OpenFront on Hard — What Actually Works
What actually works on OpenFront single-player Hard: mountains as a shield, grinding a coast with small landings, and the AI betrayal trigger.
The short version
- Expand into the pale green (plains) and leave the white (mountain) as a shield. Mountains cost the attacker more, so Tribes avoid them too
- Take the coast first, then grind them down with landings. More efficient than a frontal push — and coastline is also what blocks encirclement
- Ally behind you only. It genuinely removes you as a target, but the Nation you allied gets more aggressive
- Never let your troops fall below 20% of cap. That line is the AI's betrayal trigger
Three things mattered most on single-player Hard: using mountains as a shield, grinding a coastline with small landings, and only allying with players behind you. Everything else only starts to matter once those three hold.
Earlier pages cover “here is the formula”. This one covers “so how do you play the game”. Because of that, verified facts and playing hunches are marked differently.
🟢 confirmed in the game’s source. 🔵 an observation from play. Observations can be wrong; when one gets confirmed or refuted, this page changes.
1. Use the white land as a shield
🟢 Terrain changes losses and speed at the same time.
| Terrain | Loss coefficient | Ticks per tile |
|---|---|---|
| Pale green (plains) | 80 | 16.5 |
| Mid tone (highland) | 100 | 20 |
| White (mountain) | 120 | 25 |
Expand across the green. The same troops buy far more land. And for the same reason, a country that owns green land is the easiest one to conquer.
🔵 There is a trap in that. If all you own is green, you get carved up the moment somebody attacks — they get the same coefficient you did.
When your military is roughly even with the neighbours, the white land earns its keep. Taking it is a bad deal, so Tribes rarely bother, and when they do they crawl. Put a defense post on it and attacker losses multiply by ×5 — whatever walks in gets ground down, and your counter-attack goes out from there.
Leave countries that hold a lot of white land alone. They take forever to eat and pay little. For the same troops, hitting a green-land country is dramatically better.
→ Combat Math · Defense Post Range and Placement
2. Lean on amphibious landings
Expand by eating ordinary Tribes for a while and you look up to find Nations walling you in on every side. Fight through the front and it drags, and while it drags another Nation outgrows you and eats you.
🔵 So secure coastline as early as you can, and sprinkle troops along it at a low ratio.
- Drop the slider to 1–5% and throw at several points of coast
- Adjust with
T/Y, not the mouse. Far faster - 🟢 You get three boats at a time, so there is a limit to the sprinkling
Grinding a coastline with small landings beats pushing the front, because the defender’s troops end up split instead of massed on one border line.
🔵 And territory taken by landing seems to hold better than it should.
We went looking for this in the source and could not find a basis for it. The AI target-selection code (
AiAttackBehavior) has no concept of landed or isolated territory at all. It picks from adjacent powers, skipping Nations and humans 50% of the time.So for now, even if the observation is real the cause is probably something else — the landing site sits at the far edge of their territory so few troops arrive, or that AI is already committed on another front. This paragraph gets updated if we confirm it.
⚠️ Boats do not ignore defense posts. Land within 30 tiles of one and you eat the same ×5 as you would on foot. Pick coast without the checkerboard pattern.
→ Defense Post Range and Placement
3. Ally early — but only behind you
On Hard, an ally will knife you the moment your troops look thin. 🟢 This is confirmed in the source. A Nation’s betrayal check reads:
their troops + troops out on attack < 20% of their cap
and their troops < my troops → betray
So the trigger is the moment you drop below 20% of your cap. The check does not exist at all on Easy or Medium — only Hard and Impossible run it.
🟢 Keeping your troops up is diplomacy. Stay above the 20% line and the betrayal condition never fires. There is a separate clause for traitors — if you are a traitor and under 1.2× their troops, they betray anyway.
There is still a reason to ally. 🟢 You genuinely drop out of their target selection.
shouldAttack(other) {
// wilderness · non-human · traitor → always attack
if (!other.isPlayer() || other.type() !== Human || other.isTraitor()) return true;
...
