OpenFront AFK Teammate Takeover: Absorb a Disconnected Ally for Zero Troop Loss

When a teammate hard-disconnects, the server stops treating them as a friendly ally, so you can attack their land and take every tile at zero attacker troop loss while their own troops keep bleeding at the normal per-tile rate. In a ranked 2v2 the same disconnect can drop your team out of the winner set or end the match winnerless. Read the in-game disconnect badge before you commit, because the timer is not the ground truth.

Direct answer: who absorbs the land, and at what cost

The single most useful fact is that the takeover is one-way and, for the absorbing team, effectively free. The open-source attackLogic function checks one flag on the defender — isDisconnectedTeammate — and when it is set it zeroes the mag value that drives the attacker’s troop loss, so attackerTroopLoss comes out as exactly 0. The defender does not get the same kindness: they still shed their average troops-per-tile, computed as defender.troops / defender.numTiles, on every tick of the absorption. The golden attack test pins the concrete numbers for a standard Plains encounter between two 20,000-tile players where the defender holds 100,000 troops and the attack pushes 100,000: against a connected equivalent you lose about 38.56 troops per tick, but against the disconnected teammate you lose 0, and the defender bleeds 5 troops per tick at the same absorption speed. That speed is identical to a normal same-terrain push, so you are not paying a speed penalty to collect the land either, and you are not paying a density penalty even though the disconnected player’s army is sitting on top of the tiles you are taking.

The decision is therefore never “can I take their land” — you can, and the server is built to let you — but “is taking it the right move right now, and who should be the one to do it.” That reframing matters because the free-troop part is so consistent that it should change how you staff your frontline and how you treat a teammate’s badge, not just what button you press when a disconnect appears. A player who has internalized the zero-loss fact stops hovering over the disconnect with a small stack out of caution, because the caution is misdirected: the stack does not die, the land is free, and the only thing that is actually at risk is the time you spend walking the new border while the enemy watches. The rest of this guide is about that second question — the timing, the assignment, and the ranked edge cases — because the cost question is already answered by the code and the golden test, and the only reason it feels uncertain to a player is that the disconnect is a stressful moment and the instinct is to do something, when the correct move is often a deliberate, assigned, low-cost absorption rather than a panicked grab. The timing of that absorption is also something the code decides for you: the free window opens the moment the server marks the teammate disconnected, which is driven by the ping and connection state rather than by a countdown you can see, so the useful habit is to treat the badge as the trigger and to have the absorber already positioned before the badge appears rather than reacting to it after the fact.

What the server actually does when a teammate drops

Two separate clocks control this, and confusing them is the most common mistake players make. The first is disconnectedTimeout, set to 1 * 30 * 1000 milliseconds — thirty seconds — in the game server. A player who goes quiet for that long is marked disconnected, and that is the moment isFriendly changes behavior. The relevant method reads: if the other player isDisconnected() and the treatAFKFriendly argument is false, which is the default, it returns false, so a disconnected player is no longer friendly even if they are on your team. There is no config switch that flips this back on at v0.34.16; the non-friendly treatment is the default path, not an option the host can enable, so you cannot negotiate a softer window and a teammate cannot ask you to hold the line for them.

The second clock is emptyGameTimeout, ten minutes, which is a backstop that reaps a started game whose roster has gone completely empty — it is about ending a dead game where nobody is connected, not about marking an individual player, so do not read it as a “disconnect lasts ten minutes” rule and do not wait for it to “resolve” a teammate’s state. The practical consequence of the thirty-second clock is that a brief tab-switch, a phone ringing, a browser tab suspended in the background, or a network blip under thirty seconds usually does not mark you disconnected, and the failure mode is the reverse: a hard crash, a power loss, or a closed laptop marks you almost the instant the ping clock goes quiet. A bad habit is to treat the badge as a countdown timer you can read off your own screen. It is not. The ground truth is the badge the game draws on that player’s tiles and the roster, because the server’s ping and connection state decide the mark, not a wall-clock count you can watch. Plan your reaction around the badge appearing, not around guessing whether thirty seconds have passed, and know that the mark is effectively immediate from the server’s point of view once the connection is truly gone. It is worth separating two more things the badge tells you and the timer does not: the badge marks a single player, not a team, so a team of four with one disconnected member still has three friendly players and the win logic keeps counting the team’s land as long as at least one member stays connected; and the badge state is what every other player’s isFriendly call reads, so the moment the badge appears, every teammate’s attack against that player’s land silently switches from a loss-bearing push to the zero-loss absorption described below, with no announcement and no confirmation dialog.

