Question policy
Clear questions. Checked answers.
Every question defines the collection it is asking about. Its answer list, aliases, explanations, and exclusions are checked against published sources before release.
A boundary you can read
A country question names its country list or geographic definition. A sports question fixes an event or year. A book question identifies the series and edition differences where relevant. These conditions appear beside the question.
We favor finite lists over vague categories. “Name a famous scientist” has no clear edge; “Name one of NASA’s original Mercury Seven astronauts” does.
Sources, then explanations
We use institutional lists, standards, official historical records, and specialist reference material where available. Source links are available with each question’s feedback. Question wording and explanations are written for this game; source links support the facts.
Each list is checked for its stated scope, not just filled with a few plausible examples. Classification or naming changes can still require corrections. We keep a dated source audit with the project.
Aliases are explicit
Matching ignores capitalization, repeated spaces, supported quote and hyphen variants, and full-width English characters. Other spelling changes must be listed as checked aliases for that question.
These text checks apply to Recall challenge. In Multiple choice you click a checked label directly, so spelling is not part of the test.
We do not use an online model, fuzzy guesses, or automatic removal of every accent and punctuation mark. For example, a shorter abbreviation is accepted only when it is recorded for that particular question.
Multiple choice options are checked too
Every four-option set includes a 25-point and a 100-point correct answer, plus one or two verified exclusions from that question. If there is one exclusion, a middle-tier correct answer fills the fourth position. Options are shuffled once for the round and stay fixed while you decide. More than one can be correct; you submit only one.
Points stay hidden until you submit. Guessing from options differs from recalling an answer, so Multiple choice and Recall challenge keep separate best scores.
What “Rare” means here
Scores are editorial tiers, not live player statistics. Familiar, Less common, Uncommon, and Rare earn 25, 50, 75, and 100 points. Editors use familiarity within the question as a judgment; source lists do not measure how many players know an answer.
Every question has a 25-point and a 100-point answer. There is no speed bonus. Changes to scoring rules use a new rules version so on-device best scores are not compared across different rules.
An unlisted answer is unresolved
If your input matches neither an accepted answer nor a checked exclusion, you see Not in our list. It earns 0 points in this round, but is not marked factually wrong. An actual exclusion has a specific reason.
If you think we missed a valid answer, send the question ID, the proposed answer, and a reliable source through the contact page. Corrections can update later versions; completed local scores are not recalculated.
Local scores and open data
The game’s answer data is downloaded to your browser. It is not secret, and browser scores can be modified. This is a casual learning game with no prizes or trusted competitive leaderboard.