Why do jump scares stop working so quickly?

They are a reflex to a sudden loud signal. Once you know the timing, the surprise is gone and you start watching for the edit instead of inhabiting the space. The reaction does not survive a second playthrough.

Which games on this site use sound as a threat model?

Hunt: Showdown 1896 is the clearest example, where crows, glass and distant gunfire reveal enemy movement. Dead by Daylight inverts the idea with a terror radius that turns volume into a distance measurement.

Is randomness the same as unpredictability?

No. Pure randomness reads as noise. Effective horror uses rules you partly understand but a state you cannot observe, such as a hunt trigger whose conditions you know but whose timing you do not.

Does knowing the mechanics ruin a horror game?

It ruins games built on a single reveal. Systems-based horror survives explanation: knowing how Phasmophobia decides to hunt does not tell you when this particular hunt will start.

What is the practical way to keep a horror game tense?

Use headphones at moderate volume, avoid spawn-rule wikis on a first run, give the run a loss you care about, and stop when you notice you have switched from hesitating to optimising.

Why do you separate horror from survival in your reviews?

Because the question is different. Survival asks whether you planned well; horror asks whether you can act while uncertain. Several games sit in both, but the review has to say which pressure dominates.

Horror that lasts without cheap jump scares

There are two kinds of horror games: the ones that startle you once and get forgotten, and the ones that make you listen to an empty corridor twenty minutes after closing. The difference is not budget; it is design choices.

Choice one: sound before picture. The best horror games put information in audio: footsteps almost heard, breathing that stops. Eyes close, ears do not. Developers who understand this build darkness that stays scary with no monster on screen.

Choice two: give the player power, then make it heavy. A working flashlight is a tool; a visible battery is fear. This formula is old but it works because it moves the decision to the player: seeing costs.

Choice three: an enemy that learns. A monster that always spawns in the same place becomes moving decoration. AI that adjusts patrols to your habits makes the same corridor feel different in hour two.

Choice four: decisions under pressure. Good horror forces a pick: carry the medkit or the batteries, save at the door or push one more room. Because every option is partly right, no decision feels safe.

Jump scares have their place, as seasoning. But the night you remember years later is almost always the night you decided, yourself, to walk into that room.

Fear that survives the second playthrough

A jump scare works exactly once. It is a reflex to a sudden loud signal, and once you know the timing it becomes an appointment. Games that stay frightening across dozens of hours build something slower: the feeling that you are missing information that matters, and that the cost of being wrong is cumulative.

Five mechanisms that do the work

Asymmetric information. Phasmophobia gives you instruments that narrow down a ghost type but never confirm it instantly. You are always acting on a probability, and the game never tells you the odds. That uncertainty is the tension, not the apparition at the end.

Sound as a threat model. Hunt: Showdown 1896 is built around this: crows, broken glass, a distant concertina, a horse neighing. Players routinely win fights because they heard a team decide to move. Dead by Daylight does the reverse with a terror radius that grows as the killer closes, so volume becomes a distance measurement you cannot ignore.

Consequences that accumulate. Losing a run in Lethal Company costs the quota, the equipment and the crew's patience. Nothing resets to a comfortable baseline, so each small risk is priced against everything you have already committed.

Unpredictability with rules. Random noise is not scary; it is annoying. What works is a system whose rules you partly understand but whose current state you cannot see. Phasmophobia's hunts, Dead by Daylight's perk builds and S.T.A.L.K.E.R. 2's anomalies all sit in that band.

Deprivation of a sense of safety. Subnautica removes the horizon and replaces it with depth. The Forest removes the assumption that the empty forest is empty. Neither needs a sudden chord because the environment itself never offers a neutral moment.

Why cheap scares are a design shortcut

A loud sting is cheap to produce, tests well in a short clip, and is extremely shareable. It is also the first thing a player adapts to. Worse, it teaches the audience to watch the screen for the edit rather than to inhabit the space. Once a game has trained you to expect a sting every ninety seconds, the intervening minutes become downtime instead of dread.

This is also why horror that relies on a single mechanic tends to collapse on a second playthrough or after a popular video explains it. Systems-based horror survives explanation: knowing how a hunt starts in Phasmophobia does not tell you when this one will.

How to keep the tension for yourself

Play with headphones at a moderate volume. Most of the information design in this genre is auditory, and loud volume flattens the dynamic range that carries it.
Do not look up the spawn rules on a first run. Knowing the timetable converts probability into a schedule.
Set a loss you actually care about: a no-restart run, a quota target, or a shared save with friends.
Stop when you notice you are optimising rather than hesitating. Fatigue kills dread faster than any walkthrough.

Where KitaBertahan draws the line

We do not treat horror and survival as the same category. A survival game asks whether you planned well; a horror game asks whether you can act while uncertain. The Forest, Green Hell and Valheim sit mostly on the first side. Phasmophobia, Lethal Company and Dead by Daylight sit on the second. Hunt: Showdown and S.T.A.L.K.E.R. 2 straddle both, which is why they reward the same framework twice: manage your resources, then accept that you still do not know what is behind the next wall.