Horror Thursday — Dead Air Floors

The trigger

Two demand signals line up cleanly today. First, players are explicitly asking for more 2-player co-op games, including horror-adjacent picks, not just the same usual recommendations. Second, a game design thread is specifically asking how to make radio communication risky without making silence the obvious best play. On the market side, consumer-trust and ownership news points the same direction: a premium horror game that works offline and uses Steam Remote Play Together is easier to trust, easier to ship solo, and avoids the legal and support burden of accounts, moderation, and always-online services.

The idea

Dead Air Floors is a 2-player asymmetric horror game set in a dark office tower during a lockdown. One player is the Explorer, physically moving through blacked-out floors to restore elevator access and reach an exit. The other is the Navigator, sitting in a security room with a delayed, unreliable building map, a few camera feeds, door controls, and partial system logs. They need each other: the Explorer can see blocked halls, broken glass, and the hunter's traces; the Navigator can see route structure, power states, and safe-room locks. The catch is that using the building radio is a gameplay action. Every broadcast refreshes shared information and helps solve navigation, but it also emits a detectable pulse that lets the hunter narrow in on their floor and zone.

The core loop

• Navigator checks cameras, map drift, and locked routes.
• Explorer advances carefully, scavenging batteries, fuse parts, or override cards.
• One of them triggers a radio broadcast to sync critical information: route hints, door unlocks, breaker status, or objective location.
• The hunter reacts to the signal pulse and changes patrol pressure.
• Explorer hides, diverts power, or takes a longer route while the Navigator scrambles doors and cameras to buy time.
• Reach the next objective, restore one more system, and decide whether the next call is worth the noise.

The important detail for scope: the game risk is attached to a deliberate in-game broadcast button, not microphone analysis. Real conversation can happen however players want, but the useful building data only updates when someone commits a noisy transmission.

Why now

There is steady demand for co-op horror, but much of the market is either content-heavy online multiplayer or streamer-chaos party horror. This idea fits a smaller gap: tense 2-player coordination with a strong hook that reads instantly in screenshots and streams. It also benefits from current consumer mood. An offline-friendly, one-time-purchase co-op game is easier to position positively while players are talking more about ownership, service shutdowns, and trust. Steam Remote Play Together further expands reach without forcing a solo developer to build and maintain full online netcode.

Scope

First playable milestone:

• 1 building with 3 reusable floor tilesets
• 1 hunter behavior set
• 2 fully functional roles: Explorer and Navigator
• 1 objective chain: restore power -> unlock elevator -> reach roof exit
• 1 radio pulse system that updates shared info and raises hunter threat
• 1 local/Remote Play Together build on a single PC, no backend

That is enough for a demo and enough to prove the pitch. For a 3-6 month commercial version, keep it to one building, 4-6 scenario variants, a small pool of map lies, and escalation modifiers instead of extra monsters or a campaign.

What kills it

If silence becomes the dominant strategy, or if the Explorer can mostly solve navigation without the Navigator, the game dies immediately. The whole project depends on players feeling that they must transmit, but never casually. If that balance is off, this becomes either a tedious stealth maze or a weaker version of existing co-op horror like Phasmophobia.

Prototype only two things first: the unreliable map and the radio pulse. If those do not create real arguments about when to speak, do not build the rest.