- Upgrade your troop camps first.
- Pick % damage over % attack.
- Pick % HP over % defence.
- Run 4–5 heroes of one faction.
- Use % damage buffs (like Clear the Way) before % attack buffs (like Scouting).
- Check your own totals in any battle report.
What is the damage formula?
Every hit in a troop battle is roughly (100% + Total ATK) × (100% + Total DMG) × the skill's %. Both totals are in your battle report.
- Troop's Total ATK and Troop Total DMG are on the Stat Overview of every battle report. Mine were 2,399.5% and 752%.
- Both are plain sums. Tap Troop Buffs and you see every source added up — heroes, research, vehicle, gear, pet, VIP and the rest.
- Buffs add points to these sums. They don't multiply them. That one fact decides everything below.
- The counter bonus counts inside Total DMG.
One part is still unknown: the enemy's defence and the game's base number. Those decide the exact damage against a new target. They don't change which upgrade is best, so the ladder below holds without them.
What do these percentages multiply?
Your troops — how hard each one hits (set by its camp's level) and how many you send (set by your heroes).
That starting number is the biggest lever of all: one camp Industry level is worth more than every attack buff in the game combined. It has its own short article: where your damage really comes from.
How did I work it out?
I fought the same Lv.60 Rotting Oni again and again, changing one thing each time, and compared every damage line in Round 11.
From Round 2 onward this game gives the same result every time. My control fight at 16:27 matched a fight at 13:50 to the last digit. So when one thing changes, you can see exactly what it did.
Test 1 — an attack buff: War Frenzy (+5% ATK)
Scouting turns on War Frenzy: +5% attack for 14 minutes. If buffs multiplied my 2,399.5%, damage would rise about 4.8%. If they add, it should rise about 0.2%.
Result: every skill went from 138,420 to 138,720 — +0.22%. Buffs add.
Test 2 — a damage buff: Clear the Way (+6% DMG)
Clear the Way is an agent skill: +6% troop damage for 2 hours. My Total DMG went from 752% to 758%.
Result: every line rose +0.67% — exactly what (100 + 758 + 42) ÷ (100 + 752 + 42) predicts. Damage buffs add too, and the +42% counter bonus is part of the same total.
Test 3 — gear
Moving one helmet between two heroes changed my Total ATK by 11 points and all damage by 0.45%. Full write-up: damage is pooled, but who wears the gear still matters.
Why is % damage worth more than % attack?
Because one more point is worth 1 ÷ (100 + your current total). Your attack total is huge, so each new attack point barely moves it.
My attack total is 2,399.5%. My damage total is 752% plus the 42% counter. Add one point to each:
| +1 point of… | Math | Damage gain |
|---|---|---|
| % attack | 1 ÷ 2,499.5 | +0.040% |
| % damage | 1 ÷ 894 | +0.112% |
2.8 times more. Nothing magic about damage — its total is just smaller, so each point is a bigger share of it.
Mid-game player? The same rule works for you; only the numbers change. As an example, if your attack total were 600% and your damage total 150%, one attack point would be worth 1 ÷ 700 = 0.14% and one damage point 1 ÷ 250 = 0.40% — still nearly 3×. Check your own totals; don't take my numbers as yours.
Is there an upgrade ladder?
Yes. Raise your troops first — camp level and troop count. Then % damage, then % attack. % HP and % defence are a separate ladder, for survival.
| Rank | Stat | Why | How sure |
|---|---|---|---|
| 1 | Troop camp level and troop count | They set the starting number every % multiplies. Camp Industry 7 → 8 = +8% attack per troop. More. | Shown in-game; fight test pending |
| 2 | % Damage (incl. counter damage) | Smallest total, so each point counts most. +0.112% per point for me. | Measured |
| 3 | % Attack | Every point counts, but it's diluted by a huge total. +0.040% per point for me. | Measured |
| 4 | Hero stats (a hero's own ATK) | Hero ATK only matters through the gear that hero wears (one helmet swap: 0.45%). Hero level, stars and gear also raise troops led — that part counts under rank 1. | Measured once |
| — | % HP and % Defence | Don't add damage. They decide how long your troops survive. My HP total (647%) is much smaller than my defence total (2,237%), so the same rule suggests an HP point protects more than a defence point. | Reasoned, not tested |
"Troop stats" come in two parts: each troop's own stats from its camp (rank 1) and the % totals in the report (ranks 2 and 3).
What about the counter bonus?
It's the biggest damage number most players already have. Run 4–5 heroes of one faction and it adds 25% to 60% troop damage.
- It triggers against the faction you counter and against "non-faction enemies". My Oni report shows the +42% active against a zombie boss.
- Because it sits in Total DMG, my +42% is worth about +4.9% damage. All four attack buffs I can stack (+42% ATK) are worth about +1.7%.
- Going from 4 to 5 heroes of one faction would take my counter from +42% to +60% (per the panel) — about another +2% damage, plus ×135% hero ATK/DEF instead of ×125%.
What should I do with this today?
- Spend Precision Parts on your main troop camp first. Its Industry level sets each troop's attack, which every % multiplies.
- Open your last battle report → Stat Overview. Write down Troop's Total ATK and Troop Total DMG.
- Add 100 to each. The smaller result is where every point is worth the most. For almost everyone, that's damage.
- When you choose between skill books, research or buffs, prefer the one that says damage over the one that says attack.
- Keep at least 4 heroes of one faction in your main march so the counter bonus is on.
- Put your best gear on your highest-ATK heroes. Small, free, worth it.
All of this comes from one account fighting one monster. The enemy's defence and the game's base number are still unknown, so this predicts how much an upgrade helps, not the exact hit against a new target. I haven't tested whether hero lines like Darian's "troop skill damage" join Total DMG or work on their own. And Round 1 varies a little between fights — compare Round 11.