Town of Salem gameplay with the Lookout role panel visible.
🏛️ Town Investigative

Lookout

Watch a target at night and map who visited them.

TownTown InvestigativeTown of Salem role guide
Quick answer: Watch one player at night to learn which players visit that target.
Alignment / classTown Investigative
Primary jobTurn visitor patterns into verifiable interaction evidence for Town

Visitors, the three-name limit and disguises

  • Visitor information can connect protectors, investigators, killers and support roles to the same target.
  • You identify at most three visitors in one night. If more people visit, the game indicates additional visitors without naming all of them.
  • A named visitor is evidence about movement, with a further caveat: Disguiser can make a Mafia visitor appear under another player’s identity.

Test a visitor identity before treating it as proof

Suppose your list names Player 8 at a house where Mafia acted. That is a useful lead, but Disguiser can substitute another identity in your report. Ask for the visit claim, compare Spy information and consider whether a disguise fits before demanding a one-for-one execution.

Watch a house with a useful question

Early game

Watch players who are likely to attract meaningful activity: public information roles, obvious kill targets or players at the center of protection claims. Watching an inactive random slot often produces no useful graph.

Mid game

Use repeated visits to test claims. If a Doctor, Bodyguard or Spy claims to have visited a player you watched, the visitor log can strengthen the timeline; an unexpected visitor can also create a high-value question for the next trial.

Endgame

In a reduced lobby, a single visitor can become highly informative because fewer roles remain capable of making that visit. Recalculate old logs after deaths rather than treating each night as isolated.

Record the watched house and every named visitor

A Lookout will should record the watched player and every visitor shown, in the order and wording you received them. Avoid silently dropping a visitor because it does not fit your current theory.

Test the identity and the claimed destination

Check whether claimed visitors acknowledge the visit and whether Transporter could have moved the target. A fake Lookout often invents names that happen to sound plausible but cannot explain the surrounding night feedback.

Important interactions

  • Doctor and Bodyguard visits can corroborate protection claims.
  • A killer visiting the watched target can be exposed, but the same visitor may have a legitimate support role, so role-list context still matters.
  • Transporter can make a player appear at a destination they did not intend, complicating direct interpretation of the visitor list.

Decision example: lookout target choice

You watch a publicly confirmed Investigator and see two visitors. One later claims Doctor; the other claims a non-visiting role. That second story now has a concrete mechanical problem. The Lookout does not need to know the visitor’s exact role to create a valuable trial question.

Common mistakes

  • Watching low-interest targets without a hypothesis.
  • Equating “visited the victim” with “killed the victim.”
  • Publishing a visitor list without preserving who was actually watched.