OpenFront Cross-Water Army Commitment: How Much of Your Force to Commit Across the Sea

Direct answer: commit the minimum that wins, and keep a reserve that regrows

The question behind every boat operation is not “how many boats can I send” but “how much of my army can I afford to have gone at the same time.” A boat attack in OpenFront is a one-time projection of troops across water, not a standing army you can call back and forth. When you right-click a target and choose Boat, the game spawns Transport ships from your Port, and each boat carries 20% of the troops you selected — capped at three boats. That means at most about 60% of the force you picked is actually leaving your land, and the crucial detail is when that happens: the moment the boats launch, those troops are subtracted from your land pool. They are not parked somewhere “in reserve” on the water that you can fold back in. So the real decision is a tradeoff with a fixed shape: you trade away a large share of your strength to hit the enemy across water, and you keep a home reserve whose job is to survive the drain, hold your own coast, and regrow while the crossing is in the air.

If you want the per-landing mechanics of choosing a beach and allocating waves, that is covered by the transport landings guide, which handles the single landing rather than the whole projection; and if your question is the cheap 20%-a-boat harassment rather than a committed operation, that is the micro boat spam guide. This page sits one level up from both: it is about how much of your army to put at risk across the water at once, and how much to keep home.

The single most common mistake, and the one this whole page is built to prevent, is committing the maximum (the full 60%) on the assumption that the troops are “out there but safe.” They are not. They are gone from the pool, the operation’s outcome is largely binary, and if the enemy screens the lane with a Warship or nukes the landing, you have spent most of your army and you still hold the same amount of land. The correct default is to commit only enough to break the enemy’s weakest front — often far less than 60% — and to size the rest of your pool so that your own coast and your own regrow still work after the drain. The sections below give you the exact numbers, the two scenarios (a full send you are confident in, and a probe you can afford to lose), the failure modes, and the mode- and map-specific adjustments. Read the decision table near the end as your quick reference before you press Boat.

How a boat commitment actually drains your pool

The first thing to internalize is that a boat crossing is a transfer, not a loan. Before you launch, all of your troops are in your land pool and they count toward your troop cap. The instant a boat departs, the game runs the same troop-removal path it uses for a ground attack: the boat’s load is subtracted from the pool that is standing on your territory. If you had a 100,000-troop pool and you launch three full boats, roughly 60,000 of those troops are no longer in the pool; they are on the water and the game no longer counts them toward what you can command from land.

This matters in three concrete ways. First, your effective defensive strength drops exactly by the amount you committed. A warship screen, a nuke, or a counter-invasion that would be trivial before you launch becomes dangerous after, because the troops you would normally throw at the counter are out on the water and cannot be recalled in time to help. Second, your troop growth behaves differently while the pool is low. OpenFront’s growth is scaled by the distance from the cap, so a pool that has been drained down actually regrows faster than a pool sitting at the ceiling — but only after the pool has dropped. If you commit 60%, your pool has dropped hard, so the regrow rate is good, but you also have less total strength to spend on the beachhead fight. Third, the crossing takes real time. Boats move across water at a fixed speed, so for the duration of the crossing your committed troops are in limbo: not available to defend, not yet able to fight. The longer the lane, the longer that limbo window, and the more your home front is exposed.

The practical implication is that you should never treat the boats as a reserve you can bounce. There is a small malus if you recall a boat, but the troops do not snap back to full strength, and a recalled boat has just wasted the crossing time. Design the operation so that the outcome you want is the one where the boats land and fight, and treat a recall as a way to stop a bad operation, not a way to bank troops. Keep the home pool large enough that, even if every boat is sunk or nuked, you still hold your own coast and can let the pool regrow toward the cap over the next several minutes.

The exact numbers: 20% per boat, three boats, free, 25% retreat

Let the exact figures anchor the rest of the page, because a lot of the community’s confusion comes from approximate numbers floating around. These are the verified values from the official source at the pinned commit, so treat them as the ground truth rather than the 30%, 40%, or 150,000 figures you will see in forum threads.