The zero-loss takeover, precisely: what you gain and what you lose

Reading the attackLogic path line by line shows exactly where the generosity lives and where it stops. The attacker’s loss is mag times the traitor modifier times a clamped troop ratio times a cost made of a base plus the defender’s troop density. When the defender is a disconnected teammate the function sets mag = 0 before any of that multiplies, so the entire attacker-loss term collapses to zero regardless of how packed their land is or how big your stack is. The defender’s loss is computed separately as defender.troops / defender.numTiles and is unaffected by the flag, so their army bleeds at the normal rate while their tiles transfer to you. The golden snapshot confirms the asymmetry with clean numbers: connected human-versus-human, the attacker loses 38.5576 troops and the defender loses 5; disconnected teammate, the attacker loses 0 and the defender still loses 5 at the same absorption speed. The table below is the reference you should keep in your head:

Two things fall out of this, and both are decisions, not trivia. First, the takeover costs you nothing in troops, which means you should never be the one holding a fragile stack on the same border you are about to hand back — the safe move is to commit a healthy push, because the push itself is free and the only thing at risk is your own time on the border. Second, the disconnected player’s remaining land does not freeze: every tick of absorption, their troops drain, so a disconnected teammate who had a large army concentrated on a small footprint bleeds proportionally more per tile than one who is spread thin. The defender’s 5 in the golden case is 100,000 troops over 20,000 tiles, so the absorption is a real transfer of their army into your hands, not a cosmetic one. This is the part that should change your pre-game planning: you want your teammates’ armies to be positioned such that, if one of them drops, the land they are sitting on is land you want and land that bleeds slowly enough that a calm, assigned collection beats a scramble. A backliner who parks a huge reserve on a small coastal footprint creates a situation where, on disconnect, that reserve is both a liability you should not chase and a bleeding wound the enemy can reach if you are not the one collecting it first. There is also a subtlety in the defender’s number that matters for planning: because the per-tile bleed is the defender’s total troops divided by their tile count, a disconnected player who is land-rich but army-thin bleeds little per tick and their land stays a stable free target for a while, while a disconnected player who is army-rich and land-lean bleeds fast and the value of the land is only there if you collect it quickly, so the same badge can mean “collect calmly” or “collect now” depending on the footprint you can see on the minimap.

The decision frame: when absorbing is right and when it is a trap

Because the takeover is free, the temptation is to always do it, and that is exactly the trap. The decision has three inputs, and the correct call is the intersection of all three, not the most urgent-sounding one. The first input is the win position: if your team is already above the strict share of non-fallout land needed to win, absorbing a disconnected teammate is often a net negative, because you spend frontline time and border exposure collecting land you did not need, while the enemy team watches and repositions to press your now-exposed new border. The win threshold in this build is a strict greater-than of 80% of non-fallout land, so sitting at 80.00% is not a win and a small push can be worth making, but sitting at 85% is a win and the marginal land is worth defending rather than expanding into.

The second input is who should absorb. The player who absorbs commits a stack and holds the new border, so the right absorber is the frontliner with a healthy army and a clear line of sight on the disconnected land, not the backliner who is needed to hold the other front or to donate gold into the frontline. The community’s team-game signal is consistent that the front-line role is the most exposed and the most dependent on the team donating into it, so the absorption should be assigned to whoever can afford to leave their current line without breaking it, not to whoever is geometrically closest to the disconnected tiles, because being closest is exactly the trap that pulls the decisive stack away from the deciding line. The third input is the enemy’s response: a free absorption that moves your border forward creates a new front, and a well-coordinated enemy team will press the freshly handed border while your absorber is still consolidating. The decision rule is: absorb when it fixes a real deficit or extends a winning margin, assign it to the frontliner who can spare the stack, and only if you can hold the new border against the enemy’s expected counter. If any of those three is false, the free takeover is a gift you should decline, because declining a free thing is the highest-leverage decision in the game. One more refinement that keeps the frame honest: the win-position input should be checked against your team’s combined non-fallout share, not your own personal share, because the win logic counts the team as a unit, so a teammate holding a large chunk of that share can mean your personal number is low even though the team is winning, and chasing land in that case would be correcting a number that was never the deciding one.

Scenario 1: ranked 2v2, your partner goes AFK at the frontline

Assume the situation the ranked threads describe: you and your partner are in a ranked 2v2, your partner is the frontliner holding the main border with a large stack, and they hard-disconnect with the game mid-climb. Your own territory is at about 72% of the non-fallout land, your partner’s remaining land is roughly 11% and concentrated near the front, and the enemy team is at about 17% and actively pressing your partner’s old line. Two server rules govern the next ten minutes, and they are different rules that do not overlap. In ranked 2v2, a team drops out of the winner set the moment every member of that team is dead or disconnected, so the instant your partner’s connection is marked and your partner’s last tiles are gone, the only connected team left is yours, and the win check will award the match to your team on the next tick. That is the part most players miss: the disconnect can win you the 2v2 without you taking a single extra tile, because the enemy is now the only connected team against your connected self and the rule reduces to a single remaining team.

