Sandbox is the controlled testing mode in Anime Expeditions. The Miraheze reference says it lets players select an available map and manipulate five control categories: Game, Units, Enemies, Modifiers and Maps. Use it to answer one focused question—whether a placement reaches, whether an effect applies, or which upgrade timing fixes a leak—rather than spawning everything and learning nothing.
Interactive Sandbox control guide
Game controls
Set the pace of the experiment, then reset between samples. The reference notes that allowing base stocks to reach zero does nothing in Sandbox, so a test can continue after a situation that would end a normal match. Do not mistake that safety behavior for Story, Raid or Tournament rules.
Unit controls
Place the exact form, level, trait, potential and equipment state you want to understand. Compare one change at a time. If range is the question, keep enemies and modifiers fixed. If damage is the question, use the same target path and duration for both samples.
Enemy controls
Spawn a target that represents the failure you saw in a live mode. The reference lists a manual HP cap of 99.9 million, while naturally spawned enemies can exceed that value. Treat the number as a reference snapshot and verify the input limit in your server.
Modifier controls
Recreate the resistance, speed, durability, control or other modifier that caused the failure. Confirm that the Sandbox label matches the live stage label. A test without the relevant modifier may prove the unit's base performance but not the matchup.
Map controls
Select the same lane as the live failure whenever possible. Range and attack timing depend on path shape, so a unit that looks strong on a long overlap can perform differently on a short bend or split route.
Interactive map test selector
School Grounds
Benchmark opener cost, range across the courtyard and the five Story bosses. It is a practical first map for consistent before-and-after tests.
Flower Forest
Test speed, splitting, cleansing and long overlap placement. Keep the target and modifier fixed when comparing crowd control.
Rose Kingdom
Test sustained damage against durable or regenerating targets and measure whether separated damage zones allow recovery.
Fairy King Forest
Test support-target priority, armored pressure and boss reserves without changing the unit lineup between samples.
King's Tomb
Test summons, support removal, clone boss timing and whether essential placements are too close to one disruption.
Spirit City
Recreate a Raid Act with its boss and stage effects. Match the current Act card before trusting the result.
A reliable testing method
Write the question before touching a control. Good questions are narrow: “Does this effect apply to the Act 3 boss?”, “Does Upgrade 6 clear the first armored wave?”, or “Which of these two placements gives longer uptime?” Set the map, target, HP, modifiers, unit state and starting economy. Run a baseline, change one variable, and repeat for the same duration.
Record results that can be reproduced: target defeated at a specific time, HP remaining after a fixed interval, first leak wave, or whether a status icon appeared. A single large damage number is not enough when attack speed, multi-hit behavior or range is the real issue.
Move the result back to live play
Sandbox removes normal failure pressure, and the reference says base stocks reaching zero does not end the test. Live modes can also add reward conditions, player scaling, rotation rules and effects not reproduced by a basic setup. Treat Sandbox as evidence about the isolated mechanic, then perform one safe Story or lower-Act verification before committing resources.
Common Sandbox mistakes
- Testing several unit, target and modifier changes at once.
- Forgetting that zero base stocks do not end a Sandbox test.
- Using a different map from the live failure and comparing range directly.
- Treating the documented 99.9M manual HP cap as a permanent patch rule.
- Publishing a damage result without unit form, level, trait, equipment and duration.
- Assuming a Sandbox result guarantees Tournament or Raid performance.
Sandbox FAQ
Which maps are documented?
School Grounds, Flower Forest, Rose Kingdom, Fairy King Forest, King’s Tomb and Spirit City.
Can manually entered enemy HP exceed 99.9M?
The reference says the manual control is capped at 99.9M, while naturally spawned enemies can exceed it. Verify the current control.
Does losing all base stocks end the test?
The reference says it does nothing in Sandbox. Normal modes should not be expected to behave that way.