United States:
Transform Digital LLC
44 Montgomery Street, Suite 300
San Francisco, CA 94104
Complete Developer Guide
How to Develop a Game Like Fall Guys
Covers what Fall Guys is, its verified market data, the technical features required, how it monetises, the development process, full tech stack, team structure, and cost by build tier. All claims are source-cited.
What is Fall Guys?
Platform battle royale. 60 players. One winner.
Fall Guys (originally Fall Guys: Ultimate Knockout) is a free-to-play platform battle royale developed by Mediatonic and published by Epic Games. Up to 60 players per session compete through randomly selected obstacle courses and elimination rounds until one player remains. The format was modelled on TV game shows, specifically Takeshi's Castle and Total Wipeout. It launched on PS4 and PC on 4 August 2020. Epic Games acquired Tonic Games Group (Mediatonic's parent) in March 2021. The game moved to free-to-play on 21 June 2022, simultaneously launching on Xbox and Nintendo Switch and delisting from Steam. (Source: Epic Games press releases, 2021–2022.)
UnityEngine
60Players/session
Aug 2020Launch date
6+Platforms
F2PSince Jun 2022
76Round types
Verified market data
Fall Guys — Key Statistics
Four figures from public, citable sources. The stats section below excludes figures that cannot be independently verified.
172,213 Peak concurrent Steam players (Aug 2020) SteamDB / Tracker.gg
$185M Revenue in first month on PC (paid at $19.99) Webtribunal / TheGamer
8.2M PC players in first month of paid launch Mediatonic / industry trackers
20M Players in first 48 hrs of F2P launch (Jun 2022) Official @FallGuysGame tweet
Retention context: Average daily Steam players fell 97.6% from ~124,000 (Aug 2020) to ~3,000 (Jun 2023). By 2025, peak daily concurrent is below 1,000. The F2P transition created a spike, not sustained growth. Retention requires ongoing live-ops investment — new rounds, seasonal events, and IP partnerships — not just lower acquisition barriers. (Source: LEVVVEL player count analysis.)
Technical requirements
Core Features — What Must Be Built
Seven systems that are architecturally required for this game type. Each card states what the feature does and what breaks if it is omitted or under-scoped.
Server-Authoritative Physics
All player movement and collision is computed on the dedicated game server. Clients predict local movement but the server adjudicates every state.
Without it: clients disagree on collision outcomes; speed-hacks and position exploits become trivially easy.
Dynamic Round Selection
A matchmaking service picks rounds at runtime based on current player count and the elimination stage reached. Round selection cannot be hard-coded.
Without it: the game breaks when player count drops mid-session or a round is designed for a player count that no longer applies.
Real-Time Elimination Events
Elimination state changes broadcast to all connected clients simultaneously. Acceptable latency threshold: under 100ms. Round transitions trigger automatically on elimination count.
Without it: players see inconsistent lobby states; some clients show a round as ongoing while others show it ended.
Cosmetic Inventory & Economy
Per-account persistent item storage, virtual currency accounting, seasonal reward unlock logic, and a transactional in-game store. PlayFab covers all of this without custom database work.
Without it: no monetisation layer exists. This is where all F2P revenue flows through.
Redis-Backed Leaderboards
Per-round and all-time rankings require sub-5ms read latency under concurrent load. Redis provides this. Relational databases cannot sustain this latency profile at scale.
Without it: leaderboard reads under peak load cause database contention and visible UI lag across all connected clients.
Party & Social Layer
Persistent squad formation across rounds, friend invites, and cross-platform presence management. Epic Online Services handles this natively on EOS-based builds.
Without it: groups of friends cannot reliably play together, which is the primary social use case for this genre.
Adaptive Third-Person Camera
Field-of-view and zoom respond dynamically to local player density and obstacle proximity. A fixed camera creates spatial awareness gaps in high-density group moments.
Without it: players frequently lose track of their character in crowd scenarios, a documented source of negative reviews in this genre.