Quantity	Value	What it means for your decision
Boats per attack	3	Three Transport ships can be in the water at once per player
Load per boat	20% of selected troops	Each boat carries floor(troops/5) of the pool you picked
Maximum commit	60% of selected pool	3 boats x 20%; the rest stays on your land
Cost	0 Gold (from a Port)	Transporting troops is free once you own a Port
Retreat penalty	25% malus	A recalled boat loses a quarter of its troops
Troop cap	2 x (tiles^0.6 x 1000 + 50000) + city troops	The pool ceiling that growth is measured against
Growth shape	Fastest when the pool is far below cap	Draining the pool speeds regrow, but only after the drop

Reading the table top to bottom: the 3-boat cap is the hard ceiling, so you can never project more than 60% of a single selection across the water in one operation. The 20%-per-boat load means that a “full boat” is always a quarter of whatever you right-clicked; if you selected 50,000 troops, a full boat holds 10,000. The 60% combined maximum is the number that drives the whole commit-vs-retain decision: it is the largest share of your army you can have in limbo at once. The free cost is important because it means the only thing you are spending is troop strength and time, not gold — so the cost of a bad crossing is paid in army, which is the scarcest resource in the late game. The 25% retreat malus is why a recall is a partial loss, not a free undo. And the cap and growth rows are the reason your home reserve matters: you want the pool to be far enough below the cap that it regrows quickly, but not so low that you cannot defend your own coast.

A useful mental model is to think in thirds. A single 20% boat is a probe. Two boats (40%) is a real operation you expect to win. Three boats (60%) is a full commitment you only make when the enemy’s front is clearly broken and your home coast is safe or screened. Choosing which third you are in is the entire art of cross-water play, and the two scenarios below show what each looks like with actual numbers.

Why the old boat-banking exploit is dead

A large share of the player discussion around boats is still running on older assumptions, and a few of those assumptions are flatly wrong at the current version. The most persistent one is the “boat banking” exploit: the idea that you could push troops onto boats, hold them on the water above your troop cap, and then fold the full bank back into a bigger ground attack. Some forum posts describe sending a 30% boat toward an island target, waiting, retreating it back to your own coast, and adding its troops to a second attack — essentially treating the boat as an off-cash vault for extra troops. That does not work the way it once did, for two reasons.

First, the troops are deducted from your pool at launch, not held above the cap. The moment a boat departs, its load leaves the pool, so there is no way to accumulate a hidden bank of troops on the water that exceeds your cap and then unleash it. Second, a recall is not a clean reversal: the returning boat carries a 25% malus, so the troops you get back are fewer than the troops you sent out. The round trip costs you a quarter of the load every time. The community numbers that leak into this confusion — “30% retreat,” “you lose like 40%,” “150,000 appears with them” — are all approximations of the real mechanic (three 20% boats with a 25% malus) and they overstate both the size and the safety of the bank.

This is a version-boundary note worth being explicit about: the behavior described here is the current v0.35 line, verified against the pinned source at commit f33d00490. If you are playing on a much older client or a community server running old logic, the numbers may differ, and the banking trick may behave differently. But on the current official game, the rule is simple and it changes your play: boats project troops, they do not store them. Design every crossing on the assumption that the committed troops are gone for the duration, and that the best case for you is the case where they land and hold the beachhead. That single reframing is what separates a player who loses a third of their army to a nuke over open water from one who commits only what the situation can absorb.

The practical takeaway is that the interesting decision is no longer “can I bank troops in boats?” (no) but “what is the right fraction of my pool to project across this lane right now, given the enemy’s weakest front and my own home-defense minimum?” The next section is the framework for answering that with numbers rather than with the old exploit logic.

Sizing the commit: the home-defense minimum

The decision framework reduces to one question you answer before every launch: what is the smallest number of boats that still wins, and how much do I keep home so I survive a loss? Start from the enemy side. Read their front the way you would read any attack: where is their weakest border, how thin are their troops there, and can you break through with fewer troops than they have? If the enemy has stretched a thin front across a wide map, you may need only one boat (20%) to punch a hole. If the enemy is compact and thick, no fraction of your pool crosses a lane into that front without being over-matched, and the honest answer is “do not cross this lane right now.”

Now the home side. Your home reserve has two jobs at the same time: it must be able to hold your own coast against a counter-invasion (you cannot let the enemy land on you while your army is in the water), and it must be low enough below the cap that the pool regrows quickly enough to matter. A good rule of thumb is to leave at least as much home as you are sending out whenever the enemy has an active front of their own, and to leave more than that if the enemy owns a Warship that can screen your lane. The reason to leave more than you send is asymmetric: if your boats are lost, you have spent 60% of the pool and you are left defending with 40%; if the enemy launches a counter while you are committed, that 40% is all that stands between the enemy and your coast. A small Warship screen in your own lane raises the cost of that counter for the enemy, which is exactly why the screen and the home reserve are the same decision.

Put it together and you get a concrete procedure. Step one: pick the enemy’s weakest front and estimate the troops it would take to break it — call that the need. Step two: express the need as a fraction of your pool; if it is 20%, that is one boat, if 40% that is two, if 60% that is three. Step three: check your home reserve after that commit; if it is below the level you need to hold your coast and regrow, do not launch the full commit — reduce it or wait for regrow. Step four: decide whether you can afford the 25% malus if you must recall, and whether a Warship can screen the lane long enough. Only when the need, the reserve, and the screen all line up do you press Boat at the size the need demands.

