Sandbox is Retro Tower Defense’s testing space. Public gamepass data checked on August 28, 2026 listed the pass at 299 Robux with access to Sandbox and three extra save slots. Verify the current Roblox checkout and in-game description before buying. Sandbox is most valuable when you test one question at a time instead of building a perfect-looking board with no comparison.
What to test
Create a clean baseline
Write down the live patch date, tower tier, Tower Path, placement limit and all modifiers. Use a saved setup for the unchanged baseline. Remove support unless support is the subject of the test. Place on a marked landmark so the range comparison is repeatable.
Run the baseline twice. Targeting and control can produce small variation, so one run is not enough. Then change exactly one variable. If you alter the tower, tile, support and modifier together, the result cannot tell you which choice mattered.
Test damage without fake precision
Card DPS is useful only when the definition is known. For simple attacks, compare damage, interval, target cap and real firing uptime. For AoE, record typical enemies hit on the actual bend, not a perfect maximum. For an ability or card mechanic such as Chance’s Deal Cards, note activation conditions and downtime.
Compare equal total cash whenever possible. A maxed tower versus a half-built alternative answers an inventory showcase question, not an upgrade-efficiency question.
Test support and control
Preview the radius and count developed towers inside it. Run the same enemy without support, with one support, and—only if relevant—with a second copy. Watch whether stats or the failure point actually change; never assume stacking. Confirm whether control affects bosses or special targets.
Turn the result into a live build
Sandbox removes some pressure that a real run contains. After finding a winner, test whether it can be afforded at the required wave, whether its tile is available, and whether the team survives while saving for it. Add economy payback and opening cost back into the plan.
Use save slots for named purposes such as “boss comparison,” “AoE bend” and “support radius,” not vague loadouts. Include date and tier in the name or notes so an update does not make an old result look current.
Quick protocol
- State one question.
- Record the baseline settings.
- Run twice.
- Change one variable.
- Run twice again.
- Compare total cost and the first failure.
- Confirm affordability in a real mode.
Sandbox can show what a tower does; it does not automatically prove what is best for every map. Use it with the DPS, Support and Farm comparison methods.
Useful saved scenarios
Keep one low-cost opening test, one grouped-enemy bend, one durable-target line and one support cluster. Name each save with the patch date and the variable being tested. This makes later balance checks faster and prevents an old setup from being mistaken for current evidence.
For placement tests, use visible landmarks and a screenshot so range circles can be reproduced. For upgrade tests, reset cash and enemy state. For support tests, count the exact developed towers affected and confirm that special targets receive control.
What Sandbox cannot prove
It may not reproduce the economy pressure, unlock restrictions, teammate behavior or timing of a normal match. A tower that wins with unlimited preparation can still be a bad loadout slot if it cannot be placed before the relevant wave. After every successful comparison, write the real-match purchase sequence and test it once in the selected mode. If the live result differs, investigate affordability, path coverage and modifier rules before discarding the controlled Sandbox result.