Play Circuit Courier →First Realm
In development. An original strategy world focused on expansion, decisions and a living economy.
View project →Games built around meaningful choices.
Arcade Lab — Circuit Courier
Play 60 original circuit puzzles directly in your browser. Explore the related design tools and 50 research-oriented guides.
Explore original MojPlayWin projects, playable experiments and worlds in development.
MOJPLAYWIN / CURATED
Next World
Reserved for the next independently designed MojPlayWin title.
Experiments
Small prototypes used to validate mechanics before they become full games.
Living Cities — Build a world worth keeping
An original free-to-play, 3D city-builder and business tycoon. Open 20 businesses, manage 12 characters, discover up to 60 cities and progress through 200 missions across 20 world regions. English by default, with 19 more languages selectable inside the game.
▶ PLAY LIVING CITIES — FREE Explore game features and languages →
Local browser save · no purchases or unverified MOJ1 payments.
Designing a game that earns a second session
An engaging game starts with a repeatable player decision, not with a large asset library. Define the verb (choose, build, trade, defend), the cost of an action, the feedback it produces and the reason to attempt another round. For a browser strategy game, a small prototype might offer three decisions per turn and one transparent success condition; adding ten currencies before players understand the first decision usually increases confusion.
Measure activation (the share of new visitors who finish the first meaningful action), early abandonment, voluntary replay and the percentage of players who can explain the objective. Segment by device and acquisition source instead of treating every user as identical. Respect privacy and collect only the telemetry needed to improve play. Monetization should follow evidence of enjoyment, with clear prices and without manipulative pressure.
Apply this in three steps
- Write a one-sentence player promise and a two-minute playable loop.
- Run five observed usability sessions; note the first moment each player hesitates.
- Change one rule at a time, document the hypothesis, and retest before adding content.
Key takeaway: Practical test: if players cannot state the goal after one round, clarify the interface before buying ads.
Explore more free field notes →Balancing a strategy game's economy with measurable constraints
A strategy-game economy is a system of sources, sinks and player choices. Sources create currency or resources; sinks remove them. If production compounds without maintenance costs, veteran players may accumulate unlimited purchasing power and the early game can stop making sense. Begin with a spreadsheet that describes each turn in units rather than marketing language.
Working method
Define the player's starting stock, production rate, upgrades, maintenance and maximum productive capacity. Simulate one hundred turns for cautious, balanced and aggressive players. Compare affordability of important decisions at turns five, twenty and fifty. Use a deterministic model before adding random rewards so mathematical defects do not disappear behind probability.
Worked example
A workshop producing twelve units per turn and consuming four in upkeep generates eight net units. If construction costs sixty units, simple payback is sixty divided by eight, or seven and a half turns. That ignores construction time, capacity limits and other opportunities; add those to a commercial-quality model.
Common mistake
Balance for new and experienced players. A single expert playthrough is not enough.
Action checklist
- List all sources and sinks.
- Simulate alternative player strategies.
- Playtest for clear decisions and late-game balance.
A browser game that feels responsive on modest phones
For browser games, polished graphics cannot compensate for delayed taps or movement that freezes. Responsiveness depends on main-thread work, asset weight, draw calls, memory pressure and network behavior. A performance budget and testing on modest devices matter more than a beautiful desktop demonstration.
Working method
Measure frame timing, long tasks and the first meaningful interaction on real hardware. Avoid changing layout within animation loops, load only assets needed for the first session and defer optional worlds, soundtracks and cosmetic resources. Provide lower-detail modes and avoid unnecessary high-resolution textures on memory-limited devices.
Worked example
At sixty frames per second a frame has about 16.7 milliseconds available for input, scripting, layout and painting. A twenty-five-millisecond update loop cannot meet this target consistently. Reducing work or changing scheduling is more effective than adding a loading spinner. Observe frame-time distribution rather than average frame rate alone.