}
Non-human opponents (Nations, Tribes) and traitors are attacked without exception. Only humans get difficulty-based protection, and only on Easy and Medium. From Hard up, humans get none either.
That does not mean allying with everyone.
- 🔵 You miss the window to profit while Tribes fight each other
- 🟢 An allied Nation gets more aggressive. Also confirmed
On Hard and Impossible, a Nation holds back “75% of the strongest neighbour’s troops” (90% on Impossible). But allies and Tribes are not counted as threats in that neighbour list.
if (n.isPlayer() && !this.player.isFriendly(n) && n.type() !== Bot) { ... }
Allying lowers that Nation’s reserve requirement, so it commits more elsewhere. The common observation that “an allied bot gets cocky” is exactly this clause.
And betraying costs you: 🟢 a 30-second mark during which everyone hits you at half losses. Worse, the AI treats traitors as unconditional targets and picks the weakest bordering traitor.
So you ally only on the boring side — behind you. Leave the front open and lock the back.
🔵 A refused alliance request in the mid-to-late game is a warning. It means they are underrating your strength and war is likely. Put a defense post on that border immediately.
🟢 Looking strong is defending
Nations on Hard and Impossible also run this check:
my troops < 20% of the target's troops → "attack too weak", stand down
Keeping your troops above 20% of theirs takes you off the candidate list. Team games and Tribes are exempt, and a Nation already under attack fights back regardless.
Combine that with the betrayal condition above and it all points one way — do not run your troops dry. Diplomacy and defense hang on the same number.
→ What Happens When You Betray an Alliance?
4. Encirclement costs zero troops
🟢 Surround a power completely and it is absorbed with no troop losses at all, banked gold included.
Early on, when eating one more Tribe quickly matters, that difference is large.
🟢 But it does not trigger if they hold coastline. Touching water or the map edge breaks it.
This is why landings matter. Cutting off their coast is what opens the encirclement. Sections 2 and 4 are one tactic.
5. Shorter notes
Every map has a good seat. Knowing it solves most of the game. Continental maps are harder because Tribes expand and snowball fast; island maps are easier.
🟢 A bigger empire attacks better. This is not widely known.
| My tiles (gauge from leaderboard %) | Loss multiplier | Advance budget |
|---|---|---|
| 100,000 | ×1.00 | ×1.00 |
| 200,000 | ×0.79 | ×0.66 |
| 500,000 | ×0.57 | ×0.38 |
| 1,000,000 | ×0.45 | ×0.25 |
Past 100,000 tiles you lose less and move faster. Add the bonus for hitting large defenders and it snowballs the bigger you get. The flip side: poking a large empire while you are small is exactly as bad as it feels.
🟢 Losing a building makes the next one cheaper. The price key is min(current holdings, total ever built), not the running total — so when buildings change hands the price drops with them.
🔵 In theory you could build near the border, let a Tribe take it, then push back and keep buying cheap buildings. Whether that actually nets out is unverified — the territory you concede may cost more than you save.
🔵 Let somebody else take the bombed ground. Fallout tiles carry an attack loss multiplier. More efficient to let another power swallow them and then eat that power.
🔵 At cap, ship troops out and recall them. Growth nearly stops at cap; loading troops onto boats drops your home count and growth resumes. Recall afterwards and 🟢 you pay 25% to get 75% back.
This only pays if what you regrow beats the 25% retreat cost. At 90% of cap you are earning 30% of peak, so there is lots of room; near 50% you are already at 98% of peak and there is no reason to bother. Check where you sit with the table in Reading the Bottom Bar.
Further reading
- Named OpenFront Strategies — seven standards with evidence
- Game Modes and Options — what difficulty actually changes
- What Attack Ratio Should You Use? — where the slider values come from
- Ten Common Mistakes
- Troop Cap Calculator — how much you should be holding right now