Night 2 keeps the patient-case foundation but punishes Night 1 autopilot. The official badge confirms the second completion milestone. A current exact-game walkthrough reports an observed twelve-patient target, changed instructions, and a faceless-patient warning. Read the new tutorial text before touching the first station, because a familiar room does not guarantee an identical branch.
Reset your assumptions at the start
Pause long enough to read every changed instruction. Check whether the patient counter, task wording, or available interactions differ from Night 1. Keep one active case and let the objective choose the next action.
Do not carry a memorized ten-patient stopping point into this shift. The current community route reports twelve. More importantly, do not call the next patient until the active case visibly completes.
Repeat the full protocol, not just the fast parts
Use the form and visible patient context, then the requested diagnostic checks. Heart-rate and temperature information can still be useful, and night vision can add another perspective. Later-night coverage warns that apparently clean devices may not settle every case, so compare behavior and identity context instead of letting one green-looking result end the decision.
Follow treatment one named step at a time. If the task changes or the patient reveals something impossible, stop the ordinary branch and follow the current response. The purpose of the loop is not to force every patient through the same supplies; it is to know why the next action is safe.
Handle the faceless warning
Current Night 2 coverage describes a faceless patient or faceless visual as a significant warning and cautions against prolonged staring. Capture the useful observation, look away, and cross-check the form, task, or another signal. Do not turn the warning into a complete universal rule: the exact category and outcome were not confirmed for this checkpoint.
If a visual event creates sanity pressure, slow the queue. Resolve the active patient and use a safe between-case pause. The Sanity guide explains how to protect decision quality without claiming an exact drain value.
Work through all twelve observed cases
Track progress through the game’s counter or objective, not memory alone. At the start of each case, say what the task requires. At the end, confirm treatment or anomaly feedback, the form/computer decision, cleanup, and a changed objective.
As pressure rises, mistakes tend to come from incomplete branches rather than a lack of speed. Keep the patient route consistent and avoid several people clicking the same station. In a squad, assign a task reader and a primary operator.
Recover from Night 2 failures
If a case stalls, return to patient placement, the requested tool, the current treatment action, and the final decision. If the twelfth case is done but the night continues, follow the completion or leave prompt shown in your session.
If clean diagnostics conflict with disturbing behavior, do not invent a threshold. Recheck independent evidence and follow the game’s current prompt. The Detection guide separates high-confidence measurement warnings from community-reported later-night exceptions.
Completion and endings
The official badge proves Night 2 can be completed. A community walkthrough observes a completion sequence after the target, but it does not establish the total number of endings or all branch conditions. This page therefore guides you to the completion state without naming one scene as the only ending.
Use the visible completion feedback and badge award as the finish signal. If you want to compare a different outcome, change one documented decision at a time rather than following an unsupported “all endings” list.
Night 2 FAQ
How many patients are in Night 2?
A current exact-game walkthrough reports twelve. Treat the number as an observed UPD-era value, not a permanent official specification.
What does the faceless patient mean?
It is a significant visual warning in current community coverage. Its full entity classification, spawn rule, and exact consequences still need in-game testing.
Can I trust normal diagnostic readings?
Use them as evidence, not absolute proof. Later-night community coverage specifically warns that behavior and identity context can matter when devices appear normal.
How many endings are there?
The total and conditions are not confirmed by the official badge data. Only the completion milestone is certain.
Where should I go next?
Continue to Night 3, but read its update context and avoid importing the twelve-patient target.