OpenFront Tribes: Farm the Neutral Bots Before They Become a Threat

Direct answer: treat tribes as the first land you take, not a rival you fear

Tribes are the Bot player type. They start with exactly 10,000 Troops, cap at one third of a comparable human’s troop limit, refill at roughly half speed, and each one sends only 5% of its force per attack instead of the 20% that humans and Nations send. When you attack a tribe while it is holding land, you also take about 30% fewer losses than you would taking an equivalent human target. Together those numbers make the tribe the single cheapest source of Tiles, Cities, and gold on any standard map, which is why the correct default is to expand through them early and annex the ones that stay weak.

That default has a hard rule behind it. You annex a tribe when the attack wins cleanly and the annexed land improves your position; you keep a tribe as a neighbour when it is currently shielding a border from a stronger human or Nation and re-annexing it now would leave you exposed. You never fight a tribe first on purpose when a Nation is the real prize, and you never treat a tribe you have allied with as a permanent truce. The whole guide is the loop that decides, turn by turn, which of those three moves — annex, keep, or ignore — is right at the next border you touch.

The loop is F-A-M-E: Find, Annex, Manage, Exit. Find the soft pocket and place yourself next to tribes rather than next to several humans. Annex with a ratio you can win using the loss discount and their 5% attack size. Manage the alliance buffer by renewing relations you actually want to keep and letting the ones you do not want to keep expire on purpose. Exit when the pocket is spent, before you have expanded into the most valuable safe spot on the board. Run all four checks before every capture and again after the capture, because every annex adds a new border and changes who the next cheap prize is.

This page owns the player decision around tribes specifically. Threat Assessment scans every rival on the board and tells you which are worth touching at all. Land Combat explains how an attack resolves once you have chosen the tribe to strike. Population Growth explains the troop cap and refill that make a tribe’s cap so much lower than yours. Here the completion test is concrete: you can name the exact ratio at which the next tribe stops being a safe annex, and you know which of the four F-A-M-E checks just changed your answer.

What a tribe actually is in the current build

Start with the identity, because it is where most of the risk comes from. A tribe is not its own player class in the UI; it is the Bot player type that the spawn logic tags with a tribe behaviour script. Nations are a separate PlayerType (Nation) with their own difficulty scaling. So when the minimap shows a plain bot name rather than a Nation banner, that is the entity this guide is about, and its whole stat line is set by a handful of constants you can read rather than guess.

The four load-bearing numbers. First, starting manpower is a flat 10,000 for every tribe regardless of difficulty; Nations scale from 12,500 on Easy up to 31,250 on Impossible, and a human starts at 25,000. A fresh tribe is therefore weaker than a fresh human by 15,000 Troops before anything else matters. Second, maximum troops is the human-style formula divided by three for a tribe, so at the same land area and Cities a tribe’s ceiling is one third of yours. Third, refill rate is the standard curve times 0.5 for a tribe, so it also recovers from losses at about half your pace. Fourth, attack size is the tribe’s current Troops divided by 20 (5%), while a human or Nation divides by 5 (20%). A tribe with 8,000 Troops sends 400 into a land attack; a human with 8,000 sends 1,600.

The discount that turns those numbers into a farming decision is the defender-loss multiplier. When the attacker is a Human or Nation and the defender is a Bot, the engine multiplies the attack magnitude by 0.7. In practice that means taking a tribe costs you roughly 30% fewer Troops than taking a human of the same size and land, all else equal. That is not a small edge at the start of a game, when every lost Troop slows your own cap growth.

One structural quirk keeps the tribe from being a passive object. A tribe is inactive during the spawn phase, so it cannot attack you before you finish placing. After that, on each of its attack ticks it deletes one of its own structures, accepts the alliance requests coming at it, and runs an attack whose timing is randomised per tribe: an attack cadence between 40 and 80 ticks and a trigger gate between 50% and 60% of its force. So the same species of tribe will be visibly more or less aggressive depending on the roll each one drew, and you are reading a distribution, not a fixed timer. The table below is the cleanest way to keep the three actor types straight at a glance.

