DI33 logo DI33Honest game reviews, no hype
Games

Phasmophobia Review and Why Its Ghost Hunts Need Trust

My group got 90 hours into Phasmophobia before anyone said the obvious thing out loud: we had stopped being frightened somewhere around hour twenty, and we were still playing four nights a week. That bothered me as a reviewer.

Four investigators with torches and a thermometer spread through the dark hallway of an old farmhouse during a ghost hunt

My group got 90 hours into Phasmophobia before anyone said the obvious thing out loud: we had stopped being frightened somewhere around hour twenty, and we were still playing four nights a week. That bothered me as a reviewer. A horror game whose horror has worn off should be finished, and this one had quietly turned into something else without telling us.

What it turned into is a four player information problem with a ghost attached. You are sent into a house with a shopping list of equipment and asked to work out which of 24 entity types is in there, using three pieces of evidence, while the thing you are identifying periodically tries to kill whoever is standing in the wrong corridor. The fear is the hook. The reason people stay is that the game punishes bad communication more precisely than almost anything else in co-op.

What the evidence system is really testing

Each ghost leaves a combination of evidence drawn from a pool that includes EMF level 5, a spirit box response, ultraviolet fingerprints, ghost writing, freezing temperatures, and D.O.T.S. projector silhouettes. On the standard tiers you need three of them to close the identification, and the journal narrows the candidates as you log each one.

In practice almost none of the difficulty lives in finding evidence. It lives in reporting it. Freezing temperatures require someone to read a thermometer and say a number out loud while two other people are shouting about a door that closed. A spirit box response means nothing unless the person holding it says which room they were in. I have watched a run collapse because somebody found fingerprints, said "got prints" over proximity chat from the far end of a farmhouse, and nobody heard it.

So the skill curve is not a reflex curve. It is a habit curve: say the evidence, say the room, say your sanity, in that order, every time. The groups that adopt a reporting format stop failing contracts almost immediately, and that transition happened for us somewhere around hour fifteen.

Sanity is the real timer

Sanity drops while you are in the dark and recovers only outside the house or under lit rooms, and when the average falls past a threshold the ghost starts hunting. That single mechanic is what gives the game its shape, because it converts every decision into a cost.

Want to be thorough in the basement? That is sanity. Want to wait out a hunt in a locker? Also sanity, because the lights are off. The result is a game where thoroughness and survival pull against each other continuously, which is a far better design than a straight timer would have been.

It also creates the only genuine team economy I have seen in a co-op horror game. Pills restore 40 percent and there are rarely enough for four people, so somebody has to decide who gets one, and the honest answer is usually the person who still has a job to do rather than the person who is most scared. My group argues about this more than anything else, which is a compliment.

Difficulty tiers, and where the payout sits

The tiers change three things at once: how much evidence you get, where your sanity starts, and what happens to your money if you die. Approximate figures from the current build, since the developers adjust these between patches.

TierEvidence shownStarting sanityInsurance on deathWho it suits
Amateur3 of 3100 percentMost of your kit returnedFirst ten hours, learning the maps
Intermediate3 of 3100 percentPartialLearning to read ghost behaviour
Professional3 of 3100 percentLittle to noneThe tier most groups settle on
Nightmare2 of 375 percentNonePlayers who have memorised entity traits
Insanity1 of 350 percentNoneSmall, very practised groups

Professional is where the game is at its best, and I say that having spent a long stretch convinced it would be Nightmare. Three evidence keeps the identification honest while removing the grace period that makes the lower tiers forgiving. You still lose contracts, and you lose them for reasons you can name afterwards.

The counter take on playing it on Nightmare

The standard advice to a group that has stopped being scared is to move up to Nightmare or Insanity, because fewer evidence slots restore the tension. I followed that advice for about fifteen hours and I think it is the wrong fix.

Reducing evidence does not make the game harder in an interesting way. It makes it a memorisation test. With two evidence pieces available you identify by behaviour instead: how fast the entity moves during a hunt, whether it can open doors, whether it targets the lowest sanity player. Those tells are real and learning them is satisfying, but once learned they turn the deduction into recall. Our Nightmare runs got faster and quieter, and quieter is the opposite of what the game is for.

