Krillion Hints – How to Find a Rarer Answer
Real Krillion hints: how the tiers are decided, why a clever answer scores zero, and how to find a rare answer inside the 25 seconds you get per prompt.
Real Krillion hints: how the tiers are decided, why a clever answer scores zero, and how to find a rare answer inside the 25 seconds you get per prompt.
Most hint pages for this game give you one word and hope. The problem is that the game does not reward cleverness — it rewards being inside the answer set. So the useful hint is not a word, it is a way of thinking that keeps your answer inside the set while pushing it away from the obvious.
Your answer scores by tier, and the tier only exists if the answer is in that dive's set. An answer that is not in the set scores zero — not a low score, zero. This is why the most common regret in the game is not "I picked the obvious answer", it is "I picked something genuinely obscure that the set had never heard of".
So hints should be filters, not suggestions.
Dinosaur, Colorado and DISH, Texas both score 100 on a prompt asking for a city starting with D. "Dunmore" — a real town — is the kind of answer the set may not carry. Weird proper nouns beat plain nouns.A hint cannot make a missing answer appear. We keep one recorded snapshot per dive, and the set keeps growing while the dive runs — an answer that scores for you in the evening may not be in our snapshot, and an answer missing from our list is still worth trying.
Our practice mode takes a prompt and an answer and returns the tier and score from the current snapshot, so you can test the risky version of your answer without spending your single shot. For the current dive's answers by tier, start at today's answers.
Prompt types do not behave the same way. A city prompt and a helmet-sport prompt can be fifteen times apart in how likely a 100-point answer is, because the size of the answer set drives everything. These five pages take one prompt type each and work through the real numbers from the 18 November dive:
When a dive brings a prompt type not covered here — animals, for instance, which has not appeared in a dive we have recorded yet — a page gets built with that dive's real numbers rather than with guesses. Prompts repeat across dives, so every category page gets more evidence over time.