Business model
Revenue Model — Five Income Streams
Fall Guys launched as a $19.99 paid title and generated $185M in its first month. It moved to free-to-play in June 2022. All five revenue streams below operate in parallel and are cosmetics-only — no pay-to-win mechanic exists in the game.
| Stream | Mechanism | Known Data Point |
|---|---|---|
01 Battle Pass | Two tracks — free and premium — unlock cosmetics tied to play progression over a fixed season. Players who buy early are incentivised to play regularly to recover value. | ~$9.99/season (950 Show Bucks) |
02 Cosmetic Shop | Skins, emotes, name plates, and patterns sold individually via Show Bucks. Items are time-limited in the storefront, creating purchase urgency without gameplay pressure. | Primary daily revenue driver |
03 Premium Currency | Show Bucks are sold in bundles that do not align with item prices (e.g. 1,000 Show Bucks for a 950 Show Buck item). Leftover balances drive additional purchases. | Standard F2P purchase-friction mechanic |
04 IP Collaborations | Limited-edition cosmetics licensed from external brands (Sonic, Godzilla, film and sports franchises). Partners co-promote the collaboration, buying player acquisition for Fall Guys at zero marginal cost. | Drives both direct revenue & new-user acquisition |
05 Seasonal Events | Time-limited rounds and exclusive cosmetics tied to real-world calendar events. The primary reactivation mechanism for lapsed players. DAU data shows the steepest spikes correlate with event launches, not season starts. | Peak DAU spikes tied to event windows |
Build process
Five-Stage Development Process
Each phase has defined deliverables and exit criteria. The most expensive mistake in this genre is compressing Phase 3 to hit a deadline — retrofitting a server-authoritative network layer onto an existing codebase is effectively a full rewrite.
1. Conceptualisation & GDS
Produce a game design spec before any engine work. Define player count per session, round types, elimination rules, progression structure, and monetisation model. Player count is the single variable with the largest downstream impact — changing it from 30 to 60 after networking is built requires rebuilding the server topology and re-tuning every physics interaction.
2. Level & Character Design
Build grey-box obstacle courses in-engine to verify physics feel before committing to art. Character hitbox dimensions must be finalised here — they feed directly into server-side collision detection. If hitboxes change after networking is implemented, collision tuning must restart from zero. Art style locks in this phase.
3. Core Development & Networking
Build the server-authoritative game loop first — before any gameplay features sit on top of it. Implement the game server, multiplayer sync layer, physics integration, and matchmaking service in this order. Unity's Netcode for GameObjects and Unreal's native replication system both support this pattern correctly at 60-player scale.
4. Testing & Platform Certification
QA covers functional testing, cross-device compatibility, and stress testing at 2× expected peak concurrency. PlayStation, Xbox, and Nintendo certification each add 4–8 weeks and require specific technical compliance beyond standard QA. Anti-cheat integration runs as a parallel workstream from this phase — not a post-launch task.
5. Deployment & Live Operations
Soft-launch in one region before full rollout to surface server scaling issues and matchmaking edge cases under real conditions. Post-launch engineering is a recurring cost, not a one-time effort: seasonal content, server capacity adjustments, and matchmaking tuning are ongoing. Matchmaking quality degrades when active players drop below the minimum needed to fill lobbies within acceptable wait times.
Engineering — verified
Technology Stack
Each card below states whether the technology is confirmed (documented in developer interviews or official statements) or recommended (best-practice for an equivalent new build). Fall Guys' full internal stack is not publicly disclosed — confirmed items only.
Confirmed Game Engine
Unity + Custom PhysX Controller
Mediatonic built a custom character controller on Unity's PhysX integration. The wobbly movement feel is intentional and required significant tuning — it is not a standard Unity character controller.
Confirmed Multiplayer
Epic Online Services (post-2022)
After the Epic acquisition, Fall Guys migrated to EOS for cross-platform identity, matchmaking, and session management. Pre-acquisition: AWS. AWS GameLift remains the strongest independent alternative — it handles auto-scaling game server fleets natively.
Confirmed Economy
Microsoft PlayFab
Fall Guys uses PlayFab for account management, player stats, in-game economy, and the virtual currency store. PlayFab handles entitlements, virtual currency ledgers, and player inventory without custom database schemas.
Recommended Session State
Redis + PostgreSQL
Active session state and per-round leaderboard reads require sub-5ms latency. Redis provides this. PostgreSQL or MySQL handles persistent player data that does not require sub-millisecond access. Using a relational database for leaderboards causes visible UI lag under concurrent load.
Recommended Backend
Node.js or Go
Node.js and Go are standard for real-time game server logic because of their non-blocking I/O throughput. Matchmaking and account services can use Python or Java without performance penalty. Kafka or RabbitMQ decouples game server event streams from analytics pipelines.