Scenario A: the 60% full send against an over-extended enemy

Suppose you have a 100,000-troop pool sitting at a high fraction of your cap. Your main enemy has stretched a thin two-front defense across a wide map, and you have spotted that one of those fronts is down to roughly 30,000 troops while they hold a large territory. They have no Warship near the crossing lane, and their nukes are spent or far from the target. This is the situation where a full 60% send is the right call, and the numbers show why.

You right-click the thin front and launch three full boats. Each carries 20,000 (20% of 100,000), so 60,000 troops leave your pool and your home pool drops to 40,000. The crossing is long enough that you watch the lane, but the enemy has no Warship in it and no nuke near the landing, so the boats arrive intact. The 60,000 beachhead meets the enemy’s 30,000 on that front — a clean 2-to-1, which is the ratio that wins land fights — and you break through and annex the territory. Because you committed the maximum, your pool is now thin at 40,000, and that is exactly the trade you made: you traded 60% of strength for a decisive territorial gain. Your job now is to hold the beachhead and let the pool regrow, which it will do faster than normal because the pool is far below the cap.

The reason this specific case works is that every condition lined up: the enemy front was genuinely broken (30,000 against your 60,000), the lane was unscreened (no Warship), the landing was not nuke-covered, and your home coast was safe. If any one of those were false, the same 60% send would have been a mistake. The discipline is not “always send 60% when you have 100,000” but “send 60% only when the enemy’s weakest front is weak enough, the lane is open, and the landing is safe.” That is the difference between a full send that wins a territory and a full send that feeds 60,000 troops to a nuke over open water. When you are this committed, you are betting on the landing holding, so make the bet only on a front you have measured, not one you are guessing.

There is also a cost you should account for on the far side of the send. A 60% commit leaves a 40,000 pool that is thin against any counter-invasion, so the send is only defensible if the enemy’s own front is inactive or you have a screen. If the enemy is simultaneously preparing to land on you, a 60% send is a coin flip: you might win your landing while losing your coast. The correct check before the send is to ask whether the enemy has a live threat of their own. If they do, downgrade the send to a 40% probe even if your target front is soft, because the home reserve is what keeps a counter-invasion from ending your game while 60% of your army is in limbo. The 60% send is the maximum you can justify, not the default, and the situations that justify it are rarer than they look.

Scenario B: the 40% probe you can afford to lose

Now the mirror case: the same 100,000-troop pool, but the enemy is compact and you are not sure their front is broken. You do not have a Warship screen in the lane yet, and you suspect a nuke might be aimed at the landing. This is the situation where a 60% send would be reckless, and the right move is a two-boat probe you can afford to lose.

You launch two full boats, 40,000 troops, leaving 60,000 home. That 60,000 home reserve is the point of the whole operation: even if both boats are sunk or the landing is nuked, you still have 60,000 troops to hold your coast and regrow. If the probe lands and the front is as soft as you hoped, you break it and you can follow up with a third boat in the next crossing window to push the advantage. If the probe fails, you have lost 40% but you are still a functioning player with a 60,000 pool and a home front intact, and you have bought real information about the enemy’s strength that no map view could give you. That information has value in its own right, because now you know whether to commit a third boat or to pull back and regrow.

The probe also gives you a clean decision rule for the follow-up. After the first 40,000 land, you watch two numbers: the enemy troops that rotate onto the beachhead, and the speed of that rotation. If the enemy can throw more troops at the landing than you brought, the front is thicker than it looked, and the correct follow-up is to hold the beachhead, not push a third boat into a growing pile — a third boat into an over-matched landing just converts a 40% loss into a 60% one. If the enemy cannot out-troop you on the beach and the front is soft, the third boat is the right escalation, because you now have measured certainty that the front breaks. The probe is, in other words, a cheap way to buy the measurement that Scenario A demands: you paid 40% at worst to learn whether a 60% commitment is safe, and that payment is the rational price of the information when the front is genuinely unknown.

The discipline here is the opposite of Scenario A: you deliberately commit less than the maximum so that a loss does not end your game. The 40% probe is a controlled risk with a bounded downside, and the bounded downside is what makes it rational. Compare the two scenarios side by side: in A you committed 60% on a measured, open, safe front, so the downside was acceptable because the upside was a guaranteed territory; in B you committed 40% on an unknown, possibly screened, possibly nuked front, so you sized the commit so that a full loss left you still able to play. The choice between A and B is the whole cross-water decision, and the honest answer to “how much do I send” is “as much as the certainty on the front justifies, and no more.”