Property	Tribe (Bot)	Nation	Human
Starting Troops	10,000 flat	12,500–31,250 by difficulty	25,000
Max Troops vs a human at the same size	~1/3	scaled by difficulty	baseline
Refill rate	~half of baseline	near baseline	baseline
Troops sent per land attack	5% of force	20% of force	20% of force
Your losses when you attack them	reduced by 0.7 multiplier	normal	normal

The F-A-M-E loop, run before and after every border change

Find starts on the spawn screen, not after it. Look for a pocket where your likely neighbours are tribes, with at most one human or a single Nation behind another actor rather than a wall of humans. Being boxed in by humans is the most expensive early mistake a tribe-farmer can make, because humans refill at full speed and send full-sized attacks. A pocket next to tribes is the map equivalent of starting with a head start: every early capture through them is discounted and low-risk, and the land you take is land that would otherwise slowly build up into a mid-game obstacle.

Annex uses the two discounts together. You want a tribe whose current Troops are far below your own, because then the 0.7 defender-loss discount and its 5% outgoing attack make the trade almost one-sided. Concretely, if you are at 60,000 and a bordering tribe is at 9,000, its best attack is 450 Troops while you lose a discounted amount to take it; the annex wins and the new Tiles push your own cap up, which raises the ceiling of the next annex. The ratio that matters is your Troops to its Troops, not your land to its land. As a working rule, annex while the tribe is clearly below you and there is no stronger actor on that same border; the moment a Nation or human sits on the other side of that border, the cheap prize is no longer cheap.

Manage is the part most players skip. Tribes accept your alliance requests, and a friendly tribe will not be picked as your target while the relation holds. But the relation has a lifetime, and the tribe only auto-extends it when you ask for the renewal. So an alliance you never touch quietly expires, the tribe reverts to non-friendly, and it can then turn on you at its 5% attack size. You use this on purpose: keep the renewals on the tribes that shield your dangerous border, and let the renewals lapse on the tribes you are about to annex, so they stop buffering you and you take the land on your schedule instead of waiting for them to become irrelevant.

Exit is the stop condition. You exit a pocket once the remaining tribes are no longer safely below you, or once annexing the last one would put your centre of gravity into the most valuable safe spot on the board — the exact position a Nation or a strong human will prefer to take. Exiting means repositioning your expansion, opening a second contact with a different pocket, or turning to the economy so your cap outruns the next threat. Run the same four checks after the annex, because the new border you just created is now the input to the next decision, and the tribe that was the cheapest prize is often no longer in the candidate set at all.

Scenario one: the soft pocket decides where you actually spawn

Assume a single-player FFA with the default 400 bots. On the spawn screen you can see two candidate spots. Spot A is surrounded by four humans and one Nation. Spot B is flanked by five tribes with one human on the far side and a Nation two pockets away. The raw map looks similar, but the F-A-M-E read is opposite. At Spot A the first capture is against a full-speed, full-size human with a 20% outgoing attack and no loss discount, so your early land comes at full cost and the human is already committing to your border. At Spot B the first four captures are discounted 0.7 land attacks against 5% attackers that refill at half speed, so the same number of early captures costs far fewer Troops and builds your cap faster.

The numbers make the choice concrete. Take a representative tribe at 10,000 Troops. Its best land attack is 500 Troops. If you hold 30,000 and attack it, you take a 0.7-discounted loss to a 500 Troop incoming fight, annex the Tiles, and your cap rises because you now own more land. Do that against three of the five tribes before the far human closes the distance and you have converted a pocket of weak bots into real land at a fraction of the cost Spot A would have charged. The trap at Spot B is the one human on the far side: the moment that human has committed to your border, the far tribes are no longer the priority, and you switch from annexing to defending the new edge. The pocket gave you the head start; the human sets the deadline.

A common error is to keep annexing past the deadline because the tribes are still weak. That is the Exit check failing. The correct move is to stop the pocket, open your economy, and let your cap grow so the next contact — which is now the human’s border — is one you win on the ratio instead of by brute force. Reading the pocket correctly at the spawn screen is what turns the tribe from a passive background object into the active first advantage of the game.

The second, subtler error is to spawn at Spot B for the wrong reason — because the tribes look weak rather than because the pocket is shaped so you can clear it before the far human commits. The same five tribes can be a head start or a trap depending on the geometry. If the pocket is open and you can reach the inner tribes before the far human closes the distance, the annexes are cheap and the pocket pays. If the pocket is narrow and the far human reaches your border first, the tribes become a dead weight you are forced to defend instead of annex, and the discounted land never materialises. The Find check is therefore a geometry check as much as a numbers check: the softness of the pocket only pays if the pocket is clearable on your clock, and the spawn screen is the one moment you can test that before committing.