Confirmed Anti-Cheat
EasyAntiCheat
Fall Guys uses EasyAntiCheat on PC (now an Epic Games product). BattlEye is the main alternative. Console platforms enforce their own integrity — no client-side anti-cheat is implemented on console builds. Anti-cheat requires its own QA workstream from Phase 3 onward.
Build approach
Development Approaches — Compared
Three structural models for building a multiplayer platformer. Each has distinct trade-offs across IP control, cost, timeline, and post-launch flexibility.
In-House Team
Highest fixed cost · Slowest ramp-up
Full IP ownership and roadmap control with no third-party dependencies.
No contractual limitations on scope changes or technology choices post-launch.
Multiplayer networking engineers are a narrow, competitive talent market with high salaries.
Team must be built before production begins — adds 3–6 months to the start date.
Risk: fixed overhead continues if the title underperforms post-launch.
Contracted Studio
Fixed scope cost · Fastest ramp-up
Predictable cost against a defined scope. Studio brings an existing team and toolchain.
Production starts within weeks of contract signing, not months.
IP ownership, source code delivery, revision rights, and post-launch obligations must be explicit in the contract or they default to studio-favourable positions.
Out-of-scope changes are charged as change requests, which can exceed original estimates.
Risk: verify multiplayer-specific experience before signing — most studios are app development vendors, not game studios.
Mixed Model
Moderate cost · Requires internal lead
Studio handles engine and networking (highest-complexity work); freelancers handle art and audio (non-critical-path work at lower rates).
Reduces total cost compared to a full contracted studio by 20–35% on a typical project.
Requires an internal technical lead to evaluate deliverables and maintain integration quality across workstreams.
Coordination overhead increases with the number of independent contributors.
Risk: without a capable internal lead, the mixed model degrades into unmanageable scope fragmentation.
Budget
Cost Breakdown
Basic/POC — validates core mechanics, small player cap, limited round types. Advanced — production-ready, public release, live-ops capable. Networking and anti-cheat account for 20–30% of Advanced-tier engineering cost.
| Feature Area | Basic / POC | Advanced |
|---|---|---|
| UI/UX & Game Design | Custom visual theme, standard UI component set, single art pass. | Animated UI system, design tokens, high-fidelity visual effects, particle systems, accessibility compliance. |
| Core Gameplay | Login, lobby, basic round mechanics, real-time scoring. One to two round types. | Multiple round types, dynamic elimination logic, seasonal event framework, scalable session management. |
| Multiplayer Networking | Basic server-authoritative loop. Player cap: 8–16 per session. | Full server-authoritative physics, lag compensation, rollback netcode. Player cap: 60+ per session. Cross-platform play. |
| Backend & Infrastructure | Monolithic backend, relational database, moderate concurrent load. | Microservices, auto-scaling cloud (AWS GameLift or EOS), real-time pipeline, multi-region load balancing. |
| Economy & Integrations | Payment gateway, push notifications, basic analytics. | PlayFab economy layer, anti-cheat integration, AI analytics pipeline, real-time notifications. |
| Post-Launch Support | 3–6 months: critical bug fixes only. | 12+ months: priority support, seasonal content, matchmaking tuning, security monitoring, live-ops tooling. |
| Estimated Cost | $10,000 – $30,000 | $50,000 – $150,000+ |
Project planning
Team & Timeline
A physics prototype with basic networking takes 4–8 weeks. Full production targeting public release takes 9–18 months with 8–15 people. The networking phase and QA/stress-testing are consistently the two longest phases.
Required Roles
Lead Network Engineer
Server-authoritative loop, lag compensation, rollback netcode
Gameplay Engineers ×2–3
Physics, character controller, UI, round logic, elimination detection
Backend Engineer ×1–2
Game server, matchmaking, APIs, database, PlayFab integration
3D Artists ×2–3
Character models, obstacle assets, environment, VFX, animation
Level Designer
Obstacle course layout, difficulty curve, physics interaction tuning
UI/UX Designer
HUD, menus, cosmetic store, onboarding, platform UI compliance
QA Engineers ×1–2
Functional testing, device matrix, stress testing, cert compliance
Project Manager
Sprint planning, milestone gates, risk management
Phase Timeline
| Concept & GDS | 2–4 weeks |
| Single-player prototype | 4–8 weeks |
| Core multiplayer build | 8–16 weeks |
| Level design & art (parallel) | 10–20 weeks |
| QA & stress testing | 6–10 weeks |
| Platform certification | 4–8 weeks |
| Soft launch & monitoring | 2–4 weeks |
| Total | 9–18 months |
Level design and art run in parallel with core development. Timeline assumes 8–12 people. Smaller teams extend each phase proportionally.
