Zero-Evidence Investigation Walkthrough
Zero-Evidence broken down step by step — the elimination order one experienced investigator uses when there's no evidence to lean on.
Stage 1 — Read the first hunt
During the very first hunt, you're passively collecting several signals at once: speed, blink pattern, throw behavior, and line-of-sight acceleration. A single hunt can eliminate close to half the roster if you know what to watch for.
1. Base speed
If the ghost is moving at the standard 1.7 m/s the whole time (not visibly slow, not visibly fast), you can immediately rule out every ghost with a non-standard base or conditional speed:
- — inverts entirely: ~3.0 m/s far away, slowing to ~0.4 m/s up close. Constant "normal" speed doesn't match either extreme.
- — 1.5 m/s at high sanity, up to 2.25 m/s (and up to ~3.71 m/s combined with LOS acceleration) as sanity drops toward 0%. At anything other than very specific sanity, its speed reads as "off" from standard.
- — jumps to 2.5 m/s only near active electronics (including gear in your own hands, like a flashlight). If it doesn't speed up while you're holding on active flashlight nearby, that's a direct test against .
- — a stark binary: 1.0 m/s with no line of sight/detection, 3.0 m/s once it detects you or your electronics. Neither of those is "normal," so a steady 1.7 m/s rules it out immediately.
2. The test (distance + LOS)
only speeds up to 2.5 m/s when it has line of sight on a player more than 3 meters away and the breaker is on. If you were >3m away, in its sight, breaker on, and it didn't accelerate — not a .
3. Salt
is the only ghost in the game that never disturbs salt under any circumstance — it walks through without leaving a footprint or getting slowed. If the ghost stepped in salt and disturbed it at least once, that's a clean elimination of . is a different kind of salt tell: its behavior flips with its state — it won't disturb salt while Enraged (right after salt/incense/crucifix is used on it) but disturbs it normally once Weakened (right after a hunt ends). Salt disturbance that's inconsistent across multiple triggers, rather than a hard never/always, points at instead of a straight read.
4. The standard line-of-sight acceleration check
This is the detail worth knowing cold: most ghosts gradually accelerate while holding continuous line of sight on a player, reaching roughly 1.65× base speed (~2.805 m/s) after about 13 seconds. This is a universal mechanic layered under everything else — but there are exactly four confirmed exceptions that never get this acceleration: , , (which instead slows down as it closes in), and (whose hunt speed is instead set entirely by how many unique objects have been interacted with since its last hunt — 3.0 m/s if nothing's been touched, grinding down to a floor of ~0.4 m/s after 26-30 interactions). is a fifth wrinkle worth watching separately: it does get the acceleration, just abnormally fast, off a lower starting speed — reaching full sprint in under 9 seconds of sustained line of sight instead of the usual ~13. If you clocked a standard ~13-second ramp-up, you've ruled out all four non-accelerators plus 's fast-ramp variant at once — is usually already excluded by its own distinctive speed signature above, so the useful new information here is against , , , and .
4b. vs. vs.
If no LOS speed-up showed up, , , and are all still live and need separating from each other: watch how speed trends across the investigation rather than within a single hunt. ages with the contract — it gets progressively slower the longer the investigation runs, so a hunt late in the round should be noticeably sluggish next to one from early on, and that slowdown never resets. , instead, tracks room temperature — faster in a cold room, closer to normal in a warm one — regardless of how much time has passed or how many hunts have happened. resets to a fast 3.0 m/s at the start of every hunt and then grinds down purely off how many unique objects get interacted with before the next hunt — so a hunt that comes in fast again shortly after the last one (with little object interaction in between) points to , while a slowdown that's tied to elapsed contract time points to and one tied to room temperature points to .
5. Blink pattern
Ghosts flicker in and out of visibility during hunts at a fairly consistent rate. Two ghosts break from that baseline in opposite directions:
- blinks slower, with longer gaps of invisibility.
- blinks faster and more frequently.
6. Model
is the only ghost that swaps its model mid-hunt, on a fixed schedule of specific blink counts. If the model stayed the same the whole hunt (and the hunt ran long enough to plausibly hit those trigger points), that's an elimination of . Separately, is always a female ghost model. is the only other ghost in the game locked to a female model — so this is a free, passive check for both: if the model reads as male, and are both ruled out immediately.
7. Throw force
throws items far more often (roughly every 0.5 seconds when objects are in range) and with noticeably more force than any other ghost. Ordinary, infrequent object interaction rules it out.
Stage 2 — Detection range
Still during that same first hunt: three ghosts have hearing or detection ranges wildly different from normal, worth checking before you move on.
8.
's hunt-audio range is reduced — footsteps and vocals cut off around 12m instead of the normal ~20m. If you can hear it clearly from further away than that during the hunt, it's not a .
9.
has a drastically reduced detection range during a hunt (roughly 2.5m for electronics vs. the normal 7.5m/9m). Walk away, then turn on an electronic item (like a flashlight) from well outside that range while it's hunting. If it reacts and comes straight back, that's inconsistent with 's short detection radius — rule it out.
