Plan your next battle: get a recommendation from the planner, then use the count-ratio math and worked examples below to design an experiment the physics can actually answer.
Claim listed codes first, then spend rewards only after you know which upgrade path stays useful.
Use exact game rates when they are available. Until then, treat this as a conservative planning helper.
The game publishes no creature stats, so the only lever arithmetic can touch is how many of each side take the field. That makes ratio math the calculator's entire job — done honestly, on numbers you control.
The ratio is the handicap one side is carrying. A ratio of 1 is a fair fight; the official examples live at the extremes — 1000 vs 10 is a ratio of 100, and 500 vs 200 is a ratio of 2.5. Reading a battle as a ratio converts it into a question about creature worth: at ratio 100, each Kaiju must effectively be worth 100 Mechs for the fight to end anywhere near even.
Doubling both sides preserves the ratio: 2000 Mechs vs 20 Kaijus is the same experiment at a bigger scale, which matters because the like-unlock system keeps raising the game's maximum battle size. Scaling one side only rewrites the ratio: 500 Mechs vs 10 Kaijus halves the handicap to 50 and is a genuinely new question.
Single battles produce winners; a walk of ratios produces knowledge. If 500 Tigers lose to 200 Elephants but 700 Tigers win, the honest conclusion is that the crossover sits between 2.5 and 3.5 — a bracketed result that survives physics variance far better than any single scoreboard.
These are the only battle numbers the game has ever published, verified on the official page on 2026-10-03. They are the right starting points because every source for this site can confirm them.
Source: official Roblox game page description, checked 2026-10-03.
A planning tool is only as good as the boundary you respect around it. Three rules keep the output useful.
When a ratio-100 battle ends one-sided, the result is solid. Extreme setups overwhelm physics noise — the outcome carries real information about creature worth.
At ratios near 1, the documented chaos dominates: rerun the same setup twice before believing it. Two disagreements tell you the ratio is inside the noise band.
The experience publishes no stats and runs 1 player per server, so there is no leaderboard math to import. Your recorded trials are the database.
Two things: the planner above turns your stage and goal into a next-battle recommendation, and the ratio math on this page turns two army counts into the number that actually matters — the count ratio between the sides. All inputs are yours; nothing is fetched from a hidden game database, because the game publishes no stats.
Simple division with a twist worth remembering: divide the larger count by the smaller to get the handicap. The official example of 1000 Mechs vs 10 Kaijus is a ratio of 100, meaning each Kaiju must be worth 100 Mechs for the fight to be even. Halve the Mechs and you halve the required Kaiju value to 50.
From the official Roblox description itself, which stages 1000 Mechs against 10 Kaijus and 500 Tigers against 200 Elephants. Both were verified on the official page on 2026-10-03. The game publishes no other battle numbers.
No — and a tool that claimed to would be guessing. The game resolves battles through physics simulation with emergent chaos, so outcomes carry run-to-run variance. The calculator prepares the experiment; the battle answers it.
Because the right next battle depends on context. A brand-new player should first reproduce the official examples at small scale; an experienced tester deciding between farming knowledge or chasing the bigger-battle ceiling gets a different recommendation. The boost toggle stands in for whatever session constraint you are under — limited time, a loaded map, or a large count ceiling.
