[go: up one dir, main page]

Page MenuHomePhabricator

[Epic] WikiGames - WikiTrivia on the Apps Experiment
Open, HighPublic

Description

Background

Annual Plan KR
Staff do a lot in their regular work to make our operations faster, cheaper, and more impactful. This objective highlights specific initiatives that will both: a) make substantial gains toward faster, cheaper, or more impactful; and b) take coordinated effort and change of formal and informal practices at the Foundation. Essentially, the KRs included in this objective are the hardest and best improvements we can make this year to operational efficiency of work touching our products and technology.

Theory of Change
Run and conclude two experiments from existing exploratory products/features by staff to that provides us with data/insights into how we grow Wikipedia as a knowledge destination for our current consumer and volunteer audiences in Q1 and Q2. Complete and share learnings and recommendations for potential adoption for future OKR work in the Wiki Experiences bucket by the end of Q3.

Hypothesis

If we create Wikipedia-based games for daily use that highlights the connections across vast areas of knowledge, it will encourage consumers to visit Wikipedia regularly and facilitate active learning, leading to longer increased interaction with content on Wikipedia.

Game Requirements
  • Daily play: A game that invites players to return daily to play (new challenge each day)
  • Stats
    • Required: Make it easy for players to see how they are performing with their streak of correct answers
    • Nice to have: How their score compares to other players
  • Quick to complete: Completing game play should be relatively quick (under 5 minutes)
  • From Wikipedia: The game should be based on on this day Wikipedia content
  • Not a contribution: game Focus should be on gameplay vs. gamified contributions
  • Shareability: Encourages sharing results or game play
  • Availability and access: Easy to access via web and / or download as an app
  • Language Evaluation: Game should be available in EN and DE
Scope
  • Simple enough to be created within a sprint
  • Game only needs to be available for 30 day experimental period
  • Should be available in EN + DE in production app
  • Prompt users to “Try a game” without detailing what it is the first time, then show game
Game Mechanics
  • 1 multiple choice trivia question per day, shared at 00:01 in local timezone
  • Players must choose which event(s) occurred on this day in history
  • Streak is based on consecutive days of participation/play
  • Nice to have? Comparative stats are based on number of questions answered correctly
  • For the time limited experiment, multiple choice questions and choices can be human curated
  • Shareable on socials: what happened on this day with Wikipedia Trivia Game branding and note to download the app
  • 1-2 Question Survey at end of period (would they like games to be permanent? Other game ideas)

Stretch Goals

  • Time to answer correctly
  • Archive availability to review past questions
  • Daily notifications or alert from the app
  • Utilize the More like this API for multiple choice options
Potential Data Sources
  • Wikipedia:Selected Anniversaries
  • List of days of the year
  • OTD helper
  • More like this
Metrics

Quant

  • Impressions & CTR to game
  • Game retention over 30 days
  • DAU changes per unique that engage with feature
  • Player retention vs All users retention
  • Abandonment point (Churn)
  • Shares
  • Completion Rate (All 3 questions are answered)
  • Difficulty Score of questions

Qual

  • Interest in games
  • Desire for other game times
  • Quality of suggestions
  • Reports of offensive content
References

Event Timeline

Restricted Application added a subscriber: Aklapper. · View Herald Transcript
cmadeo updated the task description. (Show Details)
cmadeo updated the task description. (Show Details)

Feedback from game designer Francois:

  • The multi-choice is fine, but it's also quite similar to other trivia games. It seems to just keep going, so the loop isn't really closed
  • The range/slider feels more interesting. I wouldn't have it cover the full range for each question, more like the range of the multiple choice +/- a few years on either side. The point scale should >probably be something a bit non-linear adjusted to this range, so I still get some points if I'm in the right vicinity, and therefore feel smart. Same comment about closing the loop
  • Which event feels much harder for me, and it's very binary outcome. Cool idea to try, but I don't think it's a success
  • Sort events is quite interesting idea. A UX thing is it should be more clearly communicated that top is oldest. This was surprisingly fun. The loop isn't closed, but I enjoyed this quite a bit. I >wonder about variations with more events. The UX on the reshuffle to correct is a bit weird and confusing. I wonder if it would make more sense to arrange the events on a left-right timeline, and >then the correction would draw lines to when they should be relative to the others.