When to Stop Nuking and Commit a Ground Push in the Endgame

Short answer. Stop nuking the moment a ground push is more likely to convert your advantage than another bomb is. That moment is usually when you already hold a local troop majority on a front, when the enemy is bleeding port and income faster than they can rebuild, or when the Overtime threshold has dropped far enough that a steady tile grind will win before the bomb can. Nukes win fights; troops win games. Hold the line of attack, keep the Overtime math in front of you, and commit to the ground push when your troop front can actually break, not when your nuclear stockpile is still full.

Why the nuke-only posture stops winning

A lot of high-scorers reach the endgame with the same shape of board: they have spent most of the match in a nuclear posture, calling wave after wave of Atom and Hydrogen strikes against every city that shows a real population, and they have let the ground layer go quiet while doing it. The nuclear loop feels productive because every bomb lands, every detonation visibly deletes a tile, and the scoreboard goes up with each one. But the board is not a scoreboard; it is a race to hold enough of the map that you are the only side left standing when the win threshold is crossed. A bomb can only ever be the thing that takes the last few tiles if, before it lands, you already own the front line that lets it land. And a pure nuke posture is the one posture that guarantees you do not own that front line, because you have been spending your economy on warheads and silo coverage instead of on the troops who would walk in through the gap the bomb opened.

The deeper problem is timing. Nuclear weapons are a force multiplier on a target that is about to break anyway; they are a dead weight on a target that is about to be overrun. When your opponent has a healthy troop mass sitting in a defended city, a bomb that lands on that city buys you a short window, maybe a few turns, before the defenders pull back in from the surrounding tiles and re-form. The bomb changed nothing about the ratio of forces; it only deleted a chunk of the board and handed the defenders a reason to concentrate. So the more the match runs on nukes, the more it runs on the opponent’s ability to regroup between your strikes, and the less it runs on your ability to actually occupy. The endgame is where this shows up hardest, because by the endgame the map is small enough that a regrouped enemy can always reach you, and big enough that you cannot out-spam them for the last 20 percent of the board. The nuclear posture is a midgame posture that keeps working for a while in the late game, but it stops working exactly when it stops working, and the players who keep pushing bombs at that point are the ones who sit on a high score and still lose.

The concrete signal that the posture has stopped working is that your bombs are landing on cities that keep coming back. If you nuked a capital three turns ago and the same capital is still standing with a fresh garrison, the bomb did not break the enemy; it just spent 5,000,000 of your economy to reset their defensive clock. Every one of those reset bombs is a bomb that could have been a troop wave walking the city down for a fraction of the cost. So the first step in the transition is not a new strategy; it is the honest accounting that says your most recent several bombs bought you nothing you did not already have, which is the definition of a posture that has run out of value.

What actually changes in the late game

Three things change between the midgame and the endgame, and each of them pushes the balance away from pure air superiority and toward ground control. The first is the shrinking target space. In the midgame the map is large and diffuse, so a bomb can always find a fresh target, and the enemy always has somewhere to fall back to. In the endgame the map is small and contested, so the same bomb has fewer targets, and the enemy has fewer places to hide behind. When the bomb has fewer targets, its cost per useful tile goes up, and when the enemy has fewer places to fall back, their regrouping is slower and their defensive mass is more exposed to a push that actually arrives. The second change is the Overtime clock. Once the match passes 30 minutes, the Overtime window is open, and the win threshold that a player must cross starts dropping by 2 percent of tiles per minute with no floor. That means the cost of simply grinding tile by tile falls over time, and the cost of standing still gets higher over time, which is the exact opposite of what a nuclear standoff wants. The third change is income concentration. By the endgame most of the board’s ports and population hubs are in a handful of districts, so the economy is not spread out; it is clustered, and a ground push that takes the cluster takes the match, while a bomb that hits one hub and then waits for the next one to refill leaves the cluster intact.

These three changes do not happen at the same moment, and that stagger is what creates the decision window. You do not flip from nukes to ground at minute thirty; you flip at the moment when the shrinking target space, the falling threshold, and the clustered income have combined to make a ground push the higher-expected-value action. Sometimes that is a few turns after Overtime opens, sometimes it is a dozen turns before, because the Overtime window is a timer and the other two conditions are board states. The skilled player does not wait for the timer; the skilled player watches the board, sees the three conditions arriving, and makes the push while the board still favors it. Waiting for the Overtime threshold to do your decision for you is how you end up committing a push at the worst possible moment, when the threshold has already dropped so far that any push wins but only just, and only if you have not spent the last five minutes of the window nuking for the sake of the scoreboard.