The separate cancellation rule only applies if a matched player never actually spawned; a player who spawned and then disconnected does not trigger the winnerless archive, so your match stays live and can still be awarded to you. The decision is then about the free land: your partner’s 11% is free to absorb at zero troop loss, but you do not need it to win once the opponent’s team is the only connected team left, so the correct play is to not chase the disconnected land with your main stack; keep your stack where it holds the 72% and the enemy’s press, and let the win check do its work. If the enemy team also loses a player, you should reconsider, because the single-remaining-team condition is what is awarding you the match, and a second disconnect on their side does not change your position but a reconnection on their side does. The number to watch is your own 72%: as long as you hold it, you do not need the partner’s 11%, and chasing it is how a free bonus becomes an exposed front the enemy is already walking toward. One more thing the 2v2 case clarifies: with only two people on your team, there is no spare absorber at all, which is why the ranked play is almost always about not collecting the free land rather than assigning it, because the moment you commit your only stack to the coast you have left your main line unmanned with no teammate to cover it, and in a two-player team the border cost of a free absorption is paid by the only person who is also supposed to be holding the win.

Scenario 2: FFA team mode, a backliner drops and the enemy is pressing

Now assume a team game that is not ranked: you are on a four-man team against a four-man team, and your backliner — the one not holding the front, the one whose role is to hold the islands and donate into the frontline — hard-disconnects while the enemy is pressing your main line. Your frontliner is at 64% of your team’s non-fallout land, the enemy team holds about 30%, and the disconnected backliner still owns roughly 6% near the coast with a large reserve army that they were saving to donate. The free-takeover rule now cuts both ways, and the wrong instinct is the one that makes players lose. On your side, any teammate can absorb the backliner’s 6% at zero troop loss, and because the backliner’s army is concentrated on a small coastal footprint, those troops bleed at a high per-tile rate, so the land is worth collecting quickly to stop it evaporating into the enemy’s reach.

But the trap is the assignment: the backliner’s land is far from the enemy’s press, so sending your frontliner to collect it means pulling the only stack that is stopping the enemy’s 30% from climbing, and that is how a free bonus becomes the loss of the match. The decision frame says to decline the free absorption when it weakens the line that is actually deciding the match. The correct play is to have the frontliner hold the border and let the win position be decided by the main line, and to treat the disconnected backliner’s land as a secondary objective to collect only after the enemy press is blunted. If a second teammate on your side is safe behind the islands, that is the one who should collect the coast, because the absorber is the one who holds the new border and a safe second line can hold it without opening a front. The number that decides it is your frontliner’s 64% against the enemy’s 30%: you are winning the main line, so the free 6% on the coast is worth more as a reserve than as a new front you open while the enemy attacks. The lesson generalizes: the free land is only free if you can afford the border it creates, and in a four-man team the border cost is paid by whichever teammate leaves the line to collect it. The same reasoning runs in reverse for the enemy side of the screen: if the opposing team loses a backliner the same way, their disconnected land is equally free for you to collect, but the correct move is still the same discipline of not opening a second front with the stack that is holding the deciding line, so the rule is symmetric and it is a question of which free land you can reach without leaving your main line short, not of which side disconnected.

Failures and how to counter each one

The failure surface here is judgment, not a bug, because the server does the right thing on every branch, which means every mistake is yours to make and yours to prevent. The first failure is the timer fallacy: assuming you have thirty seconds to react to a disconnect, or assuming a disconnect clears itself after a fixed time. The counter is to react to the badge, not a clock, because the mark is set by the server’s connection state and the badge is the only thing you can see, and the mark is effectively immediate once the connection is truly gone, so the window to react is shorter than the thirty-second number suggests.