Common mistake
Recheck after extended play sessions because memory growth and heat can reveal problems that a short benchmark hides.
Action checklist
- Capture a trace on a low-end phone.
- Budget assets and work per frame.
- Check touch, keyboard, pause and reduced-motion support.
Designing difficulty curves with player-learning evidence
A difficulty curve should reward learning without forcing every player into the same pace. The most useful design question is not whether a level is hard, but whether the player understands what skill the level tests and what they can do differently after failing. Monitor completion, retries, time to success and abandonment together.
Implementation method
Map each mechanic to an introductory challenge, a safe practice area and a combined challenge. Record skill prerequisites and ensure the interface teaches them before testing them. Build a level dataset with target completion rate, actual completion rate, median attempts and qualitative frustration notes. Segment by experience level rather than hiding beginners in an aggregate metric.
Applied example
In a hypothetical puzzle sequence, levels one to five teach individual actions and levels six to ten combine them. If level seven takes six retries while neighboring levels take two, inspect unclear rules, visual affordances and abrupt mechanic combinations. Making enemies weaker may not solve a comprehension defect. Change one element and retest.
Important limitation
A high fail rate is not automatically evidence of engaging difficulty; repeated confusion and forced spending are warning signals.
Action checklist
- Identify the skill tested by every challenge.
- Measure attempts, abandonment and qualitative understanding.
- Retest revised levels with beginners and experienced players.
Fair real-time multiplayer under uneven network latency
Networked games must separate player intention from authoritative state. Different connection speeds can make the same encounter appear inconsistent, especially when players move rapidly or inputs are time sensitive. Good netcode balances perceived responsiveness with server validation and predictable reconciliation.
Implementation method
Define the tick rate, permitted input age, interpolation delay and server-authoritative checks. Use client prediction for immediately visible local movement when suitable, followed by reconciliation against validated server state. Log round-trip latency, jitter and packet loss separately. For competitive mechanics, document lag-compensation rules and test them under simulated impairment.
Applied example
Consider a match where one player has 35 milliseconds of round-trip delay while another has 170 milliseconds with high jitter. Merely averaging their delays misses the sudden bursts that cause apparent teleportation. Capture timelines of client inputs and server snapshots to distinguish rendering stutter from network disagreement. Replays should rely on authoritative events.
Important limitation
Do not trust client-reported damage or inventory changes as final truth. Security and fairness depend on server-side validation.
Action checklist
- Set measurable latency and fairness budgets.
- Test loss, jitter and disconnections.
- Audit authoritative events and reconciliation outcomes.
Launching player-created worlds with moderation by design
User-generated levels, characters and community posts can expand a game's creativity but introduce content, ownership and abuse risks. Moderation should be part of the creation pipeline rather than an emergency patch after a platform becomes popular.
Implementation method
Create publishing states such as draft, private test, pending review, published and removed. Restrict executable scripts and uploaded file types. Provide reporting, appeals, age-appropriate discovery controls and clear creator-license terms. Rank user content using a mix of quality, safety and diversity rather than exposure alone. Store moderation decisions with appropriate access controls.
Applied example
For a building game, a creator could submit a level containing a title, thumbnail, object list and optional script. Before public distribution, validate object limits, scan assets and review reports. When a level is removed, retain an appropriate audit trail and tell the creator which policy category applied, subject to privacy and safety constraints.
Important limitation
Automated classifiers can assist triage but should not be the sole authority for high-impact decisions or appeals.
Action checklist
- Map creation and publication states.
- Test reporting, takedown and appeal flows.
- Review licenses, privacy and age-based access.
NEW ORIGINAL GAME · 200 STAGES
Orbit Switch
Switch between two orbits, avoid hazards and save your stage progress. Touch and keyboard controls.
Play Orbit Switch →New game: Prism Relay — 150 stages
Rotate mirrors to guide a golden laser through 150 solvable puzzles. Mobile and keyboard play, undo, hints and saved progress.
Play Prism Relay free →