How to Beat the OpenFront AI: Bots vs Nations and What Actually Changes the Outcome

The short answer: OpenFront’s singleplayer AI is not one opponent but two. Bots are the smaller, slower-growing, cheap-to-absorb opponents that fill a map’s population; Nations are the larger, stronger, land-owning opponents that fight on near-human terms at High difficulty and above you on Impossible. Beat Bots by out-earning and out-annexing them early and ignoring their weak attacks; beat Nations by treating each one as a rival you manage — pressure it, split its border, and convert the land it cannot defend. Pick your map’s Bot and Nation counts and difficulty to match the kind of match you want, and the rest is ordinary play against a known, measured opponent.

What the AI actually is: two opponent types, not one

The single most common mistake is treating “the AI” as a single thing with a single strength, because the game’s interface labels everything you do not control as computer opponents. In the engine the opposition is split into two distinct player types, and they do not behave alike. The PlayerType enum has exactly three members — Bot, Human, and Nation — and the two non-human members are what you fight in singleplayer. There is no separate “Tribe” type: the small tribes that populate a fresh map are simply Bots, the kind that have no client attached to them and so are fully computer-run. That distinction matters because the two types differ in starting strength, in the ceiling they can ever reach, in how fast they refill after a fight, and in how much of their army they actually commit when they attack you.

A Bot is the low-end opponent. It starts with a modest army, its maximum troop ceiling is a fixed fraction of what a human can hold, and it grows troops at a reduced rate. Many of them are so weak that a single well-timed attack takes them out and absorbs their land with almost no cost to you. They are, in aggregate, the population of a map: the background pressure that is constant and cheap to manage if you stay ahead of it. A Nation is a different animal entirely. A Nation starts with substantially more troops than a Bot, its ceiling scales with the difficulty you selected, and it grows at roughly the human rate. Since the v0.35.0 “Smarter nations” update, Nations are also more aggressive and more coordinated, meaning they will actively press your border and contest the same land rather than sitting still. The practical consequence is that you cannot apply one strategy to both: the move that wins against a hundred Bots — keep expanding and let your economy outgrow them — loses badly against a single Nation that is allowed to match your army size and keep refilling. If you want the deeper mechanics behind each type, the companion Tribe and Bot guide covers the small-end opponents and the Nation AI behavior in v0.35 covers the coordinated end.

Read that split as a warning about scope, not trivia. “Beating the AI” is not one problem but two different problems glued together by a single label, and a large share of early singleplayer losses come from applying the wrong one of the two playbooks to a map that actually needs the other. A map stacked with Bots punishes hesitation and slow building, while a map carrying real Nations punishes the reckless, economy-first expansion you needed to win the first kind. So the single most valuable habit to build is to classify every border you look at before you spend a single troop on it, because the identity of the neighbor determines which of the two playbooks you are actually playing.

The verified numbers: the Bots-versus-Nations table

This is the reference data, taken from the tagged v0.35.9 source rather than the rounded community figures that have drifted. Use it to check any wiki or video you read against, because several widely circulated numbers are no longer what the engine actually loads. The table below is what a v0.35.9 singleplayer game actually fields for the two AI types at each difficulty.

Property	Bot (every difficulty)	Nation — Easy	Nation — Medium	Nation — Hard	Nation — Impossible	Human (you)
Starting troops	10,000	12,500	18,750	25,000	31,250	25,000
Troop ceiling multiplier	x1/3 of base	x0.5	x0.75	x1.0	x1.25	x1.0
Growth multiplier	x0.5	x0.9	x0.95	x1.0	x1.05	x1.0
Army committed per attack	troops / 20	troops / 5	troops / 5	troops / 5	troops / 5	troops / 5
Base gold income	50	100	100	100	100	100

Read this table in two passes. The first pass is identity: a Bot is a small, low-ceiling, slow-growing, low-income opponent that attacks with only a twentieth of its force — which is why a Bot’s attack is usually a rounding error you can shrug off and why absorbing a Bot costs you almost nothing. The second pass is the Nation row, which is where your real decisions live. Notice that on Hard, a Nation fights on exactly human terms: the same 25,000 start, the same x1.0 ceiling, the same x1.0 growth. That is the level at which a Nation is a true rival and you must treat it like the strongest human opponent you have ever played. On Impossible, the Nation starts 25% above you and holds a 25% higher ceiling with slightly faster refill, so the gap compounds the longer the match runs. And on Easy and Medium, the Nation is deliberately softened to sit below you in every one of those columns, which is why it is a manageable nuisance rather than a match-deciding rival. The Bot column is constant across all four difficulties — the difficulty selector never touches Bots — so you can plan around them without re-reading the table for each level. The full per-difficulty breakdown of the Nation side is the subject of the Nation difficulty guide; this page’s job is to use those numbers to decide how to actually win.

