Level Design for Casual Games: Theory, Principles, Tools, Testing

Level Design as a Composite Role

Level design combines code and design to implement game systems into actual gameplay experiences. While game designers create the rules and systems, level designers translate those rules into playable levels that players experience.

The Funnel Model

Level designers act as a funnel, taking game mechanics created by designers and funneling them into aesthetic experiences for players. The level designer is the invisible hand guiding players forward through dynamic situations.

Level Designer Responsibilities at King

Level designers at King create new levels (released every 1-2 weeks), design progressions, create content at scale, invent new power-ups and blockers, and design new game modes. The studio emphasizes knowledge sharing through internal playtesting and regular design meetups.

Core responsibilities of level designers at King

The Saga Envelope Framework

The saga envelope is King's model for casual games featuring linear progression (100+ to 1000+ levels), regular new content (every 1-2 weeks), social elements (friend leaderboards, sending lives), and design for casual audiences. This framework is applied across King games like Candy Crush, Farm Heroes, and Blossom Blast.

Key elements of the saga envelope

Why the Saga Envelope Works

The saga envelope is a strong retention driver because new content releases every 1-2 weeks give players constant reasons to return. Players come back to see new levels and new features, creating a predictable engagement loop.

From Browser Games to Saga Games

King often starts with simple 3-5 minute browser games on Gamescom, then applies the saga envelope to transform them into full saga games. Examples include Diamond Digger Saga and Bubble Witch Saga, which originated as browser games before becoming major mobile titles.

Breaking Down Your Game

Start level design by breaking your game into building blocks: game mechanics (linking, swapping, matching), game modes (high score, objectives), gameplay elements (pieces, power-ups), and blockers (obstacles that add challenge). This creates a shared vocabulary for the team.

Breaking down a game into design components

The Four Core Level Design Concepts

All levels should balance four concepts: difficulty (keeping players challenged), rhythm (varying the experience to create emotional journey), flow (guiding players toward goals), and hooks (unique elements that make levels distinct). These concepts work together to create engaging progression.

Four pillars of level design

Hooks: What Makes a Level Unique

A hook is a unique element that distinguishes one level from others—it could be a special mechanic, a twist on existing rules, a unique environment, or a new art style. In Super Mario 3D World, platforms switch colors when you jump; in Shadow of the Colossus, each boss fight uses different mechanics.

Creating Hooks in Casual Games

Hooks in casual games can come from utilizing a mechanic in a new way (like focusing on dragon creation in Lucky Lanterns), combining different game elements to create new dynamics (like pairing the Jelly Queen AI with puffles in Candy Crush Jelly), creating emotional wow moments, or introducing new modifiers like wind direction in Farm Heroes.

From Concept to Layout

After establishing a level concept, sketch the board layout on paper to visualize how the concept will work spatially. This quick iteration helps identify problems before building in the level editor and is useful during meetings to maintain focus.

The Level Creation Process

The typical level creation workflow is: translate layout from paper to the level editor, place objectives (e.g., clear ice), place game objects and power-ups, place blockers to add challenge, then run the level. Most levels break at this point because combining elements opens up unexpected gameplay possibilities.

Finding the Fun Through Experimentation

When a level breaks during testing, it's because combining mechanics creates unexpected gameplay dynamics. Finding the fun requires understanding moment-to-moment (or turn-to-turn in turn-based games) gameplay and how each action leads to the next, rather than relying on a single concept or layout.

Level Flow: The Invisible Hand

Level flow is how designers keep players moving forward toward the goal without making the guidance obvious. The level should start with limited options (one possible move) and gradually open up, giving players the illusion of choice while still pushing them forward. The less aware players are of this guidance, the more fun they have.

Engineering Flow in Turn-Based Games

In turn-based casual games, flow is engineered by starting with linear, obvious moves (like a single 3-match to activate a cascade), then gradually introducing complexity. For example, clearing clouds to reveal power-ups with a 30% spawn chance creates natural progression toward the next goal without heavy-handed guidance.

The Hint System as Last Resort

When subtle flow design fails, the hint system (hand of God) highlights obvious moves to keep players progressing. This is a safety net for when players miss the intended path, but overuse breaks immersion.

Building Progressions with Level Libraries

Level libraries (Excel sheets) catalog all levels with their game modes and blocker combinations. This overview makes it easy to create balanced progressions by selecting levels with specific properties (e.g., all levels with clouds, or clouds combined with another blocker) and arranging them into beat charts.

The Rollercoaster Experience

A good progression should feel like a rollercoaster with highs and lows, not a straight line. Vary difficulty, level length, and content introduction to create emotional peaks and valleys. This keeps players engaged and prevents monotony.

Ideal progression curve vs. monotonous progression

Rhythm: Pattern, Frequency, and Intensity

Rhythm expresses the pattern, frequency, and intensity of events in a progression. Vary game modes, alternate between hard and easy levels, mix short and long levels, and introduce new content strategically to maintain variety and prevent player boredom.

