Where the questions and scores come from
A score is easier to understand when you know what sits behind it. This independent learning edition uses its own original AI-generated question pool. It does not fetch the original Wonderhop's questions or submit your answers to its services.
An original, growing question pool
The bootstrap bank contains 100 questions: 50 everyday or nature prompts for Sky and Abyss, and 50 chronological prompts for Past and Future. Questions and accepted answers are available in English.
Each browser gets its own random seven-question daily set, saved for that date and mode. Practice draws another seven. The server prefers questions that browser has not seen in the last 30 days, then the least recently seen ones. Different players can still overlap. A scheduled AI job attempts to add seven new questions per day in total; failed batches leave the previous pool playable. The live pool may therefore be larger than this bootstrap count.
Estimated weights, not survey results
Each listed answer has a preset popularity weight. These weights total 100 for a question and are used to order or score answers. They are AI estimates prepared for the prototype. We have not surveyed a representative group or collected player responses to calculate them.
For example, a displayed 30% means a weight of 30 in that question's game data. It does not mean 30% of visitors, families, or people worldwide gave that answer. Sky favors higher weights; Abyss favors lower ones. The mode guides explain the different scoring formulas.
Historical years need careful definitions
Past and Future compare stored reference years. Generated chronology questions use one date convention per question, such as first commercial release, first publication, official opening, or launch. Their years have not completed per-answer source review. It should not be treated as a historical reference.
A reviewed chronology question needs a clear event, an answer source, and a consistent date policy. A public announcement, first working prototype, and first commercial sale are different milestones. Our what-came-first resource names its events and links the references for those examples separately.
How new answers are checked
The browser first checks the names and aliases supplied with the seven-question round, with conservative tolerance for small English spelling errors. Unmatched answers go to our server, which checks the stored round and can ask the configured AI judge. The timer pauses during that request. A refusal or service failure lets you try another response with your remaining time restored.
The judge may recognize a synonym of a listed answer or suggest a new English answer and estimated weight. New AI answers are labeled as estimates. A chronology answer also needs a year and a stated date basis. Models can make mistakes, and uncertain or ambiguous responses may be refused. New answers are cached temporarily and saved with your round; they do not automatically become part of the published question bank.
The bootstrap weights total 100 per question. A newly accepted AI answer has a weight relative to those examples, so the expanded list need not total 100. Neither kind of weight is a measured survey percentage.
Sources we can build on
- Wikidata structured data is published under CC0. It can help identify entities and dates, with reference and precision checks. It supplies no popularity distribution for this game.
- Open Trivia DB offers a documented multiple-choice API under CC BY-SA 4.0. It is not connected here and would suit a separate trivia format with the required attribution.
- Official release histories and mission pages can verify clearly defined dates. Linked examples in our timeline guide use those sources; they do not imply the entire game bank has been verified.
What stays in your browser
Your nickname, preferences, in-progress round, and recent scores are saved locally. A first-party cookie keeps your daily question selection stable. The server stores hashed browser identifiers, versioned question releases and short-lived judge decisions, without account synchronization or a global leaderboard. Read the privacy note for answer processing and storage details.
For the rules behind each result, see the Sky, Abyss, and Past & Future guides.
BACK TO THE ADVENTURE →