Content Sources & Methodology
krilliongame.org is an independent, fan-made practice edition of the Krillion format. This page documents exactly where our content comes from, how it is produced and verified, and what our data deliberately does not claim to be.
1. Original Question Bank, Not Official Data
Every question and answer on this site comes from a question bank we wrote ourselves. As of 2026-11-19 the bank holds 108 prompts carrying 3798 graded answers across 11 categories. The original 40 prompts that launched the site on September 17, 2026 are preserved unchanged in a frozen core list, so archived dives reproduce exactly what players faced on their date.
We do not copy the official game. krillion.io publishes its own daily prompts and answers; our bank is a separate, original set written for practice. Where the format overlap is obvious — “Name a …” prompts, six rarity tiers — that is the shared genre, not shared content. We are not affiliated with, endorsed by, or sponsored by the makers of Krillion.
What we deliberately do not claim: our tier grades are not player statistics. A “Rare” label means our editors judged that only a minority of players would think of the answer under time pressure — it is editorial judgment applied consistently, not measured response data from a live audience.
2. How Prompts Are Written and Graded
Each prompt follows the same pipeline. First, a category is chosen and the prompt is written as an open “Name a …” question with a large, verifiable answer space (countries, species, brands, everyday objects). Second, answers are collected per tier: the five most famous members of the family anchor Plankton; try-hard picks anchor Too Clever; the common middle is Schooler; the uncommon middle is Rare; genuinely obscure members are Deep Cut; and the rarest few — often micro-states, deep-sea species or vanished brands — are One in a Krillion.
Third, every answer gets normalised variants in an alias map (capitalizations, common spellings, abbreviations), so the answer checker recognises “UK”, “Britain” and “United Kingdom” as the same accepted answer. Finally a build-time validator rejects any prompt missing a tier, duplicate answer or duplicate identifier before it can reach players.
3. Deterministic Daily Dives and a Frozen Archive
The daily dive is produced by a seeded shuffle: the UTC date is the seed, the algorithm deals 7 prompts, and everyone playing that day sees the same set in the same order. No server chooses your prompts — the same static page computes your dive locally, which is also why the site works without an account.
Archived dives are frozen records, not retro-calculated ones. When we expanded the bank, we cut over at a fixed date (November 20, 2026): dives before that date draw only from the frozen 40-prompt core, so every archive page shows the prompts that were genuinely served on its date. The archive starts at Dive #260 (September 17, 2026) — the first dive this site served — and we do not back-fill earlier dates.
4. Limits, Corrections and Updates
Known limits: tier grades are editorial, not statistical; alias maps cover common variants but cannot cover every phrasing; and archived sheets reflect the bank as it stood on each date rather than any later improvements. Where a prompt family clearly outgrows its grades, we revise it — but past archives keep their original sheets on purpose.
Corrections are welcome and human-reviewed. Write to [email protected] with the prompt text, the answer you entered, and what you expected to happen (rejected answer, wrong tier, wrong alias). Verified fixes ship with the next site build, and major bank expansions are recorded in our public development log.
For how we handle visitor data, see the privacy policy; for the site itself, see about krilliongame.org.
Content Sources – FAQ
Where do the Krillion questions and answers come from?
Every prompt and every graded answer on this site is written and tier-rated by our own team. We do not scrape, copy or mirror the official Krillion game's daily data. That is why our bank — 108 prompts carrying thousands of graded answers — feels different from the official game rather than identical to it.
How are answers graded into rarity tiers?
Each answer is graded by instinct frequency: what almost everyone says first is Plankton, cheeky try-hard picks are Too Clever, common secondary guesses are Schooler, uncommon-but-reachable answers are Rare, genuinely obscure answers are Deep Cut, and the rarest few are One in a Krillion. Grading is editorial judgment, refreshed when a family clearly outgrows its tier.
How is the daily dive selected?
A deterministic algorithm seeded by the UTC date deals the same 7 prompts to every player on the planet for that day, with 25 seconds per prompt. Past dives are frozen: their exact prompt sets can always be reproduced, which is why our archive pages match what was actually served on each date.
Can I request a correction?
Yes. If an answer you are sure of is rejected, or a tier feels plainly wrong, write to [email protected] with the prompt text, your answer, and what you expected. Corrections are reviewed by a human and the question bank is updated in the next build.