Diagnose rising account switching and falling actives

Yesterday you noticed that account switching among users went up even as total active users went down. Lay out a concrete diagnostic plan. It should include: (a) 5–7 plausible hypotheses covering measurement artifacts (for example, deduping by person_id versus user_id, SSO changes), product changes (login/logout friction, session expiration), and behavior shifts (multi-account consolidation, spam mitigation); (b) the specific metrics and cuts you would pull (by client, geo, new versus returning, person_id versus user_id, cohorts, time-of-day); (c) at least three focused slice analyses to confirm or rule out cross-contamination between accounts; (d) event/experiment diffing to spot recent launches that correlate with the shift; (e) a minimal SQL or pseudocode outline for measuring how many active user_ids map to the same person_id relative to last week; and (f) the immediate mitigations you would try if the root cause is (i) a logging bug, (ii) authentication friction, or (iii) users rotating accounts to game policy limits. Overview: This question tests a data scientist's ability in diagnostic analytics, experiment diffing, cohort and identity-resolution analysis, metric instrumentation, and SQL-based quantitative validation.