Diplomacy, Embargoes, and Betrayal

Direct Answer

In v0.33.7, diplomacy is a timer and exit-cost problem. A request lasts 20 seconds, the same player can be asked again after 30 seconds, and the default alliance lasts 300 seconds.

Extensions need both players. Natural expiry is not betrayal. Manual breaking can create a 30-second traitor state, while embargoes and nuclear blast checks can alter trade or end an alliance before your planned attack begins.

Version Boundary and Terms

Verified for the official v0.33.7 release on 2026-08-22. This guide explains rules and decision costs, not a guaranteed best betrayal time.

An ally is a player in an active alliance. A teammate belongs to the same fixed team. Both may be friendly, but the game does not treat team membership as a timed alliance.

That distinction matters. An alliance can expire, be extended, or be broken. A team cannot be dissolved with the normal break-alliance action.

The default alliance duration is five minutes. A custom lobby can choose from one to 15 minutes, while a duration of zero disables alliances.

Legacy alliance-disable settings also block requests. Check the lobby rules before building a plan around the default 300-second window.

The mechanics below come from the official v0.33.7 release, fixed-tag source, and official tests. Tactical recommendations are labeled as decisions rather than engine guarantees.

The Four-Question Diplomacy Check

Before requesting, extending, embargoing, or breaking, answer four questions.

1. Time: How long remains on the request, alliance, extension window, temporary embargo, or traitor state?
2. Exit cost: Can you wait for natural expiry, or does an immediate break justify the traitor exposure?
3. Network: Which Port or Train relationships stop working if either side applies an embargo?
4. Fire: Will an in-flight or planned nuke violate the alliance before the ground operation starts?

This check does not choose for you. It exposes the state changes that are easy to overlook when a border suddenly becomes urgent.

Write down the next action the alliance enables. A buffer without a follow-up is still useful, but it should not consume attention that another front needs.

Alliance Requests: 20 Seconds Inside a 30-Second Cycle

An alliance request remains active for 20 seconds. The cooldown for asking the same player again is 30 seconds from the original request creation time.

The cooldown does not begin when the request expires. If a request starts at t = 0, it disappears at about t = 20s, and the same target becomes requestable at about t = 30s.

That leaves roughly ten seconds between expiry and a legal retry. Treat this as a small observation window, not as a second full 30-second wait.

If another player already sent you an incoming request, sending a counter-request accepts that pending alliance. The engine does not keep two independent requests between the same pair.

Scenario 1: a missed request

You request an alliance at 12:00. The offer remains available until about 12:20. A retry to that player is legal around 12:30.

If the border cannot survive those ten extra seconds, the request is not immediate defense. You still need enough Troops, terrain, or another exit to cover the gap.

If the other player sends a request during your wait, your counter-request can accept it. That changes the next action from “wait for cooldown” to “verify what the new friendly border enables.”

Sources: AllianceRequestExecution.ts, AllianceRequestImpl.ts, and official request tests.

Acceptance Changes the Board Immediately

An active alliance makes both players friendly for ordinary land attacks. A current attack between the new allies is recalled without the normal retreat malus.

Acceptance also clears temporary embargoes between the two players. It does not erase manual embargoes.

The game checks relevant in-flight nukes when the alliance is accepted. Nukes between the new allies that would break the alliance are removed.

That protection is specific. It is not a global ceasefire, and it does not remove third-party weapons.

Before accepting, ask whether the agreement preserves a valuable buffer, trade route, or recovery window. Then check whether an in-flight strike will be deleted as part of acceptance.

Do not accept merely to erase a missile unless the resulting friendly border also helps your position. The code proves the removal rule, not that acceptance is strategically profitable.

Extension Needs Two Decisions

Alliance extension requires both players to agree. One player cannot unilaterally turn a temporary agreement into a longer one.

The client exposes the extension window during the final 30 seconds. Completing the mutual extension resets a full alliance duration from the current tick.

It does not add a small 30-second patch to the old end time. Under the default setting, a successful extension creates another 300-second duration from renewal.

Scenario 2: renewing the default alliance

Assume a default alliance would expire at 15:00. Its extension window opens around 14:30.

Both players approve at 14:42. The new expiry is approximately 19:42, because a fresh 300 seconds starts at the renewal tick.

If only one player approves, no extension is completed. The original alliance can still expire naturally at 15:00.

Use the final 30 seconds to decide whether the buffer and network remain useful. Do not mistake the visible extension control for automatic consent.

Sources: AllianceExtensionExecution.ts, AllianceImpl.ts, and official extension tests.

Natural Expiry Is Not Betrayal

When the alliance duration ends normally, the relation ceases to be an alliance. Natural expiry does not apply the traitor status.

This gives you a low-penalty exit when immediate hostile action is unnecessary. You can let the timer finish, preserve defensive readiness, and reconsider the border after the state changes.

Natural expiry is not permission to ignore the border. The former ally becomes attackable again, and any trade or positioning assumption based on friendly status must be reviewed.

The important distinction is mechanical: expiry ends time; manual breaking is an action with a separate betrayal path.

