In a live online game such as MapleStory, one in-game item suddenly starts getting used far more than usual. How would you investigate what is happening, identify the root cause, and decide whether the team should take action? ### Constraints & Assumptions - Do not assume the spike is good or bad before diagnosis. - Verify instrumentation before acting. - Consider gameplay balance, economy health, monetization, exploits, bots, player sentiment, and live-ops events. - Include both immediate mitigation and longer-term monitoring. ### Clarifying Questions to Ask - Did total uses, unique users, uses per user, purchases, trades, or acquisition volume spike? - Over what time window and relative to what baseline? - Did a patch, event, bundle, promotion, or content release occur recently? - Is the item consumable, tradable, power-affecting, cosmetic, or progression-related? ### Part 1 - Verify And Segment The Spike How would you confirm the anomaly and narrow where it is happening? #### What This Part Should Cover - Instrumentation checks and baseline comparison. - Segmentation by server, region, platform, player level, cohort, acquisition source, item source, game mode, and time. ### Part 2 - Build Hypotheses What root causes would you consider? #### What This Part Should Cover - Intended patch/event effect, meta shift, price/supply change, promotion, creator/community influence, bug, exploit, botting, or logging issue. ### Part 3 - Validate With Data And Partners What data and teams would you use? #### What This Part Should Cover - Analytics, game design, economy design, engineering, anti-cheat, customer support, community, and live-ops. - Patch notes, source/sink trends, marketplace data, retention, conversion, suspicious behavior, and player feedback. ### Part 4 - Decide Whether To Act How would you decide what action to take? #### What This Part Should Cover - Action matrix for intended positive spike, risky balance/economy issue, bug/exploit, bot/fraud, or low-impact cosmetic change. - Guardrails and communication plan. ### What a Strong Answer Covers - Diagnoses before reacting. - Separates product opportunity from abuse or data-quality issues. - Balances player value, economy health, monetization, and trust. - Gives a clear decision framework. ### Follow-up Questions - What would make you hotfix immediately? - What if revenue rises but player complaints rise too? - How would you detect bot-driven usage? - How would you communicate a nerf? - What daily dashboard would you monitor after the spike?