The short version
Join with at least three players, keep private information private, take turns giving an answer or clue, discuss inconsistencies, and vote. Informed players try to prove they understand the shared prompt without revealing it. The Imposter tries to infer that prompt and blend in.
The two sides of an imposter round
Informed players know the central piece of information. Depending on the mode, that might be a question, a secret word, a ranking prompt, or a drawing subject. Their challenge is restraint: if they reveal everything, the Imposter can copy them; if they say almost nothing, the group cannot distinguish them from the Imposter.
The Imposter receives less information or a related prompt. That player is not expected to guess blindly. Listening, noticing the category, and constructing a plausible explanation are the role’s main skills. A fair game gives the Imposter enough context to make decisions while preserving an advantage for informed players.
A complete round, step by step
- Create or join a lobby. The host can share a lobby code or link. Use a name the group can recognize so discussion and voting are clear.
- Check the private information. Read your prompt or role without saying it aloud and avoid screen sharing the game window.
- Submit the required clue. Answer the question, add a drawing line, give a sentence, or rank the choices according to the current edition.
- Review the whole group. One unusual clue is evidence, not proof. Look for answers that fit the broad category but miss a condition shared by everyone else.
- Discuss before voting. Ask players to explain their reasoning without quoting the prompt. Give the suspected player a chance to respond.
- Vote independently. Avoid announcing a vote early just to create a pile-on. After the result, explain which clue changed your mind.
How specific should a clue be?
A useful clue is specific enough that another informed player can recognize the connection, but indirect enough that it does not hand the answer to the Imposter. Think of specificity as a dial. Early players usually keep it lower because the Imposter has heard little. Later players can turn it up when the group needs clearer evidence.
For a secret word such as “pancake,” “I associate it with breakfast” is broad but meaningful. “It is a flat batter food served in a stack with syrup” removes almost every uncertainty. On the other hand, saying “I like it” fits too many foods and provides little proof. The right middle changes with the group and with clues that have already been given.
Strategy by edition
| Edition | Evidence to watch | Best first habit |
|---|---|---|
| Question | Answers that are individually sensible but do not match the same question. | Ask for the reasoning behind one detail. |
| Word | Clues that fit the category but never become specific to the word. | Use an experience or comparison, not a definition. |
| Drawing | Lines that do not connect to the developing scene or cannot be explained. | Add one readable detail and leave room for doubt. |
| Ranking | A top three that ignores an unstated condition shared by other lists. | Remember why you ordered each choice. |
Hosting a fair discussion
The host’s job is not to lead the group toward a particular suspect. Set a short discussion window, invite quieter players to explain their clues, and stop repeated accusations from taking over. If someone misunderstood a prompt, separate that mistake from deliberate bluffing. Social deduction works best when unusual answers remain allowed.
For a first game, use familiar categories and a moderate number of rounds. Once everyone understands the clue balance, add niche categories or community packs. Public community content is reviewed before entering the catalog, but hosts should still choose material appropriate for their own group.
Common problems and simple fixes
The Imposter guesses the answer too quickly
Ask informed players to avoid definitions, exact synonyms, and several strong clues in one turn. The goal is to prove knowledge to the group, not to solve the prompt publicly.
Voting feels random
Require one reason before each accusation and compare at least two possible suspects. A player being quiet, new, or bad at drawing is not enough evidence by itself.
The same person controls every discussion
Go around once so each player can ask or answer a question. Then open the discussion. This keeps fast speakers from deciding the round before everyone participates.