The cleanest way to internalize the table is to notice which columns move and which stay still as you turn the difficulty. The entire Bot column never changes, so a Bot on Impossible is numerically identical to a Bot on Easy and you never re-plan around it. Only the Nation columns move, and they move together in one direction: from below you on Easy, to equal to you on Hard, to above you on Impossible. That single one-directional slide is the whole shape of “what gets harder as you pick a harder setting,” and it is the exact shape a winning strategy has to respect, because a plan that works against softened Nations does not survive the jump to the difficulty where they match you.

How the engine behaves each tick: the mechanics that make the difference

The numbers above only make sense once you understand the two systems that turn them into behavior: troop growth and attack commitment. Both are recomputed every game tick, and both are where the gap between a Bot and a Nation becomes real on the board. The base growth formula is identical for every player — the troops you add per tick is a function of your current troop count relative to your ceiling, rising steeply when you are far below the cap and flattening out as you approach it. That shape is why a player who has just lost most of an army refills quickly and a player already at capacity refills almost not at all. What the engine then multiplies that base rate by is what separates the AI types. A Bot’s growth is cut in half, so a Bot that is knocked down comes back slowly and a Bot near its ceiling is effectively frozen — which is exactly why the strategy of knocking Bots down and re-attacking before they recover works so well. A Nation’s growth is scaled by difficulty, so a Hard Nation refills at the same rate as you and an Impossible Nation refills a little faster, which is why a Hard or Impossible Nation that you trade with can come back to its previous size while a weakened Bot cannot.

Attack commitment is the second system, and it is the one players least expect. When a player attacks, the engine does not send their whole army; it commits a fixed fraction of their current troops. For a Bot that fraction is one twentieth; for a Nation and for you it is one fifth. So a Bot with 10,000 troops sends 500 into a given attack, while a Nation or a human with the same 10,000 sends 2,000. That fourfold difference in committed force is the mechanical reason a Bot attack feels like a nuisance and a Nation attack feels like a threat: the Bot is sending a token force, the Nation is sending a real strike. The third system is gold income. Every player earns gold at a base rate that then feeds structure and unit costs, and a Bot earns at half the base rate of a Nation or a human. That means a Bot’s economy trails yours and a Nation’s in the long run, so even before you ever touch a Bot, it is falling behind your structure output — and a Nation, earning at your rate, is the one whose structure and unit economy you genuinely have to out-build. All three systems compound with the land and Cities you accumulate, which is why the early decisions about which neighbors to expand into and which to ignore are the ones that decide the match.

Choosing your setup: bots, nations, and difficulty

Before the first tile is picked you are already deciding which of the two AI types will dominate the match, because the singleplayer setup lets you set how many Bots and how many Nations field the game independently of each other, and the difficulty selector then retunes the Nations. The most important realization is that these are two separate dials and most players only turn one. A map loaded with a large Bot count and zero Nations is a pure expansion-and-annexation match: your only job is to out-run the small opponents and convert their land, and the match is decided by your expansion speed and how quickly you absorb neighbors. A map loaded with fewer Bots and a real Nation count is a rival-match: your job is to manage the Nations, and the match is decided by how you split and pressure their borders. The difficulty level is a third, Nation-only dial: it never changes your account or the Bots, it only retunes the Nations’ starting troops, ceiling, and growth.

So the decision is three questions. First, what kind of match do you want to practice — fast expansion against the small fish, or a sustained war against strong rivals? That picks your Bot and Nation counts. Second, how strong do you want the Nations — softened below you on Easy and Medium, equal to you on Hard, or amplified above you on Impossible? That picks the difficulty, and it is a decision the Nation difficulty guide breaks down in full. Third, which map — because the map’s manifest decides how many Nations it can even place and therefore which difficulty options are actually available to you; a map with no Nations disables the difficulty control entirely. For a first serious attempt at understanding the AI, a medium map with a real Nation count on Hard is the cleanest choice: you face a Nation that is exactly as strong as you, so you learn what a true rival feels like without the compounding handicap of Impossible, while the surrounding Bots give you the expansion practice on the side. If your specific goal is to win the top-level campaign and medal, the Impossible singleplayer guide covers the extra constraints that mode adds.

