Feature

Fight what you think they have

Most strategy games show every nation's army count perfectly, every turn, free of charge. NationFall does not. If you have no spy network in a country, your estimate of their strength swings by up to 45% - they could be twice as strong as you think, or half.

Building intelligence is not a side activity. It is the difference between a campaign that works and a campaign that walks into something it didn't see coming.

The 45% problem

With no spy network in a target nation, your strength estimate is bounded by ±45%. The intel you see is a sampled estimate, not the truth. Plan an invasion against an enemy you think has 100 tanks; you may be invading 145, or chasing 55.

Every spy network level cuts the error band:

The UI shows you a center estimate plus an envelope. As your network gets sharper, the envelope tightens. As it goes stale, the envelope widens.

Neighbors and allies see clearly

Two relationships override the fog entirely.

• NEIGHBORSShared borders mean shared visibility. Adjacent nations always have perfect intel on each other - no spy network required.
• ALLIESMilitary alliances share intelligence. Pool a coalition and the whole bloc sees through each other's eyes - and your shared targets sharpen for everyone.

The strategic implication: distant powers are dark. The intelligence challenge is reaching across oceans, not staring across the border.

Stale intel drifts

Even good intel decays. Every observation is timestamped. As turns pass without an update, the engine drifts your last-known number using a noise model:

drift = (RNG × 2 − 1) × 0.3 × min(turnsSinceSeen / 12, 1.5)

Twelve turns since you last saw their army? Your estimate has drifted up to ±30% from where it was. Eighteen turns? Up to ±45%. The longer you go without fresh observations, the more your map is fiction.

In the modern WW3 scenario this is brutal. A rival you stopped watching three months ago has had time to mobilize. The number you "know" may be the number from before they started.

Recon ops cut through

A Recon Op is the surgical tool. For $1B and an Intel Level 1 network, you get exact unit counts and current war weariness on the target - for three turns.

During those three turns the envelope collapses to zero. After it expires, drift starts again. Recon ops are how you confirm the picture before you commit forces. Used right, they are the difference between a planned strike and a discovery - your discovery, not theirs.

The AI sees the same fog

Critically, AI nations are subject to the same intelligence model. Their decisions are made on their estimates of your strength, not on the truth. An AI that has no spies on you may misread your buildup completely - and either back down too soon or escalate into a force it didn't know was waiting.

This is what makes deception possible. Mass forces in a province the AI hasn't recon'd lately. Bluff with smaller forces in plain sight. The fog cuts both ways.