What worked better was custom difficulty. Keep three evidence, then push the settings that affect pressure rather than information: raise hunt duration, lower the sanity floor at which hunts begin, shorten the setup time. You keep the deduction loop that makes four people talk to each other and you restore the fear that was only ever a product of time pressure. That is the configuration my group has run since roughly hour sixty.

I will hedge on one point. Groups that enjoy mastery for its own sake may genuinely prefer Insanity, and two people I play with do. The claim is only that the tension people are chasing comes from the clock, not from removing a line in the journal.

Where it falls short

The large maps are a different and worse game. Sunny Meadows and the prison turn an eight minute identification into a twenty five minute walk, and the sanity economy does not scale to the distances involved. We treat the large maps as a novelty rather than content.

Progression after the Ascension rework is also thin. Levelling unlocks equipment tiers and then prestige resets you, which is a loop designed to extend play rather than reward it. I would rather have had more small maps. And the game remains visually plain in a way that a 2024 release would not get away with, though at the price it usually sells for that is easy to forgive.

One genuine flaw worth flagging: solo play is not a reduced version of this game, it is a different and much weaker one. Every mechanic I have praised is a communication mechanic. Alone, the evidence gathering is a chore list. If you have no group, this is not the purchase for you, and the same is true of Lethal Company for the same structural reason.

Who should buy it

Buy it if you have three friends who will talk, and if you are willing to spend the first ten hours on Amateur learning two small maps properly rather than touring all of them. Tanglewood Street and Grafton Farmhouse are enough for twenty hours.

Skip it if you want a horror game with a story, or if you intend to play alone. If what appeals is specifically the pressure of having to say the right thing to the right person under a clock, and you want a cheaper and shorter version of that feeling first, Keep Talking and Nobody Explodes tests the same muscle in twenty minute sittings. More of our co-op coverage sits in the games section.

FAQ

Is Phasmophobia worth playing in 2026?

Yes, with a group. The deduction loop has aged well and the patch cadence has kept the entity roster and equipment balance moving. Alone, I would not recommend it at any price.

How many players is Phasmophobia best with?

Four is the cap and three is the sweet spot. With four, one person usually ends up idle at the van. With two, the sanity economy gets punishing on anything above a small map.

What difficulty should a new group start on?

Amateur for the first few contracts, then Professional once everyone can name the evidence types without opening the journal. Skipping to Nightmare early teaches behaviour tells before you understand what they are tells of.

Does it still frighten you after 50 hours?

Not in the jump scare sense. What survives is the dread of walking back into a house at low sanity with one job left, which is a function of the sanity timer rather than the ghost models. Custom difficulty keeps that alive better than raising the tier does.

MR
Marcus Reyes

Marcus spent years grinding ranked ladders before deciding he liked explaining games more than climbing them.

More posts by Marcus

More in Games

01doorganico 084goad 3hot8 4465456 after5internprogram ahandfulofblessings ahmadpc ahomcybersecurity allodialresources annodex antiquesandperiodcollectibles aotc asianworldkashmir beedlectrics bloggertemplatefree bluefeatheragency bodyweightsolution campaign08blog carfixadvisor catstreetcollective cbstvsports cnfanspurchasing committeeof500yearsofdignityandresistance covertopsdjs cs-kaiyuangame cuestacomunicaciontotal cxoutlet dancebetgame deskgearhub destinyvillastay detectmagazine diamondsteelbuildings donmo dyanadropship emarsysfinance emimattl escapechan eskitarih evautoplay everythingweddings-orlando farmzone feedviewerapp foxstudiosscratch gamevui123 gleasongrade greetingsday hacklabasu haskell iw-kaiyun jagoanproperti jilidskripsi joltbusinessfinance juegodepokergratis juegospucca kaeru-festival latestmehndidesigns lestanzediottaviano litemagazines localcommunityhero lustrsazan mariashebets menswheels newwebstechs1 nikkit ootheday oxigenvpn plurfitness prefetial probudgetcarpetcleaning psychicmrsmartin pyaten qvantor rejoignezlogis rexzbot riselayouts ruchin rwxqsgd5 sakaltime shibiaowang shop-bot skrtips spotlesshaven strcoding tetapatrium topselller uneplanteartificielle vipertechcr webtech1223411 wo-kygaming yamaha-mekar yeknava yoybuysheets zenunwa yamaha-mekar jilidskripsi ankertoto ankertoto ankertoto