Three of the four levers are independent of your skill, which is what makes this pre-game decision so powerful: you are not changing how good you are at the game, you are changing which kind of game you play. A player who is comfortable with steady expansion but not with multi-front war management should load Bot-heavy, low-Nation maps and win the game that suits their strengths, while a player who wants to sharpen their war management should load real-Nation, Hard maps and deliberately take on the harder game. The difficulty and counts are not a fixed ladder you must climb in order; they are practice dials, and choosing them to match your goal is the quiet advantage over a player who just hits start and finds out mid-game which kind of match they are actually in.

Scenario A: an early Bot swarm on a high-Bot map

Picture the concrete situation. You have chosen a map with a large Bot count and zero Nations, so the entire opposition is the small end. Assume you spawn on a defensible position with a first border that is mostly Bots, and the goal is simply to convert as much of that population’s land as possible before any neighbor can organize. The numbers set the opening: every Bot starts with 10,000 troops, and because they commit only a twentieth of their force, a Bot’s first strike on you is a few hundred troops — an attack you can absorb without even pausing your build. Meanwhile your own army starts at 25,000 and grows at the full human rate, while each adjacent Bot grows at only half the rate and is capped at a third of your ceiling. That means if you simply hold your border and keep expanding, the gap between your army and any individual Bot widens every tick.

The winning move is a focused early push, not a slow build. Pick your weakest adjacent Bot — the one with the smallest army and the thinnest border — and send a decisive attack against it while your own army is at or near its post-growth peak. Because you commit a fifth of your force and the Bot commits a twentieth, even a modest fraction of your army is multiple times the strike it can answer, and a clean win annexes its land and raises your own ceiling. Do this on a tight rhythm: strike, annex, let your army refill to near its cap, strike the next weakest neighbor. The key discipline is to never let two Bots gang up on the same border at once — a single Bot is trivial, but two Bots attacking the same exposed edge can stall you while you recover. Hold a reserve, keep one edge behind your strongest border, and always strike from the side where your committed force is clearly larger than the sum the Bot can answer. Worked example: you hold 25,000 and commit 5,000 to a Bot that holds 10,000 and commits 500; you win the exchange, annex its roughly ten tiles, and your ceiling rises by the city and land you gained, while the Bots you have not yet touched are still refilling at half speed. That is the entire early game on a Bot-heavy map: stay one step ahead, absorb the weak strikes, and convert the population’s land into your own ceiling.

Scenario B: a late-game Nation that outgrows your border

Now the rival match. You are several hours into a game on a map with a real Nation count, and one Nation has become your dominant border — it has out-expanded the Bots around it, accumulated a string of Cities, and now holds a troop count and ceiling that at Hard matches yours and at Impossible exceeds it. Assume the specific situation: you hold roughly 600,000 troops at a ceiling near your current land, this Nation holds a comparable or larger army, and its growth multiplier is at or above yours. This is not an opponent you can simply outrun the way you outran a Bot, because it refills at your rate and commits a full fifth of its force to every strike. A straightforward head-on trade will end even at best, and on Impossible it will slowly bleed you, because the Nation’s 1.05 growth and higher ceiling let it recover every exchange a little faster than you can.

The winning approach is to change what the Nation is fighting rather than to out-trade it. The first tool is to force the Nation to split: if you keep a credible threat on two of its borders at once, its engine has to divide its committed force, and a split Nation commits a smaller strike to each front. The second tool is to make it pay for its own expansion: a Nation that has stretched its border thin across a long line of Cities has a thin defense behind each one, so a concentrated strike at its most exposed, least-supported City can capture it and cut the land behind, which is the same logic the nation-strike protocol and beachhead doctrine describe. The third tool is to let other strong neighbors spend on it: if another rival is contesting the same Nation, you can pressure the edge that is already weakened while the Nation is committed elsewhere. The discipline that decides this is to never trade your full army against its full army at a border where it is at rest and fully refilled — you want to hit while it is split, while it is attacking elsewhere, or at a City it cannot quickly back. Worked example: you commit 12,000 (a fifth of your 60,000) at a Nation’s exposed City that, because it is splitting to hold two borders, can only commit 6,000 in defense there; you capture the City, annex the land behind it, and its ceiling drops while yours rises, inverting the gap that was slowly favoring it.

