Strategy guide · 5-minute read

Krillion cheats that aren't cheats.

There are no exploits in Krillion — the score is a property of the crowd, and no shortcut changes what the crowd thinks. What you can change is your own search path. These are the habits that actually raise scores.

The four habits, in the order you should use them

1. Say the obvious answer, then bin it

Your first answer is the prototype — the category's best example, the word a hundred players type in unison. Naming it to yourself costs two seconds and clears it out of the way. It also tells you what the crowd is doing: wherever your first thought went, most of them went there too.

2. Bin the second one too

This is the habit almost nobody has. Once players know the obvious answer is worthless, they all skip it — and land on the same second thought. Psychologists call the convergence a focal point: people who cannot confer still converge on the choice that feels natural to everyone. The tier ladder has a name for that answer Too Clever, and it's the single most common place for a good player to lose points. The sequence "cow → goat → alpaca" is worth more than the sequence "cow → goat → stop".

3. Change the angle, not the effort

Rare answers come from a different slice of the category, not a deeper dig into the same slice. The reliable re-angles:

  • Geography: another continent, a dependency instead of a country, the smallest or least-obvious member.
  • Time: another century — a defunct airline, an extinct animal, a discontinued product.
  • Profession: the trade's insider example. Chefs and birders and railway enthusiasts all have answers the general crowd never reaches.
  • Size and extremity: the smallest, oldest, highest-altitude, least-populous member that still clearly fits.
  • Cross-category: a food that is also a colour, a place that is also a name. Prompts often hide intersections.

4. Stay inside the prompt

An answer that isn't accepted scores nothing, so obscurity only pays when the answer is unarguably correct. The failure mode is a spelling gamble or a category stretch — a city-region confusion, a tournament picked as a sport. Pick the rarest answer you are sure of, not the rarest you can imagine. In the official game a refused answer costs only the seconds it took; in our endless mode it costs a tank of air, which teaches the same lesson faster.

The rarest correct answer you're confident in beats the rarest answer you can imagine. Every time.

Reading the prompt like a setter

Find the boundary the prompt is hiding

Good prompts have a trap built into their wording. "Name a country that has territories or dependencies" is not "name a country" — the rarest answers are the countries most people forget hold territory: Denmark with Greenland and the Faroes, or New Zealand with its associated states. "Name an animal whose name starts with B — breeds count too" quietly widens the tail from animals to dog breeds, and the one-in-a-krillion answers live in the widening. Before you answer, ask what the prompt's second clause just made newly legal.

Time the commit

Treat the 25 seconds as two phases. From zero to fifteen: scan, discard, re-angle. From fifteen to twenty-five: commit to the best candidate you have and type it cleanly. Panic-typing at second 23 is how rare answers become typos, and a typo on a rare answer is the most expensive mistake in the game — the cheap answer is still on the table but the clock isn't.

What speed is worth

Nothing, directly. The tier, not the clock, decides your score, so there is no bonus for answering in five seconds and no penalty for taking twenty. The only speed decision that matters is leaving yourself enough time to type the answer cleanly and press dive before the clock — everything else is thinking time, and thinking is the entire skill.

Cheat-adjacent things that don't work

  • Answer lists from yesterday. Dives rotate prompts and answer banks shift; memorising yesterday's board trains trivia, not the skill.
  • Going maximally obscure every time. The esoteric-corner gamble misses the bank more than it lands, and in modes where a miss costs air or time, expected value beats ceiling-chasing.
  • Waiting for others to answer first. There's nothing to read — the crowd is simultaneous and invisible. The only signal you have is your own search path.

One honest advantage does exist: reps. The two-discard habit is a skill, and habits are built by volume. That's what the practice modes here are for — Unlimited for full-dive volume, Infinite for pressure reps where a miss costs air.

FAQ

Strategy questions

Are there real cheats in Krillion?

No. The score comes from how many other players would have given your answer — a crowd property where no code, pattern or exploit changes the outcome. What people call "cheats" are the strategy habits on this page: discarding prototypes, re-angling categories, committing cleanly.

What's the single fastest way to improve?

The two-discard habit. Name the obvious answer to yourself and bin it; bin the second thought too, because that's where the crowd converges. Then re-angle the category and commit to the rarest answer you're sure of. Volume reps — for example on Unlimited — build that habit faster than anything else.

Why does my clever answer keep scoring badly?

Because everyone else skipped the obvious answer too and met you there. The Too Clever tier exists precisely to catch the shared second thought. The fix is one step further out: the specialist's example, the regional variant, the forgotten member.

Should I ever go maximally obscure?

Only when you're confident the answer is unambiguously correct and fits the prompt exactly. Obscure-but-rejected scores zero, and in modes where mistakes cost air or time, a steady Schooler answer often beats a coin-flip Deep Cut across a full dive.

How much time should I spend per prompt?

All of it, deliberately. Speed earns nothing; the tier decides the score. Scan and discard for the first fifteen seconds, commit in the last ten — and leave enough time to type cleanly at the end.