Scoring reference
How Krillion Scoring Works
Six tiers, 10 to 100 points, and one trap built into the middle of the scale.
A round of Depth Engine is seven prompts. Every prompt has a hidden answer table, and every answer in that table is tagged with a rarity tier. The tier is not a measure of how difficult the fact is — it is a measure of how likely a room full of players is to say it. That distinction is the whole game, and it is why the same prompt can feel easy and still pay badly. You can Play Krillion Infinite Free in another tab if you want to see the tiers land before reading the rest.
The six tiers
The reflex answer. Most players type it before they finish reading the prompt.
The answer people pick to feel smart — which makes it the second most common one.
Correct, well known, and just off the front of the queue.
You had to think, and the thinking worked.
Requires real domain knowledge. Most rooms produce one of these, not seven.
The answer almost nobody reaches. Correct, specific, and unshowy about it.
From points to metres
Depth is simply points × 10. A Plankton answer buys 100 m, a Krillion answer buys 1,000 m. Seven prompts with a perfect run is 7,000 m, which is the hard ceiling of the scale — there is no eighth tier and no bonus multiplier. The meter on the play screen is the same number, just drawn as a water column.
The scale is deliberately non-linear. The jump from Schooler (30) to Rare (60) is worth twice as much as the jump from Plankton (10) to Too Clever (15). A round of safe answers tops out around 1,400 m. A round where you take four real swings and miss two still beats it.
The Too Clever trap
Too Clever sits at 15 points: one tier above the obvious answer, and half of what a solid Schooler answer pays. That placement is intentional. It is the tier for answers that feel like a flex and turn out to be the second most popular thing in the room — the obscure-sounding capital that every geography fan says, the cult film that is on every “underrated” list.
The feedback is built to sting a little. When you land on Too Clever, the answer tag flashes once, drains to grey, and the panel adds a single line: “Everyone thought of that.” It is the most useful screen in the game, because it is the exact moment your instinct and the crowd's instinct turn out to be the same instinct.
How an answer is matched
Matching is exact after normalisation. Your input is lowercased, trimmed, stripped of punctuation, and collapsed to single spaces. Hyphens and apostrophes inside a word are kept, so “Spider-Man” and “spider-man” both resolve, but “spiderman” does not. There is no fuzzy matching and no synonym expansion in v1 — if an answer is not in the table, it scores zero.
That is a real limitation, and it cuts both ways: a brilliant answer that nobody wrote down is worth nothing, and a lazy answer that happens to be in the table pays full price. The table grows every week, which is the only honest fix.
Why some prompts pay better than others
Tier assignments are per prompt, not global. “Brazil” is a Plankton answer to name a country in South America and a Rare answer to name a country that borders France, because the second prompt makes you remember that France still holds a border in South America. The word does not carry a rarity. The prompt does.
This is also why reading the prompt carefully is worth more than knowing a lot of trivia. The prompt is telling you which corner of your memory to search.
Ready to test it? Try the free endless mode — no signup, no unlock, seven prompts per dive.