This question and answer was more of a back and forth conversation, but the main things that I covered included: A/B testing is a method where different versions of a feature are shown to user segments to determine which performs better. The process involves forming a hypothesis, creating control and variant versions, randomly assigning users, collecting data, analyzing results, and implementing the winning version if conclusive. When designing A/B tests, focus on defining clear objectives, determining proper sample sizes, planning appropriate test duration, and considering user segmentation. Implementation requires robust technical infrastructure, careful user experience considerations, and maintaining group isolation. Analysis should employ proper statistical methods while avoiding p-hacking, examine results across different segments, and evaluate both short and long-term effects on multiple metrics. When acting on findings, have a clear decision framework, develop a thoughtful rollout strategy, document learnings, and commit to continuous iteration. For King specifically, consider impacts across their game portfolio, effects on cross-promotion, and balance between gameplay integrity and monetization, while focusing on metrics relevant to match-3 games.