StarCraft II's ranked ladder was one of the game's core competitive systems, defining how millions of active users measured skill, tracked progress, and earned recognition. As Lead UX Designer, I led the user experience for the ranking system, partnering with engineering and data science to shape each major iteration. We replaced the traditional numeric rating model with seven visual leagues, grounding competitive progression in visual identity while making rank easier to understand, celebrate, and aspire toward.
Users progressed through seven visual leagues—Bronze, Silver, Gold, Platinum, Diamond, Master, and Grandmaster—creating a clear path from beginner to the highest levels of competition.
StarCraft II's ranked play was experiencing churn. Research revealed distinct user populations with different needs. Highly skilled accounts required rankings that accurately reflected capability, while less experienced cohorts needed enough positive reinforcement to sustain engagement.
Qualitative research, quantitative analysis, and cross-functional collaboration shaped each iteration before active user telemetry validated the results.
Each major release tested a different hypothesis. Findings from one model informed the next, moving the ranked system from population-based distribution to skill-based progression and ultimately to behavioral segmentation.
The initial design distributed the first five tiers evenly by volume. Each tier held 20% of active accounts, with Master and Grandmaster reserved for the top 3%. Feedback was initial positive, finding the distribution clear and easy to understand. This established the baseline.
Qualitative feedback and MMR (Matchmaking Rating) analysis revealed that the league distribution no longer aligned with the underlying skill curve. Bronze and Silver contained too many accounts while Gold had too few. As proficiency improved, the league structure wasn't providing progression that reflected actual development.
I redesigned the league distribution around the skill curve, reducing the size of Bronze and Silver while expanding Gold.
Even after the skill-based redistribution, retention issues persisted in Bronze and Silver. Telemetry analysis pointed to a systemic pattern: different skill tiers possessed fundamentally different motivations and responded to distinct feedback loops, meaning skill grouping alone was insufficient.
I led the telemetry analysis that identified four behavioral cohorts.
Motivated by competitive accuracy. Rank needs to reflect true skill rating, and even minor rating inflation erodes trust in the ladder for this segment. Master and Grandmaster, top 5%.
Value meaningful skill progression. Advancement milestones sustain long-term engagement, so the UI surfacing explicit progress indicators when a tier boundary is within reach. Platinum and Diamond.
Broadly satisfied without being driven to advance. Progression indicators are intentionally calibrated to maintain competence sentiment without forcing tier advancement pressure. Gold.
High churn risk. This segment responded to victory milestones rather than rating precision, leading to a wins-based model where progression occurs through match victories rather than pure MMR calculations. Bronze and Silver.
The final system gave each behavioral cohort a distinct progression model. Casual tiers leveraged wins-based advancement rather than strict rating precision, which required decoupling lower-tier advancement logic from global match rating. Upper tiers remained strictly skill-based, maintaining integrity for high-end competition. Gold anchored 40% of the active base at the center of the distribution curve.
Quantitative results measured against the prior major release and validated by internal data science teams.
Recognition
I was invited as a guest speaker to the Wharton School of Business by Professor Kevin Werbach for his course "For the Win," on game design principles and behavioral motivation systems.