What Attack Ratio Should You Use?

For wilderness, 6,600 troops on plains is the ceiling — past that your advance does not get one tile faster. For player attacks, your losses bottom out at 1.67× the defender’s troops, and everything above that is thrown away.

Pushing the slider to 100% is the most common mistake in this game. Here is what the code actually does.

What the slider controls

The slider sets the fraction of your troops sent on the next attack. Those troops leave your pool the moment you click.

They cost you twice:

1. They die in combat
2. While they are out, they are not generating growth

The second one is bigger. Troop growth scales with your current troop count, so emptying your pool slows the refill itself. See Troop Cap and the 42.2% Rule.

Wilderness — 6,600 and no further

Unclaimed land is simple. Each tile costs a “budget”:

budget per tile = clamp(2000 × terrain speed ÷ attacking troops, 5, 100)

Lower is faster. But the floor is 5, so past a certain troop count nothing improves.

Terrain	Troops for max speed	Troop cost per tile
Plains	6,600	16
Highland	8,000	20
Mountain	10,000	24

You start with 25,000 troops, so a 26% slider already saturates plains expansion.

So what do you do with the rest

Expand in several directions at once. Advance rate scales with border length, so three fronts beat one.

Early game, hold the slider at 10–30% and click in multiple directions. The opening simulator lets you compare rhythms directly.

Tribes (bots) lose only half as much taking wilderness (÷10 instead of ÷5). That is why they appear to spread so fast early.

Player attacks — two different thresholds

Here is where the common advice goes wrong. The community calls it the “2:1 rule,” but the code contains two separate clamps with different break-even points.

What you are optimizing	The clamp	Troops needed
Minimizing losses	clamp(defender ÷ attacker, 0.6, 2.0)	1.67× defender
Advance speed	clamp(defender ÷ (5 × attacker), 0.2, 1.5)	1.00× defender

Read it like this:

• Matching the defender’s troop count already maxes out your advance speed
• Bottoming out the loss multiplier takes 1.67×
• Past 1.67×, neither number improves

There is a ceiling on the other side too: even against a vastly larger player, your loss multiplier stops at 2.0. So sending a timid trickle is the worst option — you eat the same 2.0 multiplier and take no ground.

40% of your losses ignore troop count entirely

This is missing from every guide I have seen. Attacker losses are a blend of two terms:

attacker loss = 0.6 × (troop-ratio term) + 0.4 × (defender density term)

Only the first 60% responds to the ratio clamp above. The remaining 40% does not change no matter how many troops you send. It is driven purely by the defender’s troops ÷ tiles owned — their defensive density.

Practical consequences:

• Driving losses toward zero by piling on troops is structurally impossible
• Instead, pick targets whose troops are spread thin, and that whole 40% shrinks
• A player with lots of land and few troops is your best target. Long borders make density worse

Situation guide

Situation	Recommendation	Why
Early wilderness	10–30%, several directions	Plains saturate at 6,600
Encircleable small player	Just enough to close the ring	Full encirclement costs zero troops
Comparable opponent	1.7× their defending troops	Loss clamp floor
Much larger opponent	Commit properly or not at all	Loss multiplier stops at 2.0
Thinly-spread large player	Attack	The 40% density term is small
Defended by a defense post	Go around, or skip	×5 losses, ×3 slower

Sent the wrong amount? Pay 25% and pull out

The red ❌ in the event panel cancels an attack in progress. It costs 25% of the attacking force; the other 75% comes home.

Paying 25% to stop feeding a losing fight is usually correct. Returning troops go straight back to work as growth.

Nukes cannot be recalled once launched.

More

• Opening simulator — compare slider values over 60 seconds
• Troop cap calculator — your cap and growth peak
• Troop Cap and the 42.2% Rule — why hoarding costs you