Failure modes and the counter that works for each

The way players lose to the AI is not a single failure but a small set of recurring ones, each tied to confusing the two opponent types. The first failure is treating a Nation like a Bot: you expand straight into it the way you would push a weak neighbor, assume the trade will be as easy as the Bots were, and get answered by a strike that commits a full fifth of a comparable army. The counter is the identity check before every expansion — is this neighbor a Bot (small, low ceiling, slow growth) or a Nation (large, at-or-above-you ceiling, human-rate growth)? If it is a Nation, do not expand into it until you have split it, weakened it, or confirmed your committed force beats what it can answer on that specific border. The second failure is the opposite: treating a Bot like a Nation and stopping to “fight a fair fight” against a neighbor you could simply out-earn and annex. The counter is the rhythm rule — a Bot is absorbed, not negotiated with; keep the strike-annex-refill loop moving and let your economy outgrow the population rather than trading equal armies with it.

The third failure is the single-front trap in the rival match: you concentrate the entire match on one Nation and ignore the others, so the Nation you are focused on spends freely while a second Nation expands unopposed into the land you never defended, and you lose the map to the neighbor you did not play. The counter is the split-and-watch rule — always leave a credible second threat, or at least a defended second border, so no rival can expand into unwatched land for free. The fourth failure is the capacity trap: you push too hard against a Border, overcommit, and end the exchange below your own ceiling while the opponent, still near its ceiling, refills faster than you because the growth curve is steeper at low troop counts. The counter is to strike from near your own cap and to avoid the moment where you are at half strength trading into a fresh enemy. The outgunned defense guide covers the specific posture to take once you find yourself in that below-cap position. The fifth failure is the assumption failure: players assume a flat “250% income cheat” on the harder levels and build an unwinnable plan around it, when the real advantage is a modest, compounding edge in start, ceiling, and refill. The counter is to use the table from this page instead of the folklore, because a modest compounding gap is exactly the kind of gap that a structural edge — more Cities, a walled region, a rival busy fighting you on a second front — can overcome. Every failure here is a misread of the two types; get the identity right and the correct counter is usually the obvious one.

Mode and map adjustments: when to change the plan

The plan above is the default, but both the match mode and the map shape change which of the two AI-type strategies is on the table. In a free-for-all singleplayer the full logic of this page applies: every computer opponent is either a Bot you annex or a Nation you manage, and you are free to pick and choose. In team-based singleplayer the picture changes because some of the Nations and Bots are on your side of the line, so the “two borders you can split” tool is partly unavailable — you cannot freely strike a teammate’s neighbor — and the match shifts toward coordinating your expansion against the opposing Nations as a unit. In a no-alliances or otherwise restricted mode the Nation-management tools are the only tools, so you lean harder on split-and-press and on letting rivals spend on each other. The map shape matters just as much. A map where your spawn is ringed by Bots is a Bot-swarm match and the expansion-rhythm plan is the default; a map where your spawn is flush against a large Nation from the first minute is a rival match from minute one and you should plan the split-and-press posture immediately rather than learning it mid-game. A map with a lot of open water and Islands changes the conversion math, because capturing a Nation’s coastal City across a strait is a different commitment than pushing overland, and the island defense guide covers the landing side of that.

The final adjustment is the one players make least: matching the map’s actual Nation count to your goal. If you want to practice beating Bots, load a map with many Bots and few Nations and play the rhythm game; if you want to practice beating strong Nations, load a map with a real Nation count and set the difficulty to Hard so each Nation is a true human-strength rival; if you want the full test, use Impossible and accept that the Nations start above you and refill a little faster, and plan the whole match around not letting that gap compound. The decision framework that ties it together is three questions asked before you move: what type of opponent is this (Bot or Nation), what is its measured strength on this difficulty (from the table), and what is the cheapest way to change the ratio in your favor on this specific border (annex if Bot, split or weaken if Nation). Answer those three and the correct next move is almost always the obvious one, because the engine is fully measured and there is no hidden strength to fear — only the two types, the numbers in the table, and the discipline to apply the right counter to each. That is how you beat the AI: not by being stronger on average, but by always knowing which of the two opponents is in front of you and playing the move that the verified numbers say will win.