Krillion Strategy: 9 Ways to Find Rare Answers & Score Deep

Quiz brain asks: “What is the right answer?” Survey brain asks: “What will most people say — and what is one step past that?” Krillion only rewards the second question. Every prompt is really a prediction market about other players’ first instincts, and your edge is knowing those instincts better than they do.

Here is the mental shift in practice. Quiz brain hears “name a dog breed” and proudly produces “Labrador” — correct, instant, and worth exactly 1 point. Survey brain hears the same prompt, predicts that 60% of players will say Labrador, and keeps thinking until it lands on something like “Basenji.” Same knowledge level, seventy-five extra points.

If you take one idea from this page, take this: the prompt is not the opponent. The crowd is.

The first answer your brain offers is, by definition, your most accessible association — and accessibility is roughly the same across millions of brains. That first answer is the Plankton answer. Do not type it. Write it down mentally, thank it for showing up, and keep going.

Worked example — prompt: “Name a dog breed.” First answer: Labrador. That is the answer to beat, not the answer to give. Second wave: Golden Retriever, German Shepherd — still crowd answers. Third wave, where the points live: Basenji, Shiba Inu (borderline now — it got popular), Leonberger. You are looking for the first answer that makes you think “wait, is that too obscure?” and then checking it against the too-clever rule: real breed, real people know it, just not most people.

This one habit alone moves most players from ~120 meters to ~200 meters. It costs nothing and works on every category.

Every prompt has a generality ladder, and the crowd clusters at the top. Your move is to climb exactly one rung down: more specific than the obvious, but not so specific that validity gets shaky.

Worked example — prompt: “Name a cheese.” Top rung: “cheese” is not an answer, so the crowd’s top real rung is Cheddar — Plankton. One rung down: Brie, Gouda — Schooler to Clever. Two rungs down: Manchego, Gruyère — Rare to Deep Cut. Three rungs down: Époisses — real, but now you are flirting with the too-clever line, because the bank needs enough players to have typed it.

The sweet spot is usually one to two rungs below the obvious. Practice spotting the ladder on unlimited mode: for any prompt, name the top-rung answer, then force yourself two rungs down before you commit.

For any “name a…” prompt about people, the most famous correct answer is always Plankton. Fame is the crowd’s sorting key. The second-most famous is where the tiers start getting interesting — famous enough to be valid and banked, unfamous enough that most players skip it.

Worked example — prompt: “Name a Beatle.” Paul McCartney and John Lennon split the Plankton vote. George Harrison is the classic second-famous answer — Clever to Rare. Ringo Starr, the drummer everyone forgets until reminded, is Deep Cut territory. All four are correct; the points spread is 1 to 5.

This generalizes beautifully: movies (not the lead actor, the supporting one), bands (not the frontman, the bassist), history (not the president, the vice president). Whenever a prompt has a fame hierarchy, aim one slot below the top.

Nobody else has your exact biography, which makes personal knowledge the most defensible rarity edge in the game. A dish your grandmother made, a river in your home state, a tool from your trade — these are answers with real validity and naturally thin crowds.

Worked example — prompt: “Name a pasta shape.” The crowd says spaghetti, penne, fettuccine — all Plankton or Schooler. Your Calabrian grandmother’s orecchiette? Rare or better, and you did not have to study anything — you just had to remember dinner. A plumber answering “name a tool” with “basin wrench” beats the entire crowd that typed “hammer.”

Audit your own life once: hometown geography, heritage food, job vocabulary, hobbies. Write the rare-but-valid answers down. This is the cheapest strategy on the page because the research is already done — it is your memory.

Daily game audiences skew young, and young crowds reach for recent things first. That makes older-but-valid answers systematically underplayed — a structural edge you can exploit on music, movies, sports, and history prompts without knowing anything obscure.

Worked example — prompt: “Name a sitcom.” The crowd floods Friends, The Office, Brooklyn Nine-Nine — all Plankton. Cheers, a massive hit with a thinner modern crowd, lands Deep Cut or better. Same logic: “name a pop star” — Taylor Swift is Plankton; Cyndi Lauper is a valid, banked, thinly-played answer.

You do not need to be old for this to work — you just need to know that old things exist. One documentary or one parents’-playlist deep cut per category is enough ammunition.

The tier is called Schooler for a reason: answers from standard curricula cluster there because every educated player learned the same examples. “Name a planet” → Mars. “Name a U.S. president” → Washington, Lincoln. “Name a Shakespeare play” → Romeo and Juliet. Correct, respectable, and worth 2 to 3 points.