Ki-Shoten-Ketsu: The Four-Act Structure

This Japanese comic structure (introduce, develop, twist, conclude) applies to level progressions. Introduce a mechanic, develop it with variations, twist it by combining with other elements, then conclude by testing all learned concepts together. This creates a satisfying narrative arc within episodes.

What Makes a Level Easy or Hard

Difficulty is subjective and varies between designers. Obvious factors include number of objectives, number of moves, and star thresholds. Subtle factors include number of blockers on the board and number of colors available (fewer colors increase cascade chances, making levels easier).

Factors affecting level difficulty

Flow Theory and Optimal Challenge

Flow theory states that players perform best when their abilities match the challenge level. Too easy leads to boredom; too hard causes anxiety. However, deliberately breaking flow occasionally with easier or harder levels creates emotional variety and prevents monotony.

Designer Intuition vs. Data

Designers often misjudge difficulty based on intuition alone. A designer might estimate a level requires 8 attempts, but data shows 24 attempts. This gap shows why testing with real players is essential—designer intuition about difficulty is frequently skewed.

Example of designer intuition vs. real data

The Testing Cycle at King

Testing follows a structured progression: designers self-test 15-20 times, internal team testing with other designers, qualitative user testing with feedback, and finally soft-launch testing with thousands of players for quantitative data. This cycle repeats continuously for live games.

Magician Meets Mathematician

Level designers need both intuitive design sense (the magician's gut) and analytical rigor (the mathematician's data analysis). Balancing creative instinct with hard metrics ensures levels are both fun and properly tuned.

Key Metrics for Balancing

Three main metrics guide balancing: player drop-off curves (percentage of players progressing through levels), cumulative attempts (total lives spent to reach a level), and average attempts per level (difficulty of individual levels). Steep drop-offs indicate difficulty spikes.

The Level 65 Case Study

Candy Crush's infamous Level 65 started as the hardest level in the original game. Players complained extensively ('this level just sucks'). Through iterative tweaking—removing chocolate, reducing frosting—it was transformed into a fun level. This demonstrates how data-driven iteration can fix broken levels.

Level 65 transformation through data-driven iteration

Continuous Optimization

Live games require perpetual testing and balancing. As new players join and old players progress, metrics shift. Designers must continuously monitor data and adjust levels to maintain optimal difficulty and retention for both new and experienced players.

Randomness as Core Mechanic

Most casual games at King include randomness in their core mechanics. Chain reactions in Blossom Blast and cascades in Candy Crush create inherent unpredictability. This randomness is a feature, not a bug, and must be managed rather than eliminated.

Semi-Deterministic Approaches

To balance randomness, designers use semi-deterministic methods like giving objects a probability of spawning as power-ups (e.g., 30% chance). This provides some control while maintaining unpredictability. Different games use different methods—some use seeding systems to filter good vs. bad random states.

Preventing Unwinnable and Overpowered States

Designers must avoid two extremes: unwinnable states (no valid moves after a few turns) and overpowered states (cascades that never stop). Subtle mechanics like introducing new colors to stop cascades or ensuring move availability prevent these extremes while maintaining challenge.

Team Size and Specialization

Candy Crush has had four dedicated level designers per team. Smaller games like the speaker's current project have three designers who all do level design. As games scale to thousands of levels, individual designers specialize in different episodes or content types.

Team size varies by game scale

Progression Oversight and Difficulty Budget

Early in development (first 50-100 levels), one designer often oversees the entire progression to ensure coherent difficulty curves. As games scale to thousands of levels, oversight becomes impractical. Instead, teams use a 'difficulty budget' per episode—ensuring the right mix of easy and hard levels without absolute difficulty increasing too steeply.

Critical Early Levels

The first 50 levels are crucial because they determine whether players continue or drop the game. Designers are particularly meticulous about early progression. Later levels become more self-contained, with each episode having its own highs and lows rather than contributing to an absolute difficulty curve.

Linear Progression as Strong Retention Driver

Linear progression (moving from level 1 to 1000+) is proven effective for casual audiences because it's easy to understand and provides constant forward momentum. However, King is experimenting with nonlinear progressions like branching paths and open-world designs.

Challenges of Nonlinear Design

The biggest obstacle to nonlinear progression is UX—players accustomed to linear progression may not understand branching paths or choice-based systems. While there's strong motivation to explore alternatives, King is still searching for the best way to present nonlinear progression to casual audiences.

Designer Motivation and Avoiding Burnout

Creating thousands of levels can lead to monotony. King mitigates this through internal transfers, allowing designers to move between projects or teams. Some designers specialize in content creation; others prefer variety. The company's size enables flexibility to keep designers engaged.

Retention First, Monetization Second

Level design prioritizes retention and fun first. Monetization is considered later through KPIs like retention, progression, difficulty, and monetization metrics. The philosophy is that good retention naturally leads to good monetization, not the reverse.

Balancing Fun and Monetization

Designers start by building fun levels, but sometimes levels are too easy or repetitive. At this point, monetization considerations may influence design. However, the speaker cannot fully discuss King's specific monetization strategies due to company confidentiality.