GDC Festival of Gaming is part of the Informa Festivals Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

March 1-5, 2027
Moscone CenterSan Francisco, CA

Agenda

Demystifying Collaborative Causal Inference for Game Development

Drew Vande Guchte  (Principal Insights Analyst, Riot Games)
Pass Type: Academic Pass, Festival Pass, Game Changer Pass, Indie Pass, Start-Up Pass - Get your pass now!
Track: Business Strategy, Production & Team Leadership
Audience Level: Entry-Level
Format: Lecture
Vault Recording: TBD
Audience Level: Entry-Level
Did a new battle pass actually improve long-term retention? Does a new game mode create incremental engagement or simply cannibalize other parts of the game? When A/B testing isn't possible, game teams increasingly turn to non-experimental causal inference—but methods like Propensity Score Matching and Double Machine Learning can often feel like statistical black boxes. This session pulls back the curtain on these techniques through intuitive live-service game examples, building conceptual mental models without requiring advanced mathematics. Product managers, designers, and analysts will gain a shared understanding of how these methods work, where they work best, and how to evaluate causal claims through assumptions, limitations, and validation. Attendees across product and analytics will leave with practical frameworks for selecting and interpreting causal methods, understanding their tradeoffs, asking better questions, and collaborating more effectively to make stronger product decisions.

Takeaway

Attendees will understand causal inference not as opaque algorithms, but a multi-discipline framework for tackling difficult product questions when A/B tests aren't possible. They'll leave with an actionable playbook for framing causal questions, selecting appropriate methodologies, and critically evaluating and interpreting results with confidence across product, design, and analytics.

Intended Audience

Designed for product managers, designers, analysts, and data scientists working on live-service games, this session benefits anyone who evaluates product decisions using player data. No prior knowledge of causal inference or advanced statistics is required; the presentation builds intuitive conceptual understanding through practical game development examples.