Failure modes and counters: warships, nukes, and being over-matched

A cross-water commit fails in a small number of distinct ways, and each has a specific counter that you should check before you launch. The first and most common is the Warship screen. If the enemy has a Warship in your lane, its shells will sink your transports before they land, and you will have deducted the troops from your pool for nothing. The counter is to check the lane for enemy Warships before you launch, and if there is one, either move your own Warship to contest the lane or reduce the commit so the loss is survivable. A screen is also why you keep a home reserve: if the enemy can screen your lane, they can often counter-invasion your now-exposed coast, and the reserve is what holds them off.

The second failure is the nuke on the landing. A MIRV over your beachhead is far worse than a 25% retreat; it can delete the committed troops outright and the landing site with them. The counter is to read the enemy’s nuke posture: are their silos aimed at your likely beachhead, and is the warhead available? If the landing is nuke-covered, do not commit the maximum into it. A smaller probe that the nuke cannot fully delete, or a landing offset from the aimed grid, is the mitigation. The community is loud about this, and the correct posture is to assume any open-water landing near a defended coast is nuke-able until you prove otherwise.

The third failure is simply being over-matched at the beachhead: you cross the lane safely, the boats arrive, but the enemy’s front on the landing side is thicker than you measured and you lose the landing fight. The counter is to measure the front, not guess it, and to commit only up to the ratio that wins. If the enemy has 60,000 on the landing front and you bring 40,000, you are over-matched regardless of how safely you crossed. The counter to that is either to break the front from a different, softer angle first, or to wait and regrow so your next commit is larger. A fourth, quieter failure is the timing error: you commit correctly in size but at the wrong moment, so that while your troops are in limbo the enemy shifts their front, spends a nuke elsewhere, or simply reorganizes, and the soft front you measured has hardened by the time the boats land. The counter to timing is the same probe logic from Scenario B: send a small first boat to confirm the front is still soft at landing time, and only follow with the rest once the boats are already close enough that the front cannot meaningfully change under them. The window between “the boats launch” and “the boats land” is the most vulnerable part of the operation, and it is exactly the window where a measured front can stop being true.

Tying the three together, the pre-launch checklist is: Is the lane Warship-clean? Is the landing not nuke-aimed? Is the front I am landing on soft enough that my commit wins the ratio? If all three are yes, launch at the size the front demands. If any is no, reduce the commit, fix the screen, or wait. The players who lose armies over open water are almost always the ones who crossed a lane that failed one of those three checks on the assumption that “the troops are out there but safe.” They are not safe. They are gone, and the lane, the nuke, and the front are the three things that decide whether they come back as territory or as a loss.

Mode and map adjustments, and the decision table

The commit-vs-retain math is the same everywhere, but the defaults change with mode and map, so adjust the numbers you carry. On Four Islands and other team modes, the water lanes are shorter and the enemies are more predictable, so a 40% probe is often the whole operation and the home reserve you keep is mostly there to cover your teammate’s back. Team modes also reward screening the lane, because one shared Warship can protect multiple crossings. On large FFA maps with many hostile fronts, the 60% send is much more dangerous because the moment you commit, another player can invade your exposed coast from a different direction; the home reserve on a large FFA map is what keeps you alive, so size it larger and lean on two-boat probes rather than full sends.

Map shape drives the crossing time, which drives the limbo window. On a map where the lane you need to cross is short, the committed troops are in limbo for a few minutes and the home front is exposed for a short time, so a larger commit is tolerable. On a map with a long open-ocean lane, the same 60% sits in limbo for much longer, and the whole game can shift while your army is on the water; on those lanes, favor the probe or screen the lane hard. The strait and chokepoint maps sit in the middle: the lane is narrow enough to screen cheaply, which lowers the cost of a larger commit because the screen is easy to hold.

The decision table below is the quick reference to use before you press Boat. It maps the situation you are reading to the commit you should make and the home reserve you should keep.

Situation on the front	Lane / nuke	Right commit	Home reserve to keep
Enemy front clearly broken (you hold the ratio)	Warship-clean, not nuke-aimed	Full 60% (3 boats)	Enough to hold coast + regrow
Enemy front soft, some uncertainty	Warship-clean, low nuke risk	40% probe (2 boats)	More than the 40% you send
Enemy front unknown or thick	Any screening or nuke risk	20% probe (1 boat) or hold	Max, do not commit yet
Lane has an enemy Warship	Screened	Do not cross, or screen first	Full, contest the lane
Landing is nuke-aimed	Nuke-covered	Reduce commit, offset landing	Enough to survive the loss

Use the table as the floor, not the ceiling: if the situation is worse than any row describes, the answer is to hold and regrow rather than to commit. The whole page reduces to one line — commit only what the certainty on the front can justify, keep a home reserve that holds your coast and regrows, and screen the lane so the crossing cannot be recalled cleanly.