Determine Player Preference for Local Game Creators

Quick Overview

Roblox analytics prompt on measuring player preference for local creators, covering engagement metrics, availability adjustment, observational causal methods, selection bias, matching, fixed effects, and guardrails.

Determine Player Preference for Local Game Creators

Company: Roblox

Role: Data Scientist

Category: Analytics & Experimentation

Difficulty: hard

Interview Round: Onsite

##### Scenario Roblox analytics team wants to verify whether players prefer games created by local creators. ##### Question Define primary and secondary metrics to measure player preference for local creators. If an A/B test is impossible, outline an analysis plan that credibly estimates the causal effect while mitigating selection bias. ##### Hints Discuss time-per-session, sessions count, matching, IV, DID.

Overview: Roblox analytics prompt on measuring player preference for local creators, covering engagement metrics, availability adjustment, observational causal methods, selection bias, matching, fixed effects, and guardrails.

|Home/Analytics & Experimentation/Roblox
Roblox logo
Roblox
Jul 12, 2025
hardData ScientistOnsiteAnalytics & Experimentation
32
0

Player Preference for Local Game Creators

A Roblox analytics team wants to understand whether players prefer games created by local creators, such as creators who share the player's country, region, or language.

Assume you can label each game session or impression with whether the creator is local to the player, and you have standard engagement and retention telemetry.

Constraints & Assumptions

  • Define "local" explicitly and discuss data quality for location and language.
  • Separate preference from availability and ranking exposure.
  • If an A/B test is infeasible, propose observational methods and state assumptions.
  • Address selection bias: players may choose local games for reasons unrelated to locality.

Clarifying Questions to Ask Guidance

  • Is locality based on country, region, language, or creator community?
  • Are local games equally available in every market?
  • Are local games already favored by ranking or discovery surfaces?
  • What business decision depends on the analysis?

What a Strong Answer Covers Guidance

  • Primary metrics such as local-game CTR, playtime per session, completion/return rate, retention, or availability-adjusted local preference index.
  • Secondary metrics such as sessions per player, D1/D7 retention, spend, social co-play, creator outcomes, diversity, and satisfaction.
  • Guardrails for overall engagement, content quality, creator displacement, latency, and safety.
  • Observational plan: matching, within-user comparisons, fixed effects, propensity scores, difference-in-differences, instrumental variables, or natural experiments.
  • Controls for rank position, surface, predicted game quality, player tenure, device, time, region, language, and social graph.
  • Validation: covariate balance, placebo tests, sensitivity analysis, pre-trends, overlap checks, and triangulation across methods.
  • Clear statement that observational evidence is weaker than randomized evidence unless assumptions are credible.

Follow-up Questions Guidance

  • How would you distinguish local preference from local content supply?
  • What instrument might affect local exposure without directly affecting engagement?
  • What if local games improve retention but reduce creator diversity?
  • How would you design a future randomized test?
Loading comments...