Doppelganger GuideintermediateUpdated: 7/23/2026

Shift At Midnight Red Flags & Anomaly Full Checklist

The complete Shift At Midnight red flags checklist. Spot doppelgangers faster with this exhaustive behaviour-anomaly list, weighted by reliability and frequency.

Red flags and behaviour anomalies are the second layer of the doppelganger detection workflow, sitting on top of the ID cross-check. This article compiles every documented red flag into one weighted checklist, ordered by reliability, with practical notes on when each flag is most useful. Print it, screenshot it, or memorise it — but do not try to play late-game shifts without it.

The list is long because doppelgangers are good at any single trick. The value of a checklist is that it forces you to look for multiple independent anomalies before deciding. A customer with one red flag might be human; a customer with three is almost certainly doppelganger. Per the official Steam News hub, the Q4 2026 Endless Mode build rebalances these exact values against the 1.0 launch baseline; treat the figures above as the launch reference set and re-check after each patch.

How to Use This Checklist

Core Strategy

  • Primary action: Lock in the core strategy before any customer interaction, since the timing window is tighter than it appears.

  • Secondary action: Track monster spawn timers, weapon cooldowns, and teammate callouts before committing to the core strategy.

  • Failure recovery: If the primary action fails, fall back to the defensive stance instead of restarting the sequence.

  • Checklist: Confirm monster positions, weapon readiness, and customer flow before triggering the core strategy step.

The checklist is organised by reliability, not by customer type. Each red flag has a reliability rating based on community observation during launch week. Treat high-reliability flags as decisive and low-reliability flags as supporting evidence only. Co-op coordination via Proximity Chat lets the lookout clerk call a spawn 8 seconds before monsters arrive, while solo players have to rely on the audio cue timing, which triggers roughly 5 seconds later.

For additional official context and live updates, see the official Shift At Midnight Discord.

ReliabilityWhat It MeansAction Threshold
Very HighAlmost always a doppelganger1 flag is enough
HighStrongly indicative2 flags is enough
MediumUseful supporting evidence3+ flags is enough
LowSuggestive onlyCombine with others

The action threshold column tells you how many flags of that reliability level you need before defaulting to execute. Use it as a quick reference during shifts. The underlying mechanic is tied to the procedural shift generator, so no two runs feel exactly the same even when following the same strategy.

Very High Reliability Flags

Key Mechanics

  • Primary action: Lock in the key mechanics before any customer interaction, since the timing window is tighter than it appears.

  • Secondary action: Track monster spawn timers, weapon cooldowns, and teammate callouts before committing to the key mechanics.

  • Failure recovery: If the primary action fails, fall back to the defensive stance instead of restarting the sequence.

  • Checklist: Confirm monster positions, weapon readiness, and customer flow before triggering the key mechanics step.

These flags are decisive on their own. If you spot one, the customer is almost certainly a doppelganger. Beginners tend to overlook this detail, but mastering it early saves significant frustration once monster hunts begin in earnest during shifts five and beyond. This is one of the highest-impact micro-decisions you make each shift, because it influences both your survival odds and your final shift score for the achievement tracker.

Red FlagWhy It WorksWhen to Look
ID photo is from a different decadeDoppelgangers reuse older ID photosCompare jawline carefully
Name spelling does not match databaseDoppelgangers improvise namesRead letter by letter
Customer contradicts earlier answerThe classic two-question test catches thisAsk any factual question twice
Customer refuses to show ID at allReal humans hand over ID willinglyPolite refusal is itself a tell
Customer stares at you without blinkingNon-human microexpressionWatch for 5+ seconds

The "stares without blinking" tell is the most underrated. Real humans blink, look away, and shift their gaze during a 30-second interaction. Doppelgangers, especially the higher-tier variants, hold eye contact with an intensity that feels off when you actually pay attention. It is one of three mechanics that the Discord #mentor channel asks every new clerk to confirm before joining a co-op run, alongside board priority and sidearm ammo discipline.

High Reliability Flags

Solo Play Tips

  • Primary action: Lock in the solo play tips before any customer interaction, since the timing window is tighter than it appears.

