A/B testing is essential for data-driven game development, but running effective experiments in games requires careful planning. I'm interested in what aspects of your game you test most frequently - UI elements, difficulty curves, monetization offers, or new features? What sample sizes and test durations do you find necessary for statistically significant results? How do you handle the complexity of multivariate testing when multiple systems interact? What tools do you use to manage experiment deployment and analysis? Have you encountered situations where A/B tests showed clear winners but implementation failed? How do you segment test groups to ensure fair comparison while accounting for player skill levels or engagement tiers? What are your biggest challenges with A/B testing, and what surprising insights have tests revealed about player preferences?