Scenario 3: wait or break

Thirty-five seconds remain on an alliance. You do not need to attack immediately, and the other front can hold.

Waiting avoids the manual-break traitor path. Breaking now may expose you for 30 seconds, only five seconds less than the remaining alliance time.

That comparison does not prove waiting is always correct. It shows why the remaining timer belongs in the decision before you pay an avoidable exit cost.

Embargoes Are Separate from Alliances

Trade permission is blocked if either player embargoes the other. An alliance is not required for trade, and an alliance does not override an active manual embargo.

A normal attack causes the target to create a temporary embargo against the attacker. That temporary state expires after roughly five minutes.

A manual embargo is persistent. It does not use the same automatic expiry path.

When two players accept an alliance, temporary embargoes between them are cleared. Manual embargoes remain, so the new allies may still be unable to trade.

This is why “we allied, therefore trade is fixed” is unsafe. Check the actual trade connection after acceptance.

Embargo All excludes the player using it, Bots, and teammates. It has a ten-second cooldown.

The action can cut several external relationships quickly, but its exclusions and cooldown still matter. It is not a universal permanent lock on every player.

Sources: EmbargoExecution.ts, EmbargoAllExecution.ts, PlayerImpl.ts, and GameImpl.ts.

Scenario 4: temporary versus manual embargo

Player A attacks Player B at t = 0. B’s temporary embargo against A is expected to end around t = 300s if nothing else changes.

At t = 120s, both accept an alliance. The temporary embargo is cleared immediately.

If B had also placed a manual embargo, that persistent state would remain. Trade can stay blocked even though the pair is now allied.

The practical response is to inspect the route, not to infer trade status from the alliance icon alone.

Manual Breaking and the Traitor Window

Manual breaking follows a different path from expiry. If the betrayed player is connected and not already a traitor, the breaker receives traitor status for 30 seconds.

The betrayed recipient records a -100 relation change toward the breaker. Nearby players who are not on the recipient’s team record -40.

Relation values are engine state. They do not guarantee that a human opponent attacks, forgives, or changes plans.

The visible tactical risk comes from combat modifiers. When the defender is a traitor, the attacker’s loss multiplier is 0.5.

Tile processing cost is multiplied by 0.8 against that defender. This means lower processing cost under the same conditions, not a guaranteed completion time.

Path shape, available Troops, counterattacks, and tick order still affect the real advance. Do not translate 0.8 into a promised 20% faster conquest.

Sources: BreakAllianceExecution.ts, execution combat utility, and AttackExecution.ts.

Scenario 5: the 30-second exposure

You manually break at t = 0 while the other player is connected and eligible for the betrayal path. Your traitor state runs for 30 seconds.

An attacker who fights you during that state receives the 0.5 attacker-loss multiplier and 0.8 tile-processing-cost multiplier.

At t = 30s, the traitor timer ends. Damage, lost territory, or changed relations created during the window do not rewind.

The decision is therefore not “Can I win after breaking?” It is “Does immediate action justify a known 30-second vulnerability and the relation changes?”

Nuclear Strikes Can Break an Alliance First

Nuclear alliance breaking uses two independent triggers. A strike can cross a weighted allied-territory threshold, or it can threaten an allied structure.

The configured threshold is 100. Allied tiles inside the inner blast radius count as 1; allied tiles in the outer ring count as 0.5.

The weighted total must be strictly greater than 100. A total equal to 100 does not trigger the tile-count branch.

The structure branch is separate. If an allied structure would be destroyed inside the outer blast radius, the alliance breaks regardless of the weighted tile total.

When a launch is judged to break the alliance, the execution rejects relevant pending requests, breaks the active alliance, and follows the normal break logic for traitor handling.

Sources: Config.ts, NukeExecution.ts, and official nuke tests.

Scenario 6: strict territory threshold

An explosion covers 101 allied tiles in its inner area. The weighted count is 101, so the territory branch breaks the alliance.

Exactly 100 allied inner tiles produce 100. That does not pass the strict > 100 comparison.

Two hundred allied outer-ring tiles also produce 100 because each counts as 0.5. Tile count alone still does not trigger.

These examples isolate the weighting rule. A structure in the blast can independently change the result.

Scenario 7: three tiles and one Port

Suppose only three allied tiles are inside the outer blast radius. Their weighted total is far below 100.

One of those tiles contains an allied Port that the blast would destroy. The structure branch triggers, so the alliance breaks.

Do not use a territory counter alone when aiming near Cities, Ports, Factories, SAM Launchers, or other allied structures.

The MIRVWarhead Boundary

The v0.33.7 NukeExecution path skips this alliance-break check for the MIRVWarhead unit type.

That is a local code boundary. It supports a narrow statement: this warhead execution path does not trigger the check at that location.

It does not prove that the MIRV carrier, every stage of a MIRV launch, or every resulting interaction can never break an alliance.

Keep the public wording narrow. “MIRVWarhead skips this check” is supported; “MIRV never breaks alliances” is not.

The same caution applies to teams. Team membership is not an alliance and cannot be removed by the break-alliance action.

That does not prove teammates are immune to every nuclear damage path. This guide makes no teammate nuclear-immunity claim.