The Overtime threshold is your real endgame clock

The single most useful number in the endgame is the Overtime win threshold, because it tells you how much of the board you must own to win, right now, and how fast that number is falling. At the start of a match the base win threshold is 80 percent of the tiles owned; one side that owns 80 percent of the map wins. The Overtime mechanism is disabled by default in a standard match, which means the 80 percent threshold holds until Overtime is triggered. When Overtime is active, it opens at 30 minutes by default, and from that point the required win percentage drops by 2 percent of tiles for every minute the match continues, with no floor, so the longer the match runs the less of the map a player needs to hold to take it. The practical reading of that is that a match that runs long is a match that becomes easier to win by a tile grind and harder to win by a standoff, because the standstill that a nuke-only posture tries to maintain is precisely the thing the falling threshold punishes.

The reason this matters for the transition is that the threshold is the number a ground push is chasing. A bomb does not move the threshold; it moves the board, and it moves it in a way that can be reset. A ground push moves the board in a way that sticks, and it does so by moving the threshold-crossing line closer and closer with every tile it occupies. So the endgame question is not “can I afford another bomb” but “how many tiles do I need to win right now, and how many of those can I take with a push that holds.” If the threshold has dropped to, say, 60 percent and you already own 55 percent with a local troop majority on one front, you do not need to delete the enemy’s last capital; you need to own the front and grind the remaining five percent, and a ground push is the only action that grinds. If instead the threshold is still at 80 percent and you own only 45 percent, a ground push is too expensive and a nuke is still the right tool, because the gap is too big for a push to close and a bomb is the cheapest way to delete a chunk of the enemy’s tile mass and shrink the gap. The threshold is the dial that tells you which side of that line you are on, and the disciplined endgame player keeps the dial in front of them and flips the posture when the number, not the scoreboard, says the push is now the cheaper win.

The transition decision framework

The transition is not a mood; it is a check you run against the board before you commit troops. Run the same five questions in the same order every time, and the answer will be the same shape: either the push is the higher-expected-value action, or it is not, and if it is not the bomb still has work to do and you keep the nuclear posture. Question one is the front test: do you already hold a local troop majority on at least one connected front, meaning more of your ground force than the enemy’s within the tiles that matter for the next push? If you do not hold a local majority anywhere, a ground push is a gamble on a front you do not own, and a nuke is still the cheaper way to buy the majority, so keep nuking. Question two is the income test: is the enemy bleeding port and population faster than they can rebuild, so that their next few turns of income are already smaller than their last? If the enemy’s income is climbing or flat, a push into that front will meet a fresh defensive wave and burn your troops for nothing, so keep the pressure on with bombs that delete the ports feeding that front. Question three is the threshold test: given the current Overtime threshold and your current tile ownership, how many tiles short are you of winning, and can a one-turn push close more of that gap than a one-turn bomb can? This is the math that the other two questions feed. Question four is the regroup test: if you commit a push this turn, will the enemy be able to pull a fresh mass in from an adjacent district and reset the front before your push is spent? If the answer is yes, you need to break the regrouping path first, which is often itself a small ground action, not a bomb. Question five is the irreversibility test: once the push is committed, can you still fall back to the nuclear posture if it stalls, or does the push commit your economy so far that a stall is a loss? The push should be reversible enough that a bad first turn is a setback, not a collapse, which is the difference between a transition and a surrender of the posture that got you here.

When four or more of the five answers point the same way, the posture is wrong and the push is the move. When the answers split, you are still in the nuclear regime and the right action is to keep buying the conditions, deleting the enemy ports that feed a regroup, until the front test and the income test both come back yes. The framework exists so the transition is a decision you make and can repeat, not a reaction you have when a bomb happens to land well. Players who transition on a lucky detonation are players who have not run the five questions, and players who never transition are players who have run them and found the front test never comes back yes, which usually means their economy has been going to warheads and silo coverage instead of to the troops who would make the front test pass.

Scenario one: the 2.5 billion nuclear burn

This is the board shape that the community keeps describing, and it is the clearest failure mode of the nuke-only posture. A player reaches a mid-to-late state where they have poured the equivalent of a 2.5 billion nuke budget into the enemy, deleting capital after capital, and yet the enemy is still holding a real share of the board and the match is still open. The player’s instinct is that they are winning because the bombs are landing and the enemy’s tile count is going down, but the deeper read is that the player has spent 99 percent of their economic power on deletions that the enemy keeps replacing, and the player’s own ground front is thin or absent. The enemy is not losing the board; they are losing individual cities on a schedule the player set, and every city they lose is one they refill from the ports the player has not taken. The 2.5 billion is not gone; it is parked in the form of the enemy’s ongoing ability to rebuild, and the player’s only remaining asset is a growing scoreboard.

