The site
Data & Editorial Policy
This page explains where answer data comes from, how it is checked before publication, and what this site adds on top of it.
How data moves from import to publication
- 01
Official reveal import
New published Krillion reveals are retrieved automatically and loaded into our database.
- 02
Schema and consistency validation
Every import must pass deterministic checks: a question needs text, accepted answers, and a fixed tier and score. A tier that disagrees with its score is rejected.
- 03
Duplicate and relation checks
The same question recorded in several games is consolidated to one canonical question, repeated answers are deduplicated, and slug collisions are flagged and held back.
- 04
Derived tools and statistics
Search, archives, answer checking, rare-answer finding and the statistics on this site are computed from the validated set, and labeled as calculated by Krillion Answers.
- 05
Representative editorial review
A human checks a representative sample of published question pages for wording, category, duplicate identity and obvious data errors, and uses the findings to tune the checks above.
- 06
Corrections
Reported errors are verified against the official reveal and corrected in the published set.
Where official answers come from
Official answers, rarity tiers, scores and quips are retrieved automatically from published Krillion reveal data. We label that set as answers from Krillion’s official reveal, not as answers published by this website.
How imported data is checked
Automated imports are not published on trust. Each import runs through schema and consistency checks, duplicate and relation checks before it can appear on the site, and anything that fails is held back until it is fixed. The checks are deterministic and re-run against the whole database, so one bad record cannot quietly change a published page.
What we add
Krillion Answers adds search, filters, archives, answer checking, rare-answer finding, score calculation, hints, session statistics, database insights and strategy notes. Those layers and the statistics computed from them are our publisher-added value.
Editorial review and statistics
Automated checks catch structural problems, not meaning. A representative sample of published question pages is reviewed by a human for wording, category, duplicate identity and obvious data errors, and the findings are used to tune the checks. Statistics such as answer counts, distributions and comparisons are calculated by Krillion Answers from published reveal data and are never attributed to Krillion.
Official vs suggested answers
Suggested answers are additional ideas that have not been verified against the official answer set. They receive no Krillion tier or score and are labeled as suggested, not official. A suggestion becomes an official reveal answer only if it later matches an active official answer.
Corrections
If an official answer, score or prompt is wrong, or a page shows data that disagrees with the official reveal, use the Contact page and include the page URL plus, when possible, the relevant official reveal. Verified reports are corrected in the published set and reflected on that page.