On this page
QUICK ANSWER
Direct answers
What is the best first tree?
A sustain-first tree: Waystone quantity, tier and reliable map completion before expensive mechanic or boss multipliers.
When should I specialise?
When the character can replace failed Waystones, clear the chosen mechanic without repeated deaths, and sell its rewards without filling the stash.
Which Atlas Master should I use?
Jado for general utility and unique-focused maps, Doryani for heavily modified Waystones, and Hilda only when empowered unique monsters are already comfortable.
Should I copy a finished screenshot?
No. Copy the objective and the pathing logic. Exact nodes change with patch, unlock state, and the mechanic you can actually finish.
OVERVIEW
How to use this guide
An Atlas tree is a production plan for maps, not a permanent character build. The early job is to keep Waystones flowing; the middle job is to make one mechanic repeatable; the late job is to spend risk and currency only where the build can convert them into completed encounters.
Patch 0.5 adds the Atlas Master layer on top of ordinary passives. Master selection is a per-map decision, so the useful question is not which Master is universally best, but which one improves the next Waystone without making it fail.
REFERENCE TABLE
Progression framework
Progression framework for the current article topic.
| Stage | Readiness | Priority | Stop condition |
|---|---|---|---|
| Stage 1 — establish sustain | Fresh endgame | Waystone quantity, tier progression, completion reliability | Do not buy premium tablets or boss pressure while replacing failed maps is difficult. |
| Stage 2 — choose one engine | Stable low and mid tiers | One mechanic, nearby density, and its reward multipliers | Pick a mechanic the build clears quickly and the player is willing to sell. |
| Stage 3 — improve conversion | Repeatable high tiers | Better routing, tablet coverage, map modifiers and loot handling | Measure profit after consumables, failed maps and selling time. |
| Stage 4 — add risk | Defenses and boss damage proven | Hilda, powerful bosses, high-mod Waystones, premium encounters | Test a small batch before scaling the investment. |
REFERENCE TABLE
Decision matrix
Decision matrix for the current article topic.
| Choice | Strength | Best fit | Watch for |
|---|---|---|---|
| Jado | Uniques, Strongboxes and flexible mapping utility | General progression and maps where broad value matters | Do not assume unique quantity is liquid profit; many drops have little demand. |
| Doryani | Rewards and survivability around heavily modified Waystones | Characters that already run four-modifier rare maps consistently | The conditions must be present when the map opens; a mild map cannot use the full package. |
| Hilda | Harder unique monsters and larger rare/unique reward potential | Boss-ready builds with reliable single-target damage | Empowered uniques turn a marginal build into failed maps, so begin with low-cost tests. |
REFERENCE TABLE
Scenario recommendations
Scenario recommendations for the current article topic.
| Situation | Recommendation | Action | Risk control |
|---|---|---|---|
| Waystones are running out | Sustain passives | Keep modifiers moderate and complete more maps | Adding more difficulty before completion improves nothing. |
| Maps are stable but loot is scattered | One mechanic cluster | Remove mechanics you skip and strengthen the one you finish | Atlas points spent on ignored encounters have zero realised value. |
| Four-mod maps are routine | Doryani | Run a measured batch with the same map-quality rule | Separate the Master's value from unusually lucky drops. |
| Bosses die quickly and safely | Hilda | Test empowered uniques on replaceable Waystones | Stop if deaths erase more maps than the reward gain replaces. |
| You want low-friction general mapping | Jado | Use broad utility and sell only genuinely liquid uniques | Do not spend minutes pricing every low-demand item. |
CARD GRID
Practical workflow
Run the process in order and preserve a rollback point before spending heavily.
Step 1
Write one sentence for the tree's job: sustain, one mechanic, bossing, or general mapping. If the sentence contains all four, the tree is unfocused.
Step 2
Allocate the shortest efficient path to that job, then reserve points for Waystone sustain and survival-friendly routing.
Step 3
Prepare a fixed test batch and record starts, completions, deaths, Waystones returned, saleable rewards, and consumable cost.
Step 4
Change one cluster or one Master at a time. A complete respec makes it impossible to identify the useful change.
Step 5
Keep a rollback screenshot or planner link so a failed experiment does not strand progression.
CARD GRID
Common mistakes and fixes
These failures cost more time than a small difference between two reasonable choices.
Copying a high-budget tree
Why it fails: The reference assumes damage, defenses, tablets, and market access that are not present.
Do this: Return to sustain and one accessible mechanic.
Valuing gross drops
Why it fails: Listed value ignores failed maps, consumables, unsold stock and trading time.
Do this: Track net realised value per hour.
Changing the Master mid-map
Why it fails: Master effects are chosen for the map when it is opened.
Do this: Confirm the active Master before inserting the Waystone.
Keeping every mechanic
Why it fails: Travel points and interruptions dilute the farm.
Do this: Remove encounters that are skipped or consistently failed.
DIAGNOSTIC
Quick diagnostic
Choose the symptom that best matches the current problem.
PROBLEMS & ANSWERS
Frequently asked questions
Can I respec later?
Yes. Treat the first tree as a progression tool, not a permanent commitment; keep enough currency and a recorded rollback plan.
Is Hilda always the most profitable?
No. Her upside matters only if harder unique monsters are completed consistently and the resulting items are liquid.
How large should a test be?
Use enough maps to include ordinary variance—at least a deliberate batch rather than one lucky map—and keep inputs comparable.
What if a published tree uses old node names?
Follow its priorities, then rebuild the route against the current in-game tree and patch notes.
Should every map be fully juiced?
No. Investment should rise only when completion rate and reward conversion remain healthy.
REFERENCE TABS
Staged point routes: first 20, 40 and 60 points
Use these as a routing plan, not a fixed screenshot. Node availability, league affixes and your build change the best exact path; the priorities below stay constant.
Spend the opening points on Waystone sustain and completion, not on a mechanic you cannot finish yet. A steady map flow is what funds every later investment.
- Take the Waystone Quantity and Waystone Drops cluster near the tree centre to keep maps flowing.
- Add the Map Completion and bonus-completion nodes so failed maps are cheap to replace.
- Reserve points for the shortest path to your first repeatable mechanic rather than spreading thin.
Sustain first
Allocate Waystone quantity, tier and completion before any boss or premium pressure.
One anchor
Pick the single mechanic your build clears quickly and path toward its reward nodes.
Verify
Run a small batch; only add difficulty once completion rate stays healthy.
SCREENSHOTS & VIDEO
Walkthrough video guide
Asmo
The Ultimate PoE 2 Endgame Progression Guide (Common Mistakes, Tricks, SSF & Trade)
A full early-Atlas routing guide covering the Master layer, Hilda's passive, tower and tablet strategy, and the biggest mistakes players make.
What to watch for: Unlock all three Masters early, then select the one that matches what you are about to farm; swap for free on each new map.
Open source ↗SOURCES & CHANGELOG
Sources and verification
Current 0.5 Atlas Masters and specialisation model.
CHANGELOG
Revision history
- Replaced the short summary with a decision framework, comparison matrices, a practical workflow, failure recovery, and current-version boundaries.
- Added a walkthrough video section and concrete 20/40/60 point-allocation routes (tabs); replaced the external hero image with a local webp per the image-fingerprint contract.
Information is compiled from public web sources and in-game observations; some visuals are original editorial diagrams. External information and images remain the property of their authors or platforms; contact us for removal requests.