Clients We Have Worked With
Due diligence
Evaluating a Development Partner
Most app development studios cannot build a 60-player server-authoritative multiplayer platformer. Four criteria that distinguish genuinely capable studios from general vendors.
Multiplayer Game Portfolio
Ask for shipped multiplayer games — not mobile apps or web platforms. The disciplines are different.
Networking-Specific Evidence
Ask specifically about server-authoritative architecture, lag compensation, and rollback netcode. These are not backend development skills.
Post-Launch Scope in Contract
QA, platform certification, and live-ops support must be explicitly scoped and costed before signing.
Live Title Retention Data
A shipped game with players through a second content season is proof of capability. A prototype is not.
FAQ
Frequently Asked Questions
Each answer below contains only verifiable claims. Sources are cited inline.
What does it cost to develop a game like Fall Guys?
A proof-of-concept with simplified physics and 8–16 players: $10,000–$30,000. A production build with 60-player real-time multiplayer, cosmetic system, and cross-platform support: $50,000–$150,000. AAA-scale productions with custom engine work, full live-ops infrastructure, and multi-platform certification: $500,000+.
The multiplayer networking and anti-cheat layers account for 20–30% of total engineering cost at the production tier. This is the most consistently underestimated line item in early-stage budgets.
How many players does Fall Guys support per session?
Up to 60 players per session. A dedicated game server manages all physics state and broadcasts authoritative positions to each client. Client-side prediction is applied to the local player's own movement to mask network latency. Supporting 60 simultaneous physics-simulated players requires one capable server per active session — server compute cost scales directly with concurrent sessions, not total registered users.
Source: Fall Guys store page descriptions; multiple developer interviews confirming 60-player architecture.
How long does the full development cycle take?
A single-map physics prototype with basic networking: 4–8 weeks with 2–3 engineers. A full production build for public release — multiplayer networking, QA, platform certification, and soft-launch — typically takes 9–18 months with 8–15 people. Load testing at 2× expected peak concurrency adds 4–6 weeks by itself. Platform certification on PlayStation, Xbox, or Nintendo each adds 4–8 weeks independently.
Did the F2P transition actually grow Fall Guys' player base?
Short term: yes. 20 million players joined in the first 48 hours of the June 2022 F2P launch. (Source: official @FallGuysGame and @EpicGames tweets, June 25, 2022.)
Long term: no. Average daily Steam players fell 97.6% from ~124,000 in August 2020 to ~3,000 in June 2023. By 2025, peak daily concurrent on Steam was below 1,000. (Source: LEVVVEL player count analysis; Tracker.gg Steam data.) The F2P transition removed the acquisition barrier but did not create engagement. Sustained engagement requires ongoing live-ops investment in new content, events, and IP collaborations.
Unity or Unreal Engine for this type of game?
Fall Guys uses Unity. For a new build, Unity remains the stronger default: larger pool of multiplayer engineers, more mature mobile export pipeline, and Netcode for GameObjects is specifically designed for the server-authoritative model required at 60-player scale.
Unreal Engine's native replication system is technically equivalent. Choose Unreal only if the studio already has deep Unreal expertise — switching engine mid-project is not a viable option regardless of direction. Godot 4 lacks production-verified track record at 60-player concurrent scale.
What is Fall Guys' confirmed technology stack?
- Game engine: Unity with custom character controller on PhysX (confirmed — developer interviews)
- Anti-cheat: EasyAntiCheat on PC (confirmed — Epic Games product documentation)
- Player economy & accounts: Microsoft PlayFab (confirmed — Microsoft case study)
- Cross-platform services: Epic Online Services post-June 2022 (confirmed — Epic press release)
- Server infrastructure pre-2022: AWS (confirmed — Mediatonic interviews)
Session state layer, internal event streaming, and backend service framework are not publicly confirmed. Recommended equivalent for a new build: Redis for leaderboards and session state, PostgreSQL for persistent player data, Kafka for event streaming.
Sources: Mediatonic developer interviews; Epic Games public statements; Microsoft PlayFab case study; GDC talks on game backend architecture.