10.
is completely blind during hunts and tracks purely by sound, but with detection range far beyond normal — roughly 10m crouched, 15m walking, 30m sprinting, even through walls and doors. Staying crouched and silent lets you walk right past it in plain sight during a hunt; if it beelines for you anyway with no noise made, it's not a .
Stage 3 — Active room tests
With the field narrowed, head to the ghost's favorite room and run a couple of setups at once.
11. Crucifix + firelight test ( vs. )
Place a crucifix and a lit firelight in the same room. is built around extinguishing flames as part of its hunt-trigger logic — if a nearby lit flame is within range when it would otherwise start a hunt, it prioritizes blowing that flame out instead of triggering the hunt. So: if the crucifix burns (meaning a hunt was attempted and blocked) while the firelight is still lit, that's inconsistent with 's known behavior — rule it out. While you're waiting in the room, if the ghost hunts while you're physically inside its favorite room, that rules out — it's specifically restricted from hunting, doing events, or most EMF-tier interactions while a player shares its room. Caveat: watch room boundaries carefully on small maps — a can step just outside its room to interact and create a false positive, so don't rule it out from a borderline case.
12. Orb check
While in the room, check for ghost orbs. 's ability spawns an orb in its favorite room as an unofficial "extra" piece of evidence — one that shows up even outside its official evidence set, and even on zero/low-evidence difficulties, since it isn't officially tied to the evidence system. Seeing an orb here is a solid tell.
Stage 4 — Smudge test ( vs. )
One more active test before falling back on passive reads: if you've still got a smudge stick, its hunt-cooldown side effect narrows two of the trickiest ghosts in one shot.
13. Smudge test ( vs. )
Smudge the ghost and time the next hunt:
- needs only ~60 seconds after a smudge before it can hunt again (vs. the normal 90s).
- is the opposite extreme — it needs a full 180 seconds (3 minutes) before it can hunt again.
- A hunt landing well outside either window rules out both.
Stage 5 — Varying speed
These ghosts don't have one fixed hunt speed — theirs shifts predictably, either on a timer or off an external trigger. Catching the pattern across more than one hunt (or one hunt with a state change) is the tell.
14.
randomly roll one of two fixed hunt speeds each time (slower ~1.5 m/s or faster ~1.9 m/s) — the point is that repeated hunts should show visible speed variance between them. If every hunt has come in at the same consistent speed across multiple hunts, that's a mark against .
15.
alternates between two states roughly every 2 minutes: Aggressive (~1.96 m/s, hunts at ≤65% sanity, hunts run 20% shorter) and Calm (~1.45 m/s, hunts only at ≤10% sanity). It starts every contract Calm, first switching about a minute after the exit door is opened, and can even flip state mid-hunt — a hunt that visibly speeds up or slows down partway through is a strong tell.
16. , again
's three-state cycle also shows up as a raw speed change, not just the salt behavior from earlier: Normal (1.7 m/s) → Enraged (1.96 m/s, triggered by salt/incense/crucifix) → Weakened (1.36 m/s, right after a hunt ends) → back to Normal. A ghost that visibly speeds up right after you use salt/incense/crucifix on it, then visibly slows down right after a hunt ends, is about as confirmed as gets without evidence.
Stage 6 — The stragglers
The rest of the roster comes down to behavior patterns rather than a single clean test — worth checking last since they lean on probability or absence rather than a hard rule.
17.
must fully open or fully close any door it interacts with — leaving a door at any half-open state is impossible for it. If the ghost's room has no doors to observe, this test simply can't run, and stays on the table until you can test it another way (e.g. tracking with motion sensors whether it gets trapped in-room for ~90 seconds after a smudge, another documented trait).
18.
By this stage, you're watching whether the ghost seems to be actively roaming toward you specifically, the way a stalks its one chosen target. If you're unsure, listen for the distinct wail on a parabolic mic or sound recorder — it has roughly a 1-in-3 chance of producing it on a given response. After several non-scream responses in a row, the odds swing hard against , though — being probability-based — this isn't an absolute proof the way a salt-and- test is.
The Hardest to Pin Down: , , , , ,
If you've run this whole flow and you're still stuck, there's a good chance you've landed on one of the ghosts the wider Phasmophobia community broadly agrees are the hardest to pin down with zero evidence:
- — its only real tell (preferring darkness, killing lights right after they're turned on, never turning lights on) is something any ghost can coincidentally do; you need to see the pattern repeat before it means anything.
- — without incense/smudges left to test its short hunt-cooldown, and without a Ouija board or crucifix interaction to watch, it has very little that separates it from a "generic" fast, aggressive ghost.
- — its defining trait (never changing its favorite room) is a non-event you can only confirm by absence over time; it gives you nothing active to test for.
- — its door test needs an actual door in its favorite room to run at all; without one, you're left relying on the weaker, indirect smudge-trap sign instead of a clean test.
- — its only tell is a ~180s re-hunt after a smudge; burn your last smudge stick early or on the wrong ghost, and there's nothing left that separates it from a default one.
- — ruled out only by catching it hunting while you share its room, which means being in the right place at the right time, and small-map room boundaries can produce a false positive either way.
If every test in this flow has come back neutral, whatever's left standing among these six really is close to a coin flip. At that point: pick one, and treat it as an educated guess rather than a certainty.