Before launching the first run
Confirm the practical basics before strategy. The live English minimum set lists an Intel i5 or newer-generation i3 / AMD equivalent, GTX 1060 6 GB, and DirectX 11. Recommended lists an Intel i7 or AMD equivalent, including AMD FX 9500+ Series in the checked copy, RTX 3060, and DirectX 12. Both list Windows 10 64-bit, 16 GB RAM, broadband, and 15 GB available storage.
Decide whether the first session is solo or co-op and set an appropriate learning goal. The official store supports single-player and online co-op for one to four players. A group should agree on short callouts and patience around interactions. A solo player should plan more conservative information gathering because there is no second perspective to catch a missed route, resource decision, or environmental change.
- Set voice or text callouts before entering.
- Choose one system to learn rather than every system at once.
- Keep the official store and this guide’s verification date available.
- Do not import demo-era assumptions into the current build.
Use the first minutes to establish control
Resist the urge to convert forward motion into the only measure of progress. Look at the room, test movement space, identify usable objects, and note the route back. Procedural generation makes observation more dependable than a route remembered from somebody else’s run. When an interaction changes the environment, pause and update the mental map. The room after an interaction may demand a different plan from the room before it.
In co-op, one player can describe the forward space while another watches the retreat and a third tracks important resource decisions; smaller groups can combine responsibilities. These are editorial roles, not claims about the four in-game classes. In solo, alternate between forward observation and a deliberate rear check. The routine feels slow only until it prevents the group from becoming confused under pressure.
Make the first spell decision explicit
Casting uses FAITH, and lower FAITH exposes the run to FEAR according to the official description. For the first few casts, name the problem being solved. Was the spell used to create space, obtain information, support movement, or answer immediate danger? A precise purpose makes it easier to judge the result without relying on undocumented values. It also teaches when another tool or a reposition might have preserved FAITH.
Avoid the opposite mistake of refusing every spend. A saved resource has value only if the team survives long enough to use it. Compare the consequence of spending with the consequence of delay, then keep a fallback. Beginners learn the system faster when each use has a hypothesis. If the outcome differs from that hypothesis, record the observation without declaring a permanent mechanic from one event.
Coordinate without crowding
Staying together does not require occupying the same patch of floor. Maintain enough spacing to see, move, and interact without blocking each other, while keeping the group close enough to share information. Before a risky action, say who is acting and where others will hold. This prevents several players from responding to the same problem while another direction goes uncovered.
Use compact calls that describe facts: location, movement, interaction, resource intent, and retreat. Avoid a stream of guesses. If somebody is uncertain, the useful response is to slow the decision, not demand false confidence. When the bunker separates attention, repeat the current objective. Shared intent is more valuable than everybody holding a different detailed plan.
Recover from confusion before pushing
When a room becomes chaotic, stop adding new decisions. Move toward the last understood position if it remains safe, restore spacing, and restate the known facts. Teams often compound uncertainty by opening another route, using another resource, or changing several roles at once. A recovery phase should simplify the situation. Solo players can use a physical pause to rebuild the route in memory and identify the next reversible action.
If the run ends, review only a few moments: the first lost retreat, the first unexplained spend, and the first time the team’s shared picture diverged. Those are actionable. Blaming a class, spell, or random layout without evidence teaches little. The next run should change one behavior, making the comparison understandable even when procedural generation produces a different bunker.
Choose the next briefing from the failure
If resources felt opaque, continue to the FAITH and FEAR guide. If route uncertainty dominated, study bunker doctrine and the generator tradeoff. If responsibilities overlapped, use the builds page as a capability-planning framework rather than a loadout prescription. Co-op and solo pages address the different information burdens of each mode without pretending their fundamentals are unrelated.
Check the article verification stamps whenever a system appears different. Early Access can change balance and terminology, and no guide should overwrite direct current evidence. The known-issues page also explains how to distinguish an official present notice from community reports and historical demo patches. That distinction keeps beginner learning grounded in the build that actually exists.