The dodge is simple: when you catch yourself giving the textbook example, ask what the textbook’s second example was. Shakespeare → Macbeth is still crowded; Othello starts getting interesting. U.S. president → not Washington, but James K. Polk — real, banked, and rarely anyone’s first thought.

Textbook answers are also the best diagnostic: if your answer feels like it belongs on a classroom poster, the crowd is already there.

Rarity is not evenly distributed inside a category. Geography rewards small countries; movies reward supporting cast; food rewards regional dishes. Players who learn one category’s map deeply outperform players who know a little about everything.

Pick your two weakest categories and drill them in unlimited mode for a week. Then read the full category guides — each one maps where the tiers live for that category, with worked examples:

• Geography: small countries, second cities, and tributaries beat capitals — geography guide
• Food: regional dishes and heritage ingredients beat global chains — food guide
• Movies: supporting actors and directors beat lead stars — movies guide
• Sports: retired legends and niche sports beat current MVPs — sports guide
• Music: bassists, drummers, and one-hit wonders beat frontmen — music guide
• Science: the second discoverer and the obscure element beat the famous one — science guide
• Animals: uncommon breeds and deep-sea species beat pets — animals guide
• History: the vice president and the failed revolution beat the headline event — history guide

The timer is generous if you budget it and brutal if you do not. The winning split: 12 seconds thinking (run tactics 1–7), 8 seconds typing, 5 seconds verifying spelling and validity. Most players invert this — 3 seconds thinking, 22 seconds staring — and it shows in their depth.

Two clock rules. First, never lock in early: an extra 5 seconds of thinking has positive expected value on every prompt, because your third idea is almost always rarer than your first. Second, never change your answer in the final 3 seconds — that is when good answers become typos, and a typo on an otherwise Krillion answer still only scores if the meaning survives.

If you are blanking at 10 seconds, drop to your anchor answer (tactic 9) rather than freezing. A banked 30-point answer beats a 0-point blank every time.

An anchor answer is a pre-chosen, rare-but-valid response for a common prompt shape — your emergency ripcord when the clock is eating you. The best anchors are answers you have already seen land in Rare or Deep Cut during unlimited practice, so there is no guesswork.

Build one anchor per category this week. Test each in unlimited mode first — an anchor is only an anchor if the tiers confirm it. Below are starter shapes to get you going; replace them with your own finds as your notebook grows.

Prompt one: “Name a movie starring Tom Hanks.” Thinking trace: Forrest Gump surfaces first — Plankton, rejected (tactic 1). Saving Private Ryan, Cast Away — still crowd answers, rejected. Second-famous rule (tactic 3): not the iconic roles — Big, the 1988 comedy, or Turner & Hooch. Decades check (tactic 5): 1980s Hanks is thinner than 1990s Hanks. Too-clever check: Big is a real, famous-enough film — safely banked. Answer: Big. Expected tier: Deep Cut, 85 points.

Prompt two: “Name a breakfast food.” Trace: pancakes — Plankton, rejected. Waffles, eggs — Schooler, rejected. One level deeper (tactic 2): crepes, shakshuka — shakshuka feels exciting but risky; is it banked? It is a real dish, but the crowd for it might be too thin — too-clever trap territory. Life-mining (tactic 4): what did your family actually eat? Grits — real, regional, valid, and far from the crowd’s first three ideas. Answer: grits. Expected tier: Rare, 4 points.

Notice what happened in both: the first answer was sacrificed, the exciting-but-risky answer was checked against the bank, and the final pick was rare-but-safe. That three-step rhythm — reject, reach, verify — is the whole game at high level.

Strategy is half tactics, half avoiding self-sabotage. These four habits show up in low-depth dives over and over, and none of them have anything to do with knowledge.

• Answering fast. Speed earns zero points — only rarity does. A 5-second Plankton answer and an 85-second Deep Cut answer take the same clock time. Use the time.
• Joke answers. Funny is not a tier. Joke answers land outside the bank and score 0, and they waste a prompt you will never get back.
• Maximum obscurity. The rarest answer you can think of is usually too clever — outside the bank, worth 0. Rare-but-recognized beats obscure-and-alone every time.
• Repeating yesterday’s winners. Tiers drift as the crowd learns. An answer that hit Krillion last week is more crowded this week. Hunt fresh; never autopilot.