthDerived

$14.99assumed list price per unlockAssumption

$22,935net revenue per month after 15% feesDerived

Three scenarios on the same arithmetic
Scenario	Monthly players	Conversion	Price	Buyers	Net per month
Low	20,000	1.5%	$11.99	300	$3,057
Base	60,000	3.0%	$14.99	1,800	$22,935
High	200,000	4.5%	$14.99	9,000	$114,674

Net = players × conversion × price × 0.85. Campaign expansions, cosmetics and any portal revenue share are excluded, as are payroll, marketing, legal and tax. The base case annualises to roughly $275,000.

$275kbase case, annualised net revenue

$12.74net per unlock after the 15% fee assumption

37.5×high scenario over low, same arithmetic

Use of funds, 18 months

Engineering30%Payments and accounts, onboarding, telemetry, performance work on low-end GPUs.

Art and audio24%Combat VFX and UI polish, an original score, voiced transmissions for key beats.

Marketing and community22%Portal and storefront presence, capture and trailer production, a devlog cadence.

QA and balance14%The normal-speed playthrough of all forty operations, plus device and browser matrix testing.

Runway buffer10%Unallocated reserve against a slipped gate on page 11.

18months of runway assumed

6people, founder plus five hires

Q2 2027web launch target, page 11

Illustrative. Not a forecast, projection or commitment. Every input on this page is an assumption: there is no telemetry, store page, price test, marketing spend or paying player. The arithmetic is printed so that each assumption can be replaced with a measured number after the closed beta, which will change these outputs. The percentage split is a plan for a round that has not been raised.

MARS INVADERS RoadmapNext 11 / 12

Five gates from repository to storefront.

Quarters are relative to October 2026. This is a plan, not a commitment. Each gate depends on the one before it, and the first gate decides whether the game is fun at normal speed.

01Prove the game at normal speed
- Play all forty operations at normal speed and fix the balance problems that only show up there.
- Interactive visual QA of combat, HUD and the save controls in a real browser.
- Delete the Mars app's /developer/test/ debug harness and its build entry before anything ships.
02Closed beta with instrumentation
- First telemetry: where players stop, which operation kills the run, what they never click.
- Onboarding and tutorial pass for the first twenty minutes.
- Device and browser matrix, including low-end integrated GPUs.
03Web launch
- Free operations 1–3, paid Campaign I unlock; accounts and payments built.
- Portal listings where the portal carries acquisition.
- Replace the page-10 assumptions with measured conversion.
04Desktop wrapper
- Wrap the same bundle for Steam and itch.io; no second codebase.
- Storefront assets, trailer, and localisation of campaign text.
- Sell the campaign expansions to players who finished Campaign I.
05New campaigns and cadence
- Further campaigns on the same persistent-fleet contract.
- Regular operation drops instead of a single release event.
- Decide, on data, whether to keep going or to stop.

Next. Nothing on this page is built, and no date here is a commitment. The quarters assume the round on page 12 closes in Q4 2026 and the hires on page 12 are made; a slipped gate moves every gate after it. Gate 01 is the real risk: no operation has been played through and signed off at normal speed, so the difficulty curve is unproven.

MARS INVADERS Team and askIllustrative 12 / 12

One author built all of it. That is the risk and the evidence.

Founder Built

Daniel de Witte

Sole author of every commit in this repository: the simulation, the Three.js renderer, the campaign catalog and its writing, the UI, the asset import pipelines and the twenty theater illustrations. No other contributor and no outsourced code.

680 commits between 4 and 30 September 2026, one author, 27 days. Daily counts come from the repository’s own history.

Hires to make Next

Gameplay and balance designer — owns the normal-speed playthrough and the difficulty curve across forty operations.
Technical artist — combat VFX, HUD and the visual polish pass the engine can already carry.
Composer and sound designer — an original score and voiced transmissions over the existing audio library.
QA, part-time — the device and browser matrix, and the regression discipline the test suite cannot cover.
Community and marketing, part-time — devlog cadence, portal relationships, launch beat.

Five people, none of them hired or promised. The single-author bus factor is the main risk in this company.

The ask Illustrative

$750,000Placeholder round size · assumption

Eighteen months of runway for the five hires above and the plan on page 11, allocated by the split on page 10: prove the balance, run a closed beta, launch on the web, then wrap for desktop.

This number is a working placeholder sized to an 18-month plan for a team of six. It is not a term sheet, a valuation, a committed raise or a validated budget. The real figure follows from gate 01 and the first measured conversion.

mars-invaders · cataclysm · mars-developer · cataclysm-developer Four independent static-assets Worker targets use apps/mars-invaders/wrangler.jsonc, apps/cataclysm/wrangler.jsonc, apps/mars-developer/wrangler.jsonc and apps/cataclysm-developer/wrangler.jsonc. Each publishes only its own app-local dist. This deck belongs at /developer/deck/ on the Mars developer origin. The prior combined-deployment address, mars-invaders.mars-invaders.workers.dev, is historical; no live independent deployment is claimed here.

Illustrative. The round size, its eighteen-month duration and the team of six are assumptions for this page, not decisions. The commit history, the single-author fact and the repository contents can be checked; nothing about money or hiring on this page is real yet. Built means implemented in the repository as of 1 October 2026, which is not production readiness.