Scenario two: the alliance buffer that is about to expire

You have allied a bordering tribe to keep a Nation from being able to probe one of your edges. The tribe has been friendly for several minutes and is currently at 14,000 Troops while you hold 90,000. The Nation on your other border has been attacking you, and you have just won the exchange and now hold a clear ratio. The question the loop forces you to answer is whether that friendly tribe is still a shield or whether it is now a safe annex you are letting expire.

The Manage check decides it. If the tribe’s border is the only thing standing between you and the Nation’s next probe, you keep the renewal and treat the relation as an active shield; you do not annex it now, because annexing would remove the buffer at the exact moment the Nation is pressing. If, however, the Nation has been redirected and your other borders are quiet, the tribe is no longer shielding anything valuable, and re-annexing it now is a 0.7-discounted capture against a 14,000 Troop force that sends only 700 into an attack — a one-sided trade. You let the renewal lapse, the tribe reverts to non-friendly, and you annex on your own schedule. The difference between the two answers is which border the tribe is covering, and the loop makes you check that instead of assuming the relation still matters.

The failure that this scenario exists to catch is treating the friendly icon as permanent. Because the tribe only auto-extends when you ask, a shield you stop renewing quietly becomes a non-friendly neighbour that can re-target you at 5% of its force. In the first case you were renewing on purpose; in the second you were letting it lapse on purpose. Either way the decision is explicit. The third, accidental case — where you simply forget the relation exists and it lapses while the Nation is still hot — is the one to design out of your own routine by checking the Manage step before every capture and before every renewal window opens.

The same scenario also teaches you the cost of the buffer in the reverse direction. A friendly tribe you keep renewing is, in effect, spending your diplomatic attention to hold back a Nation’s probe. If that Nation redirects and your other borders go quiet, the shield is now a cost you are still paying for a threat that has left, and the correct move is to stop the renewal and annex the tribe while it is still a one-sided trade. The ledger makes this visible: the moment the Nation’s border goes cold, the tribe’s entry flips from “active shield, keep renewing” to “expired-soon shield, annex on schedule,” and you are no longer paying to hold back a threat that is not coming.

Failure modes, and the rival that punishes each one

The first failure is aggressing a Nation before you can delete it. A tribe pocket is cheap, so the temptation is to spend the head start on a Nation that is the real prize. But a Nation sends 20% of its force, its cap scales with difficulty, and you do not get the 0.7 loss discount against it the way you do against a tribe. If you are not already at a clean deleting ratio, the Nation will refill faster than you can bleed it and turn a losing border into a losing game. The loop’s rule: annex the tribes, but only turn on the Nation when the ratio means you can delete it in a few bounded exchanges, not while you are probing.

The second failure is being boxed in by humans instead of sitting near tribes. The F-A-M-E Find check exists entirely because a human on a border refills at full speed and sends full-size attacks, so a pocket that looks soft but is hemmed by humans is the worst early position on the map. The fix is the spawn-screen read: choose the pocket with the most tribe neighbours and the fewest human neighbours, and accept a slightly worse immediate shape if it keeps you out of a human wall.

The third failure is over-valuing a tribe alliance as permanent. The Manage check covers it, but the damage shows up when a forgotten renewal lapses and a tribe that used to be a shield becomes a non-friendly neighbour probing you at 5% of its force while your other border is hot. The fix is to renew only the shields you actually want and to treat an expired relation as a deliberate annex window, not a surprise.

The fourth failure is treating tribe numbers as fixed. Every tribe’s cadence and trigger are drawn from a random range, so two tribes that look identical on the minimap can be on very different timers. The fix is to judge each border by the tribe’s current Troops and the ratio to your own force, not by the tribe’s name or an assumed schedule. Each of these is punished by a specific rival, and each maps to exactly one F-A-M-E check, which is why running the loop before and after every border change is the whole defence against them.

The fifth failure is ignoring the loss discount in the ratio you trust. Because attacking a tribe costs you about 30% fewer Troops than an equivalent human, a ratio that looks marginal against a human can already be a clean win against a tribe. The inverse is the trap: a ratio that looks comfortable against a tribe can still be a losing trade against the human or Nation sitting on the other side of the same border. The ledger keeps the two ratios separate — your discount-adjusted ratio against the tribe, and your raw ratio against the stronger neighbour — so the annex you are making against the weak bot does not quietly become the overextension that the strong neighbour was waiting for.