  • Secondary action: Track monster spawn timers, weapon cooldowns, and teammate callouts before committing to the solo play tips.

  • Failure recovery: If the primary action fails, fall back to the defensive stance instead of restarting the sequence.

  • Checklist: Confirm monster positions, weapon readiness, and customer flow before triggering the solo play tips step.

These flags are not decisive on their own but become decisive in combination with any other flag. Endless Mode, when it lands in Q4 2026, is expected to amplify this behavior, so learning it well now pays dividends long-term as the meta evolves. The 0.9.4 network-issues-patch beta adds a brief 0.5-second slow-mo when the Brute breaks a board, which the Bun Muen team flagged for inclusion in the Q4 2026 release alongside Endless Mode.

Red FlagWhy It WorksWhen to Look
Customer mirrors your posture exactlyMimic behaviourCompare stance over 10 seconds
Customer over-performs politenessReal humans get boredWatch for excessive smiles
Customer hesitates on addressImprovising a fake addressAsk twice if uncertain
Customer looks at the door repeatedlyWants to leave before you finishCount glances in 30 seconds
Customer's eyes do not track your handLack of focus suggests non-humanMove your hand near their face
Customer's clothing is inconsistent1990s theme means period clothingLook for modern accessories

The "eyes do not track your hand" test is borrowed from real-world detection training. Humans instinctively follow movement near their face; doppelgangers occasionally fail this simple reflex.

Medium Reliability Flags

Co-op Synergy

  • Primary action: Lock in the co-op synergy before any customer interaction, since the timing window is tighter than it appears.

  • Secondary action: Track monster spawn timers, weapon cooldowns, and teammate callouts before committing to the co-op synergy.

  • Failure recovery: If the primary action fails, fall back to the defensive stance instead of restarting the sequence.

  • Checklist: Confirm monster positions, weapon readiness, and customer flow before triggering the co-op synergy step.

These flags are useful as supporting evidence but rarely decisive alone. Combine with two or three other flags before defaulting to execute. According to recent Steam Community discussions, this is the most frequently asked technical question about the mechanic, and the answer has been refined as the game has patched.

Red FlagWhy It WorksWhen to Look
Customer's voice is monotoneDoppelgangers struggle with intonationListen for natural inflection
Customer takes too long to answerProcessing a script rather than recallingTime the response
Customer's hands do not moveNervous humans fidgetWatch for hand motion
Customer's smile is delayedReaction is performed, not naturalWatch the timing
Customer has no smell of coffee/foodReal customers come from outsideTrust your nose if close enough
Customer's shadow does not matchLighting inconsistencyOnly on certain light setups
Customer asks the same question twiceTrying to gather data on youNote repetition

The "shadow does not match" flag is rare and situational, but it is the strongest non-ID evidence in the game. If you ever spot it, treat it as a very high reliability flag regardless of what else is happening. Updated numbers are expected in the Q4 2026 Endless Mode update; the official Steam News hub and the #patch-notes channel on the official Discord carry the live tuning values within hours of each push.

Low Reliability Flags

Common Pitfalls

  • Primary action: Lock in the common pitfalls before any customer interaction, since the timing window is tighter than it appears.

  • Secondary action: Track monster spawn timers, weapon cooldowns, and teammate callouts before committing to the common pitfalls.

  • Failure recovery: If the primary action fails, fall back to the defensive stance instead of restarting the sequence.

  • Checklist: Confirm monster positions, weapon readiness, and customer flow before triggering the common pitfalls step.

These flags are suggestive but easily fooled. Use them only when you have already accumulated two or three stronger flags. Treating shifts 1-3 as a skills lab lets you fail safely; by shift 7 the same approach carries real consequences and reset wipes that progress back to the save point.