Choosing Between Expiry, Extension, and Break

Choose natural expiry when you do not need immediate hostility and want to avoid the manual traitor path.

Choose extension when both sides still gain from the buffer, trade access, recovery time, or a shared threat. Confirm both approvals before relying on the new end time.

Consider manual break only when acting now is worth the 30-second traitor exposure, relation changes, and loss of a friendly border.

Consider embargo when denying trade has value by itself. It need not be paired with an attack, and a manual embargo can outlast later diplomatic changes.

Recheck nuclear plans before every diplomatic transition. Acceptance can remove qualifying in-flight nukes, while a planned blast can break the alliance before impact.

These choices are not a ranking. Map topology, remaining Troops, other neighbors, and the value of connected infrastructure determine which cost matters most.

Counterplay and Failure Modes

The ally is buying time

A request may create a buffer, but it also gives the other player time to recover or reposition. Keep enough defense to survive rejection, expiry, or a later break.

Do not empty the shared border merely because ordinary land attacks are blocked during the active alliance. The state has a timer.

The extension is not mutual

One approval does not complete renewal. If the other player does nothing, plan for the original expiry.

Use the final 30 seconds to shorten exposed fronts, protect infrastructure, and stop creating routes that only work while friendly status remains.

Trade stays blocked after acceptance

The usual cause is a manual embargo. Alliance acceptance clears temporary embargoes only.

Verify the actual Port or Train relationship. Do not spend on a route until both directional embargo states allow trade.

The traitor modifier attracts attackers

Other players can exploit the 30-second traitor window. The lower attacker-loss and tile-processing multipliers make you a cheaper defender to attack during that state.

Before breaking, check every border, not only the former ally. Nearby players also receive the relation event unless they are on the betrayed player’s team.

The nuke calculation ignores structures

A weighted total of 100 looks safe only for the tile branch. One allied structure in the outer blast radius is enough to take the other branch.

Inspect structure positions separately from territory coverage. The two checks are joined by OR, not by AND.

Human behavior is treated like a formula

The relation penalties are game state, not a prediction model for human decisions. A person may retaliate, ignore you, or cooperate later.

Use the numbers to explain the engine consequence. Use map and match context to discuss possible responses without promising one.

Mode and Map Adjustments

In FFA, timed alliances can reshape several open borders. The 30-second traitor window may matter more when multiple nearby players can reach you.

In team modes, distinguish ally from teammate every time. A teammate is friendly through team membership, not through the temporary alliance lifecycle described here.

On compact maps, a nuclear outer radius can include more allied structures and territory. Inspect the OR check carefully before launching near a dense shared border.

On large maps, a five-minute alliance can create a long repositioning window. Distance may also make a relation change less immediately exploitable.

On trade-heavy maps, manual embargo persistence matters because Port and Train plans may survive longer than one alliance cycle.

None of these map notes changes the timers or multipliers. They change how much positional value each rule carries.

FAQ

How long does an alliance request last?

Twenty seconds. The cooldown for requesting the same player again is 30 seconds from the original request creation time.

How long is a normal alliance?

The default is 300 seconds, or five minutes. Custom lobbies can use one to 15 minutes, and zero disables alliances.

Does an alliance extend automatically?

No. Both players must approve. The extension window appears during the final 30 seconds, and success resets a full duration from the renewal tick.

Is natural expiry a betrayal?

No. Natural expiry ends the alliance without creating the manual-break traitor state.

How long does traitor last?

Thirty seconds when the manual-break eligibility conditions are met. During that state, attackers receive 0.5 loss and 0.8 tile-processing-cost multipliers against the traitor defender.

Does an alliance clear every embargo?

No. Acceptance clears temporary embargoes between the pair. Manual embargoes persist and can still block trade.

How long does an attack embargo last?

The temporary embargo created by a normal attack lasts about five minutes. A manually applied embargo is persistent.

Do exactly 100 weighted allied tiles break an alliance?

No. The tile condition is strictly greater than 100. An allied structure in the outer blast radius can still trigger the separate structure condition.

Can MIRV never break an alliance?

That broad claim is unsupported. Only the MIRVWarhead branch is shown to skip this check in NukeExecution; do not generalize it to the whole MIRV sequence.

Are teammates immune to nuclear damage?

This source set does not prove that. It only shows that team membership is not removed through the normal alliance-break action.

Next Reading

Use nuclear deterrence before placing a strike near allied territory or structures.

Review team naval control when diplomacy protects Ports, coasts, or shared sea routes.

Use Port versus Factory to understand the trade network an embargo may interrupt.

Return to the first-match guide if you need the basic attack, alliance, and map-reading controls before using this decision framework.

If a diplomatic change leaves a player exposed, switch to the Recovery Playbook. For ongoing donations, targets, and logistics, use Team Roles.

Official Sources

• OpenFrontIO v0.33.7 release
• Alliance request execution and request model
• Alliance extension execution and alliance model
• Manual alliance break execution
• Embargo execution and Embargo All execution
• Player trade and friendly-state implementation
• Game acceptance cleanup
• Combat modifier utility
• Nuke alliance-break execution and official tests