A boss-wave layout should answer one question: can the base maintain useful pressure after the first contact point fails? Current community guides connect bosses to later progression and rewards, but they do not provide a stable official attack table. Build around observed lane behavior and preserve a normal-wave baseline instead of guessing a specific script.
Capture the normal-wave baseline
Save the most reliable wave layout before moving anything. Record the item at each contact point, where ranged damage begins and ends, the first lane to leak, and the final core condition. If layout slots are available, copy it into a separate boss slot. Otherwise, take enough notes to rebuild it.
Run the baseline once more under the current server and update. A result from a previous event is not a clean comparison. Use the same starting resources and avoid buying a permanent upgrade between the baseline and first boss test.
The Wave Layout guide provides the normal testing method. Boss preparation begins only after that arrangement clears consistently enough to reveal what the durable target changes.
Create three defensive layers
The front layer establishes contact and gives weapons time to act. It should not contain every valuable item. The middle layer holds protected damage and another delay or control point. The final layer catches a breach and prevents a single local failure from immediately exposing the core.
Spread important weapons across useful coverage zones. A concentrated cluster can look powerful until a boss reaches it and removes the entire damage plan at once. Distributed pressure also reveals which segment contributes most. Keep one fallback capable of acting even when the front has collapsed.
Use traps where the observed path reliably crosses them. Do not place a trap solely because a copied screenshot uses that tile; your inventory, range, update state, or lane behavior may differ.
Measure sustained uptime
During the attempt, record when each major weapon starts attacking, when it stops, and why. A weapon can lose uptime because the target leaves range, the front breaks too early, another lane becomes uncovered, or the weapon itself is removed. Those causes require different fixes.
Compare the boss run with the normal baseline. If the same lane fails before the boss matters, repair the general layout. If the base handles groups but stalls on the durable target, increase sustained pressure or contact duration. If damage is sufficient until one breach removes everything, distribute key pieces.
Do not rank by visual effects. The useful metric is whether a controlled change delays the first failure, protects more uptime, or improves the final core result.
Change one variable per run
Choose one adjustment: move one weapon, replace one block, relocate one trap, add one fallback, or buy one justified upgrade. Keep all other positions stable and rerun the same encounter when possible. Write the observed result before making the next change.
If the encounter is event-timed, use a comparable normal wave between event windows to test placement. The official Tsunami description says the event occurs every hour as of 2026-08-05, but the live countdown decides availability. Do not wait indefinitely for a run when static layout testing can continue.
A simple run record includes date, server or update label, encounter name, starting layout version, one change, first failure, final result, and visible reward. Screenshots are helpful evidence, but the written change is what makes two attempts comparable.
Spend upgrades after placement checks
When moving a piece fixes the failure, save the currency. When the same well-covered item still falls short, inspect the live permanent upgrade panel and choose the smallest purchase expected to change the result. Retest immediately before buying something else.
More rolls are justified when the inventory lacks a role such as durable contact, protected pressure, or a fallback. A strength upgrade is justified when the right role already exists but a controlled test shows a numerical limit. Economy is justified only when its return arrives before the next meaningful boss investment.
Use the Upgrades guide for the decision order and Best Base for broader failure-led design.
Record rewards without inventing rates
After a clear or failed attempt, capture the exact reward screen. Record names, quantities, rarity labels, and whether the item goes to inventory, Index, relic storage, or crafting. One reward does not prove that it is guaranteed, and several misses do not prove that it is impossible.
If a relic or material appears, preserve the first copy until its destination is understood. Continue with Relics and Crafting rather than relying on an undated rate table.
Boss-wave mistakes
- Overwriting the only stable wave layout.
- Stacking every valuable weapon behind one front block.
- Buying several upgrades before comparing a run.
- Calling a weapon weak when it was out of range or idle.
- Publishing an attack sequence from one uncontrolled attempt.
- Treating one reward result as an exact drop rate.
Should I use my normal layout for bosses?
Use it as the baseline, then adapt a copy for sustained and distributed pressure. Preserve the original for wave and offline progress.
What should I change first after a loss?
Change the item or position responsible for the first repeatable failure. Do not redesign from the final screen alone.
Are exact boss attacks listed here?
No. The audited evidence did not support a stable boss-by-boss attack catalogue, so live observation takes priority.