Understand the verified relationship
The official description gives a directional relationship rather than a complete numerical model: spells consume FAITH, and lower FAITH increases exposure to FEAR. That is enough to shape tactics, but not enough to publish exact breakpoints, rates, or guarantees. The safe conclusion is that a cast changes both the immediate situation and the risk context of later decisions. Every spend deserves a purpose and an after-plan.
Spell use is not the only pressure on the resource. The official store also says certain eldritch horrors can drain FAITH. Avoid turning either source of loss into a universal rate or threshold without current official support. Read live interface cues directly, note whether pressure came from a cast or a threat, and preserve enough clarity to change the plan when the source of the drain changes.
Use the five-question decision loop
First, name the immediate problem. Second, ask whether movement, timing, a weapon, teamwork, or a physical interaction offers another answer. Third, state what the spell should accomplish. Fourth, identify the fallback if the result is incomplete. Fifth, reassess FAITH and the environment after the cast. This process can take seconds once practiced and prevents an old plan from surviving after its assumptions have changed.
The loop does not imply that conservation is always correct. A resource protected through a failed run provided no later value. It asks the player to compare risks explicitly. Under pressure, use a shortened call: problem, cast, opening, fallback. Teammates then know how to support the decision and whether another simultaneous spend would duplicate it.
- Problem: what must change now?
- Alternative: can position or environment answer it?
- Purpose: what opening should the spell create?
- Fallback: what if the result is incomplete?
- Recheck: what changed after the spend?
Protect a purpose, not a magic number
A reserve is most useful when connected to a job. The group might protect an answer for escape, recovery, information, or control near a consequential transition. That purpose can change as the bunker changes. An arbitrary percentage copied from another build can create false confidence, while a purpose-based reserve tells the squad what loss it is accepting if it spends now.
When the protected situation arrives, use the reserve deliberately rather than continuing to save from habit. If new information makes the original purpose irrelevant, allocate it again. This flexible budgeting acknowledges Early Access volatility and procedural variation. It also supports solo play, where the same resource may need to cover more than one responsibility and conservative thresholds may be appropriate.
Respond to FEAR pressure by simplifying
When pressure rises, reduce the number of new commitments. The official description names alcohol and ally spells as ways to reduce or manage FEAR. An active generator also restores FAITH, linking the light decision to the same resource problem. These are recovery methods to consider against the current situation, not evidence for exact quantities, timers, or a guaranteed rescue.
Failing to manage the state carries a transformation risk described as becoming a GHOUL or something worse. That is a real reason to respond, but it does not justify inventing a meter breakpoint. In co-op, state the observed change and name the recovery action so teammates do not duplicate it. In solo, pause when possible, rebuild the route picture, and choose the next reversible action. Panic spending can deepen the problem, while refusing an urgent response can also end the run.
Coordinate one shared resource picture
Individual decisions accumulate into group exposure. Before a planned sequence, announce which player acts first and what signal allows the next action. Afterward, summarize whether the intended opening exists. This keeps several players from solving the same problem independently. The summary should remain qualitative unless the live game explicitly provides stable values everyone can see.
Assign one player to restate the group plan during a complex moment, but let leadership move with information. A player watching the retreat may understand movement risk better than the caster. A player at an interaction may understand environmental timing. Shared resource discipline comes from combining those perspectives, not from giving one person an unsupported formula.
Review decisions in build context
After a run, choose one or two consequential spends and reconstruct the decision: known facts, intended purpose, alternative considered, outcome, and remaining options. Record the date or build context. Do not conclude that an option is universally efficient from one procedural layout. The review should improve the next decision loop, not manufacture a permanent ranking.
Check official announcements after updates and compare this page’s last-verified time with the current build. Historical demo patch notes can explain why old reports circulate, but they do not establish present behavior. The known-issues page keeps those evidence categories separate. For practical application, continue to the spell guide or generator briefing.