Red FlagWhy It WorksWhen to Look
Customer looks nervousCould be a real humanConfirm with other tells
Customer looks too confidentCould be a confident humanConfirm with other tells
Customer fidgets a lotCould be anxietyConfirm with other tells
Customer avoids eye contactCould be shy or culture-specificConfirm with other tells
Customer asks about the weatherSmall talk is normalNo action needed

Low-reliability flags are listed here so you do not over-weight them. A nervous real human looks similar to a nervous doppelganger in many cases. The tell matters only when combined with stronger evidence. Story Mode spawns one Brute event every 4 shifts; the Endless Mode preview compresses that to every 2 shifts from night 8 of an endless run, per the developer notes on Steam.

The Three-Strike Decision Rule

Pro Tips

  • Primary action: Lock in the pro tips before any customer interaction, since the timing window is tighter than it appears.

  • Secondary action: Track monster spawn timers, weapon cooldowns, and teammate callouts before committing to the pro tips.

  • Failure recovery: If the primary action fails, fall back to the defensive stance instead of restarting the sequence.

  • Checklist: Confirm monster positions, weapon readiness, and customer flow before triggering the pro tips step.

The optimal decision rule is the three-strike rule. Count independent red flags, weight them by reliability, and act when you cross the threshold.

Strike Count (Weighted)Recommended Action
0 strikesRelease
1-2 strikes (low only)Ask twice, then release
3 strikes (any combination)Execute
1+ strike (very high)Execute immediately

The weighted count lets you treat one very-high flag as equivalent to three medium flags. This is why the eye-contact and ID-mismatch tests are so powerful — they alone justify execution.

Advanced Techniques

  • Primary action: Lock in the advanced techniques before any customer interaction, since the timing window is tighter than it appears.

  • Secondary action: Track monster spawn timers, weapon cooldowns, and teammate callouts before committing to the advanced techniques.

  • Failure recovery: If the primary action fails, fall back to the defensive stance instead of restarting the sequence.

  • Checklist: Confirm monster positions, weapon readiness, and customer flow before triggering the advanced techniques step.

Red flag frequency shifts across the 13-shift campaign. Early shifts are forgiving; late shifts are brutal. Plan your thresholds accordingly. Solo players often find this the hardest part of the workflow, while co-op teams can distribute the responsibility across multiple roles for a much smoother experience.

Shift RangeAverage Red Flags per DoppelgangerThreshold
Shifts 1-32-3 flags per doppelganger3 flags = execute
Shifts 4-71-2 flags per doppelganger2 flags = execute
Shifts 8-101 flag per doppelganger2 flags = execute
Shifts 11-130-1 flags per doppelgangerAny flag + ID mismatch = execute

Late-game doppelgangers are designed to mimic humans more closely, which is why the threshold drops. By shift thirteen, even a single red flag combined with an ID anomaly is decisive.

Frequently Asked Questions

When To Use This

  • Primary action: Lock in the when to use this before any customer interaction, since the timing window is tighter than it appears.

  • Secondary action: Track monster spawn timers, weapon cooldowns, and teammate callouts before committing to the when to use this.

  • Failure recovery: If the primary action fails, fall back to the defensive stance instead of restarting the sequence.

  • Checklist: Confirm monster positions, weapon readiness, and customer flow before triggering the when to use this step.

How many red flags is "enough"?

The weighted count is the safest metric. One very-high flag is enough; three medium flags is enough; five low flags is enough. Anything in between requires the two-question test. Community wiki consensus puts this strategy in the top three priorities for clearing any shift cleanly, and it pairs especially well with the recommended weapon loadout.

Do red flags ever appear on real humans?

Low-reliability flags are common on real humans. The trick is that real humans rarely accumulate more than one flag, while doppelgangers accumulate two or three. The most recent community polls on Discord show that this exact pattern catches new players out roughly once every two shifts until they internalize the timing.

Are the flags random or scripted?

The flags are procedurally generated per customer based on a seed. Two runs of the same shift will have different flag combinations, but the underlying types are drawn from a fixed pool.

Do flags change in Endless Mode?

Endless Mode red flags are subtler and require higher thresholds. The base rules still apply, but expect to need four or five flags for confidence late in an Endless run. Solo players who clear this consistently report a noticeable jump in confidence going into shift five, when the monster spawn frequency first accelerates.