Choose a target with a hypothesis
An Investigator checking a contested claim has a clear purpose; a Doctor protecting a likely kill target has a clear purpose. Random actions are harder to evaluate because success and failure do not answer a planned question.
Track intended target and resolved target
Transporter or control can change where an action lands. A clean will records whom you selected, while the day discussion reconstructs whether that was the player actually affected. This distinction prevents false contradictions.
Resolve control effects before result interpretation
If a player was jailed, blocked, transported or controlled, determine that chain first. Only then decide whether the missing or unexpected result is evidence of a fake claim. Night feedback is a system, not a set of independent events.
Use visits as evidence, not alignment proof
Lookout and Spy can show movement, but Town and evil roles both visit. The useful question is whether the visit is compatible with the claimed role and target, not whether visiting a dead player automatically means murder.
Reconstruct the night in a fixed order
For discussion, collect control and redirection claims, protection claims, results and deaths in a consistent checklist. That is a reading method, not the game engine’s universal action-priority order. When two effects conflict, identify the particular roles and their restrictions instead of treating this list as a timing rule.
Resolving a confusing night
Sheriff claims they checked Player 4, Transporter says 4 and 9 were swapped, and Lookout saw the Sheriff visit 9. The Sheriff result belongs to the transported destination, not necessarily Player 4. Resolve the transport first; only then compare the result with Player 9’s claim.
Keep selected targets and claims on separate lines
My action: Sheriff selected 4; result innocent.
Claimed redirect: Transporter says 4 ↔ 9.
Corroboration: 4 confirms; 9 has not answered.
Working interpretation: result may concern 9.Do not replace your original target with the inferred recipient. If the transport is later disproved, that preserved record prevents a second false accusation. When multiple effects overlap, leave the unresolved link explicit.
Common mistakes
- Choosing targets with no question in mind.
- Calling a result impossible before resolving Transporter or control.
- Treating a visit as alignment proof.
- Recording only the final theory instead of the raw target and feedback.