The second failure is absorbing when you are already winning, which converts a free bonus into a new, exposed front the enemy can press; the counter is the win-position check from the decision frame — if you already clear the strict share, defend rather than expand. The third failure is the wrong absorber, where the backliner who should be holding the other front or donating goes and collects the disconnected land, leaving the main line short; the counter is to assign the absorption to the frontliner who can spare the stack and hold the new border, and to make that assignment explicit before the game. The fourth failure is misreading the ranked 2v2 rule, where a player assumes a teammate’s disconnect ends the match as a loss when it is actually the single-remaining-team condition awarding the match to the connected team; the counter is to know the two distinct rules — the never-spawned cancellation that archives a match winnerless, and the in-game disconnect that reduces the team set — so you do not panic or concede a match the server is about to award you. The fifth failure is assuming the disconnected land is frozen in place; it is not, the defender’s troops bleed every tick, so a concentrated army on a small footprint drains faster and the land that is free to take now is the land that is least worth taking in a minute. Each counter is a habit: check the badge, check the win position, assign the absorber, know the two ranked rules, and act before the free land bleeds away. The unifying point is that none of these failures is a bug the server owes you a reminder about, which is the inverse of most common gameplay advice; the server is doing exactly what the open-source code says on every branch, so the whole safety net is yours to build before the game starts, and the single highest-value habit is deciding in the lobby who the named absorber is and where they stand, because every one of the five failures above becomes much easier to avoid once the absorption is a pre-assigned duty rather than a reaction to a badge flashing on your screen.

Version boundary and what this guide does not assert

The zero-loss teammate takeover and the ranked 2v2 disconnect rules are present in the tagged v0.34.16 source: the isDisconnectedTeammate branch in attackLogic, the isFriendly check that returns false for a disconnected player, the thirty-second disconnectedTimeout, and the checkWinnerTeam logic that drops a team once all of its members are dead or disconnected. They are absent at the v0.31, v0.32, and v0.33 tags, where the related golden attack test and snapshot files do not exist at all, so the feature was introduced between v0.33 and v0.34 and landed on the main line by the merge of the corresponding PR on 2025-10-06. The release listing for that window has gaps — deleted or renamed releases — so the verifiable boundary is the presence of the code at the v0.34.16 tag plus the merge date of the feature PR, not a clean release note that you can point to, and that is exactly why this guide cites the tag and the merge rather than a release body.

The practical line for a player is therefore: if you are on v0.34 or newer, everything in this guide applies to your client; if you are on v0.33 or earlier, the zero-loss teammate takeover is not in your build and a disconnected teammate’s land behaves like a normal, loss-bearing target, so the free-absorption advice does not apply and you should treat the disconnect as a normal loss-bearing attack. This guide does not assert a reconnect grace period longer than the thirty-second mark, because the client does not expose a longer configurable window at v0.34.16, and it does not assert what a clan or a leader can do to prevent the absorption, because the open-source client defines the mark and the loss, not a social override. If a community video you are watching predates v0.34, it cannot document this feature, and the disconnect behavior it shows is the old behavior, so treat the version boundary as a filter before you trust any walkthrough of what a teammate’s disconnect does. The two community walkthroughs gathered for this guide both postdate v0.34 and cover the current behavior, which is the only behavior this guide describes. A final boundary worth stating plainly is the difference between “present at v0.34.16” and “guaranteed stable forever”: the code you are playing today is the v0.34.16 tag, and this guide is a description of that tag, so if a future build changes the zero-loss behavior, the disconnect timeout, or the ranked win logic, the numbers and the advice here stop being authoritative for that newer build and you should re-verify against the newer tag rather than trusting a version number you are not on.

Modes, maps, and how to set up for a clean absorption

The rule is the same in every mode, but the value of a clean absorption is not, because the mode decides what the disconnected land is worth to you and the map decides how fast you can collect it. In ranked 2v2 the mode is the whole game: the disconnect can decide the match through the single-remaining-team rule, so the setup is to keep your own connection stable and to know that your partner’s drop can award you the match rather than cost it, and the map’s front-line width matters less than the fact that a team of two has no spare member to absorb with, which is why the ranked decision is about not chasing the free land rather than about who collects it. In a FFA team game the value of the absorption depends on the front line: on a map where the fronts are short and the land is dense, a disconnected backliner’s concentrated army bleeds fast and the coast is worth a quick collection by a safe second player; on a map with long, thin fronts, the disconnected land is far from the deciding line and the correct setup is to assign the absorption to whoever is not holding the front, because walking down a long thin front to collect free land is exactly how you open a second front.

The map also decides the speed, because the absorption runs at the same absorption speed as a normal push, so on a small map the land transfers quickly and on a large map a distant disconnected territory takes longer to walk down, which is another reason to send the safe player rather than the frontliner, and a reason to pre-position a second stack near the likely disconnect points before the game starts. The single setup habit that covers all of it is to decide, before the game, who is the absorber when a teammate drops — the safe second line, not the front — and to let the badge, not the timer, trigger the assignment. A team that has named its absorber in the lobby, and has positioned that absorber so the likely disconnect land is in reach without breaking the main line, turns the disconnected teammate from a stressful scramble into a calm, assigned, free collection, which is the entire point of reading the badge instead of the clock.

• Team victory threshold: turn 80% into a safe finish
• Ranked 2v2: partner, lanes, and the ladder
• Team economy and space: who holds the front and who donates