Home › In live-ops, an A/B TEST measures the impact of …
In live-ops, an A/B TEST measures the impact of a change by which method?
ARunning the change only for the dev team for 2 weeks
BRandom user split A vs B; compare metrics over time
CSwitching the change on globally and watching trends
DPolling players for opinions on the design change
Answer & Solution
Correct answer: B. Random user split A vs B; compare metrics over time
Kohavi GDC + Microsoft Experimentation Platform: random allocation; identical population except the change; statistical test for lift. Causal claim possible.
Related questions
Per F2P live-ops, an ECONOMY SINK (vs source) is a design element that does which?In live ops + UA, the CAC payback period is which metric?In live ops + UA, COHORT ANALYSIS groups users by which property to study behaviour over tIn F2P ad monetisation, REWARDED VIDEO ads differ from interstitials in which respect?In live-ops design, a BATTLE PASS combines which two engagement mechanics?Per public GDC talks, the FREE-TO-PLAY WHALE / dolphin / minnow segmentation describes whiIn F2P monetisation taxonomy, a GACHA / LOOT BOX mechanic differs from a direct shop purchPer Eric Seufert + GDC F2P talks, LIFETIME VALUE (LTV) of a user is most commonly approxim