The Cave Diver Route Planning
Build a multi-body campaign so the cave grows familiar on purpose, not by accident.
Route Planning is the campaign layer of The Cave Diver walkthroughs. Movement pages teach hands. Dead-end pages teach honesty. This page teaches how to spend a sequence of lives toward the flower without turning the save into a graveyard of unmarked experiments.
The full game is framed as a roughly three-to-five-hour descent for players who can learn from failure. That estimate collapses if every session reinvents the map. Signs persist across deaths; planning is how you exploit that persistence. Start from the Walkthrough Hub if you still need orientation, and keep Map plus Mechanics open while you design a plan.
Define the campaign, not the perfect line
A route plan in this game is not a speedrun string of rooms. It is a ranked list of questions you want the cave to answer. Examples: Which early fork is a hard dead end? Which branch contains a journal cluster? Which squeeze is skill-gated versus truly impassable? Write those questions before you launch a session. The Progress Checklist is a good place to store campaign milestones.
If you are still Demo-focused before the August 12, 2026 Steam release, plan Demo campaigns the same way. The Demo Walkthrough already suggests run types; route planning simply sequences them across days.
Information budget per body
Assign each life a primary job:
- Scout body: enter one new fork with strict reverse rules.
- Label body: fix ambiguous signs without seeking depth.
- Journal body: prioritize readable remnants in a chosen sector.
- Push body: attempt a known-hard squeeze only after labels are clean.
- Audit body: walk old marks cold and correct legend drift.
When a body tries to do all five jobs, it usually dies with noisy data. Read Death and Bodies to keep the fantasy of lineage tied to practical role assignment.
Sign strategy as the save system
Your durable map is the sign field. Before a campaign day, decide severity language with Sign Legend and test edge cases in the Sign Planner. Then commit: no improvisational glyphs mid-panic. If a new situation appears, pause in a safe pocket and extend the legend deliberately.
Signs explains persistence; this page tells you when to spend the action of placing one. Place marks at decision points, confirmation points, and failure borders. Do not spam marks in linear corridors where they add noise. Noise is how route planning becomes route fog.
Branch sequencing
Use Branching Paths to classify forks before you schedule them. A good sequence often clears cheap information first: soft side tunnels that can be scouting quickly, journal-adjacent pockets, and false shortcuts that would otherwise keep tempting you during push runs. Hard irreversible-looking drops come later, when your legend can absorb a stuck death without confusion.
Early Cave Routes and Dead Ends feed this sequencing. Early cave discovers your personal failure modes. Dead ends convert those modes into borders. Route planning schedules the next border to test.
Lore goal without lore derailment
The flower story matters. Generations told of a bloom that can cure ailments, even paraplegia of the legs. Men enter and do not return. Some blame the cave; others speak of an older evil. That narrative pressure is part of the horror. A strong route plan acknowledges the goal while forbidding unplanned depth spikes “because the story wants it.” Journals help—see Journals—but journals are not permission to abandon entrance marks.
Session templates you can copy
Ninety-minute scout night: fifteen minutes reviewing signs and Controls Cheatsheet, forty-five minutes on one fork question, twenty minutes labeling, ten minutes notes.
Deep push night: only allowed if yesterday’s audit found no ambiguous marks on the approach path. Push bodies that ignore this rule recreate Demo chaos at full-game scale.
Recovery night after rage: no new forks. Only Dead Ends style honesty and checklist updates. This protects the campaign from emotional debt.
Hardware and expectation checks belong in the plan
Route planning fails if your PC cannot sustain focus. Confirm System Requirements (Windows, Vulkan-capable GPU, and the rest of the Steam minimums) before a long campaign weekend. Use Demo vs Full and the Review Hub to set length expectations so you do not abandon a good map because the game is “taking too long” relative to a trailer.
Cross-links to keep beside your notes
Movement baseline: Controls, Guides. Spatial brain: Map Hub, Sign Legend, Branching Paths. Systems brain: Mechanics, Signs, Journals. Player tools: Tools, Progress Checklist, Sign Planner.
A simple campaign board
Column one: known safe spine of signs. Column two: open questions ranked by cost. Column three: confirmed dead ends. Column four: journal findings that change risk reads. Column five: next body’s assigned job.
Update the board after every life. If you cannot update it, the life did not finish. That standard sounds harsh; it is how a punishing physics-horror cave becomes solvable.
Closing principle
The cave grows familiar with every new body only if you force familiarity through marks and plans. Without planning, familiarity is just trauma recognition. With planning, familiarity becomes a route. Keep crawling toward the flower, but make each crawl answer a question you wrote down before your hands touched stone.
Frequently Asked Questions
Quick answers to common The Cave Diver questions.
Do I need a written plan for short Demo sessions?
Even a three-line plan helps: one question, one fork, one marker update. Short sessions without intent often repeat the same death.
How many new forks should one body open?
Prefer one primary new fork per scout body. Extra forks are allowed only if entrance marks and reverse plans are already clean.
What if my plan breaks mid-run?
Switch the body to a label job. Protect the map first, then rewrite the plan after the session instead of improvising depth.
When should I attempt a major push run?
Only after an audit finds no ambiguous signs on the approach and your checklist shows the required scouts are complete.