Adjust the loop for maps, FFA, Team, and Humans versus Nations

The loop changes shape with the bot count and the map geometry. The default single-player lobby spawns 400 bots, and the schema caps the number of bots at 400, so a standard map is a dense field of cheap prizes. On a custom or lower-bot map the pockets are sparser and you have to carry a head start further before the next pocket opens, which makes the Find check stricter and the Exit check earlier. On tall maps the pockets are stacked along the long axis, so a single early sweep can chain several discounted annexes in a row; on wide maps the pockets spread out and you are more often choosing between two pockets rather than sweeping one. Either way the per-tribe numbers do not change, only the density and the shape of the next contact.

In FFA the alliance mechanics work exactly as described, and the Manage check is fully active: you renew the shields and you let the rest lapse. In Team mode the same tribe can be on your side, which changes what “friendly” means and which borders count as your own; the annex decision still runs on the same ratio, but the shield you are protecting is your team’s edge rather than your own, so the Exit check becomes a team-level decision. In a Humans-versus-Nations lobby with no bots there is nothing to farm, and this whole guide collapses to the Nations page; the tribe only exists on a map that actually spawns the Bot type. The practical read is that the bot count is the first knob: the more bots, the more the early game is a farming problem, and the fewer bots, the sooner you are playing the Nations and humans directly.

One more adjustment is the map’s spawn density. On a map where the 400 bots are clustered into a few dense pockets, the Find check is easy but the pockets close fast, so the Exit check fires earlier and you convert the head start into cap growth sooner. On a map where the bots are spread thin across the whole field, the Find check is the hard part and you may have to choose between two pockets instead of sweeping one, which means you carry the alliance management load longer because you are juggling several renewals at once. The density does not change the per-tribe trade, but it changes how many pockets you must run in parallel, and that is the difference between a farming game and a juggling game.

Across every mode the two constants that never move are the 0.7 defender-loss discount and the 5% tribe attack size. Those are what make the pocket cheap and the annex safe, and they are the reason the loop works on a 400-bot map and on a sparse custom map alike. The density and the mode only change how many times you get to run the loop and which rival you turn on after the pocket is spent.

Keep a pocket ledger and hold the version boundary

The whole loop is bookkeeping you do in your head, and the cleanest form of it is a small ledger for every active pocket: the tribe’s current Troops, your Troops, the ratio, whether a human or Nation shares that border, and whether you are renewing the relation or letting it lapse. You update the ratio after every capture because the annex changes both your land and your cap, and you flip a pocket from Find to Exit the moment the ratio or the border stops saying “safe annex.” The ledger is what keeps the four checks honest when the map is busy, because it turns “this tribe looks weak” into “this tribe is at 14,000, I am at 90,000, and no one else is on that border.”

Hold the version boundary as a fact, not a footnote. The current build is the one where tribes are the weakened, fast-expansion variant that was introduced in the v31 line, so the 10,000 flat start, the one-third cap, the half refill, and the 5% attack size are the numbers a player actually faces today. The most recent v34 change to tribes is cosmetic only — it changes how a tribe is coloured on the board, not any of the numbers in the table above. If you are reading an older guide that describes tribes as standard-strength or that shows them in a different colour, that guide is describing a pre-v31 or pre-v34 state, and the decision loop here is the one that matches the current build. The ledger, the four F-A-M-E checks, and the version boundary together are the complete, checkable definition of a tribe game in OpenFront.

That boundary matters because the tribal decision is the one that is most likely to be outdated by a patch. The Nations page and the land-combat page rest on mechanics that change slowly, but the tribe’s whole edge comes from the gap between its numbers and yours, and a single balance change — a higher tribe cap, a full-speed refill, or a larger attack size — would collapse the 0.7-discounted annex into a normal trade and flip the loop’s default from “annex early” to “avoid.” So before you commit a pocket based on this guide, confirm the three numbers in the table are still the ones in the build you are actually playing: the 10,000 start, the one-third cap, and the 5% attack size. If any of them has moved, the Find check is still valid, the Manage check is still valid, and only the Annex ratio has to be re-estimated. That is the whole reason to hold the version boundary explicitly instead of assuming the tribal edge is permanent.