Frontlines Easy, Normal or Hard: What Each Changes

Difficulty in Frontlines changes five things about the AI: how often a nation decides, how willingly it attacks, how many attacks it runs at once, whether it builds ships and nukes, and a growth and gold bonus of +0%, +8% or +30%. It also changes how many of your opponents are real aggressors.

Hard is not a smarter AI. It is the same AI, faster, bolder, and paid more. Easy never builds a warship or a nuke.

We read the numbers from the game code (config.js, ai.js, game.js) and ran 72 headless games on the real engine to see what the setting does to a player who does nothing.

The three settings side by side

Everything in this table is a number in the game code or follows from one. The decision rate is derived below.

	Easy	Normal	Hard
Decisions per minute, one nation	15	22	33
Troops committed to an ordinary attack, up to	72%	82%	92%
Attacks at once	1	2	3
Traders, warships, carrier	none	traders and warships	plus one carrier
Silos and nukes	never	yes	yes
Growth and gold bonus	+0%	+8%	+30%
Aggressive nations among 9 opponents	2	4	5

Normal is the default in the menu, with 9 opponents. The counts in the last row come from creating real games and counting who spawns, not from the formula.

Hard nations decide more than twice as often

Every AI nation runs a decision pass, then waits a random time between one and two "think intervals". The interval is 26 ticks on Easy, 18 on Normal and 12 on Hard, and Frontlines runs 10 ticks per second. That averages 3.9, 2.7 and 1.8 seconds between decisions.

Each decision is a chance to expand, build or attack, so a Hard nation gets 33 of them a minute and an Easy nation 15.

The one that matters to you is the strike check. A nation that borders you and has at least 12% more available troops than you rolls a dice on every decision: 4.0% on Easy, 5.8% on Normal, 7.6% on Hard. Combined with the decision rate, that is about 0.6 strike checks per minute on Easy and 2.5 on Hard, four times as many. Alliances and truces block it, as covered in how alliances work.

Courage decides who they pick on

Strike checks are one trigger. The general one is courage: a nation attacks its softest neighbor when its available troops exceed a multiple of the neighbor's. The multiple shrinks as the match ages, falling by 45% over 50 minutes.

Setting	Troops needed over the target, minute 0	Minute 25
Easy	1.45x	1.12x
Normal	1.10x	0.85x
Hard	0.85x	0.66x

An Easy nation wants a 45% edge before it starts a war. A Hard nation attacks while it is still outnumbered. Courage is not the only way into a war: a nation also attacks when it has been attacked, when its population is full, when its target is nearly finished or when it holds a 1.3x edge.

Hard attacking at 0.85x sounds reckless, and the combat rules punish it. The price per enemy tile rises as the defender's army outnumbers the attacking force, up to a 3-to-1 gap, and the code's note on the rule is that outnumbered attackers barely dent the garrison. That is why outnumbered attacks stall. Expect a Hard neighbor to start some wars it cannot finish.

The growth bonus is the real handicap

Population growth and gold income both multiply by one number. For a human it is 1 plus 0.25 for a capital plus small boosts from railways and airports, capped at 1.6. For a bot, the code adds the difficulty bonus on top, with a comment that reads:

AI handicap: bots grow faster

— src/game.js, recomputeBoosts

Setting	Bot, no buildings	Bot with a capital	You, with a capital
Easy	x1.00	x1.25	x1.25
Normal	x1.08	x1.33	x1.25
Hard	x1.30	x1.55	x1.25

A Hard bot is 30% ahead before it builds anything. To match its x1.55 with a capital, you need 0.30 of extra boost from railways and airports. That is 8 railways, or 4 airports and a railway, each standing within range of a city and a factory. Reaching the 1.6 cap costs 5 airports (73,500 gold, because each copy costs more than the last) or 9 railways (120,750). The cost curve is in what to build first.

So the bonus is the one part of Hard you cannot out-play on the map. You have to out-grow it with a capital and boosters, or take its land before the compounding starts.

Easy builds no navy and no nukes

Easy nations still build ports and can ship troops across water, because that code has no difficulty check. What they never build is a trader, a warship or a carrier, and they never build a silo.

On Normal and Hard, a nation builds a silo only after it holds 4% of the land and has no free land beside it. It fires only at the biggest nation it is not allied with, and only when that nation holds more than 1.5 times its land (more than 1.4 times for a doomsday weapon). Launches are at least 250 seconds apart. If you are the runaway leader, you are the target, and one SAM site may not be enough.

Only Hard, the setting that allows three attacks at once, buys a carrier. It wants 16,800 gold in hand first, and whether one is worth the price is the question in warship or carrier.

More opponents means more nations, not just more people

The number of opponents and the difficulty both feed into who spawns. Nations are the aggressors. Settlers expand and defend, and only attack you when your troops are within about 20% of theirs. A settler holding 18% of the land starts no wars at all. Pirates stay small, around 1.2% of the map, and raid from the coast.

Opponents	Easy (nations / pirates / settlers)	Normal	Hard
5	1 / 1 / 3	2 / 1 / 2	3 / 1 / 1
9	2 / 1 / 6	4 / 2 / 3	5 / 1 / 3
14	4 / 2 / 8	6 / 3 / 5	8 / 2 / 4
29	7 / 4 / 18	13 / 6 / 10	17 / 4 / 8

Hard with 29 opponents is 17 aggressive nations. Hard with 5 is 3. The opponent count changes the match about as much as the difficulty does.

An idle player lasts about 6 minutes on every setting

To see the combined effect, we ran the real game engine with a player who never touches anything: Earth map, normal size, 9 opponents, free-for-all, seeds 1 to 24 per setting. After the idle player dies, the test keeps the simulation running so we can watch the AI. The script is scripts/research/frontlines-difficulty.mjs.

Setting	Idle player eliminated within 20 min	Median minute of elimination	Biggest AI nation at minute 5
Easy	15 of 24	6.8	19.6% of land
Normal	17 of 24	6.7	25.5%
Hard	19 of 24	5.3	32.3%

Easy is not a safe place to leave a match running. The gap between settings is smaller than the gap between seeds, because what kills an idle player is mostly who spawns next door. The cleaner signal is the land race: by minute 5, the median largest Hard nation holds 1.6 times as much land as the median largest Easy one.

Missiles are rare in the first 20 minutes. Across 24 Normal and 24 Hard games, the AI launched 6 in total, and none on Easy. Treat this as a bench test, not a statistic about human matches: one map, one idle opponent, 72 games.

OpenFront limits troops, Frontlines adds an economy

We make Frontlines and CONQUEST, so read this as a description, not a ranking. The OpenFront source caps what an AI nation can hold at 50%, 75%, 100% or 125% of a human's maximum troops, depending on Easy, Medium, Hard or Impossible. Its language file labels those tiers Dumb, Average, Smart and Very smart, and says public free-for-all games use Medium.

Frontlines has three tiers. The difference is where the handicap sits: OpenFront limits the AI's army size, Frontlines speeds up the AI's economy. The same question for our other game has its own answer in CONQUEST's four bot levels.

Pick the setting by what you want to practice

A sensible split:

• Easy to learn the economy. One attack at a time, no navy, and 6 of 9 opponents are settlers, so there is room to build.
• Normal for a standard match. It is the menu default, and nukes and a navy appear.
• Hard for online practice. Online bots are Hard, and the 30% bonus is the real test of your build order.

Start a solo match at playfrontlines.com and set Difficulty and Opponents in the menu. The FAQ covers the controls.