The fix for this exact shape is a hard posture flip, and the five-question framework tells you where to flip. The front test comes back no on most fronts, because the player has not built a troop mass; but the income test may come back yes on the ports that are feeding the enemy’s rebuild, because the player has been deleting the ports’ cities but not the ports themselves, or the ports have been left under a light garrison. So the first move is not a big push; it is a targeted ground action against the specific ports and the specific district that feeds the enemy’s refill, which is a much smaller commitment than the push the player was afraid to make. Once that district is ground-held, the enemy’s refill rate drops, the next wave of nukes you still have stops meeting a fresh garrison, and the front test on the surrounding districts starts to flip to yes, because the enemy can no longer afford to mass a defensive wave in three places at once. The 2.5 billion that was parked in the enemy’s rebuild capacity is now a threat that is actually going to close, because the player finally owns the front that lets the bomb’s deletion stick. The lesson is that the nuclear burn was never the problem; the absence of the ground front that would have let the burn convert was the problem, and the burn only gets converted the moment the ground front appears.

Scenario two: the ground advantage you already built

The other endgame shape is the one where the transition is easy and the only question is timing. A player has run a mixed posture all match, nuking the enemy’s heavy cities and ground-holding the ports that feed them, and by the late game they already hold a local troop majority on two fronts and the enemy’s income is flat or falling. The front test and the income test both come back yes, and the only thing standing between the player and the win is the threshold test: how many tiles short of the current Overtime threshold are they. If the player owns 58 percent and the threshold has dropped to 62 percent, the player needs four percent, and a single committed push into the one remaining enemy district can close more than four percent in a turn or two if the front is as good as the two fronts the player already holds. This is the shape where the push is clearly the higher-expected-value action and the only real question is whether to commit the whole economy to it or hold a reserve.

The disciplined answer in this shape is to commit the push but leave a reserve that covers the one front where the enemy still has a real mass, because that mass is the thing that resets a stalled push. The push goes into the district with the flat or falling income, where the enemy cannot refill, and the reserve sits on the district where the enemy can still mass, where the reserve’s job is not to win but to deny the enemy the ability to break the push’s rear. The nukes that remain in the stockpile are not wasted; they are the insurance that says if the reserve front starts to crack, a bomb on the enemy’s regrouping path buys the reserve a turn to re-form. This is the transition done right: the ground push does the winning, the reserve holds the line the push depends on, and the last nukes are the safety net that lets the push be committed with confidence instead of with a prayer. The timing in this shape is still the Overtime dial, because the push is a grind and the grind only needs to reach the threshold, so the player commits the push at the moment the threshold is low enough that the push’s projected tile gain crosses it, and that moment is usually a few turns earlier than the player’s instinct, which is to wait for a more certain window that will cost a turn of the falling threshold. The player who has built the ground advantage through the match does not transition; the player just stops pretending the nukes are still doing the work and lets the troops do the work they have been building toward all match, and the reserve plus the last nukes are the two things that let that hand-off happen without a gap where the enemy resets the front while the player is deciding whether the push is ready.

How to convert nuke superiority into a push

Converting a nuclear advantage into a ground advantage is a sequencing problem, and the sequence matters more than the total amount of force. The first step is to pick the push’s path from the ports, not from the cities. The ports are where the enemy’s income and their refill capacity live, and a push that runs through the ports takes the enemy’s economy first and their tile mass second. A push that runs through the cities is a push that meets the enemy’s defensive mass head-on, which is the most expensive way to take the map. So the path of the push should be the path that takes the clustered income, even if that path is longer than the path to the enemy’s largest city, because the path that takes the income is the path that leaves the enemy with nothing to refill the city with. The second step is to time the push to the enemy’s refill gap, which is the window between the enemy’s last port being hit and their next port coming online. The push should land in that window, because a push that lands while the enemy has a full port behind it meets a fresh wave, and a push that lands in the refill gap meets a front the enemy cannot reinforce. The third step is to commit in one wave, not in dribs. A push that arrives in pieces is a push the enemy resets piece by piece, because each piece meets the full defensive mass before the next piece can combine with it. A single committed wave that breaks the front and then holds it is the only push shape that converts, because the break has to happen before the enemy can regroup, and the regroup only fails to happen if the wave is bigger than the enemy’s reset capacity all at once.

The fourth step is to hold the break with the nukes still in the stockpile, which is the part of the conversion that players skip because they feel they are “wasting” a bomb on a held front. The held break is the thing the enemy will try to reset, and the enemy’s reset is a defensive mass that pulls in from an adjacent district. A bomb on that adjacent district, on the regrouping path, is not a waste; it is the insurance that says the break you spent the push to open will still be open next turn. So the conversion is: pick the port path, time it to the refill gap, commit in one wave, and hold the break with the last nukes. Every one of those four steps is cheaper than the alternative it replaces, and the total conversion cost is usually far below the cost of the nuke-only posture that preceded it, because the conversion spends force on things that stick and the posture spent force on things that the enemy reset. The player who can run this sequence is the player who transitions; the player who can only run the posture is the player who sits on a high score and loses the match that the score says they were winning.

When the push fails and how to recover

A ground push can fail, and the disciplined player has a recovery plan before the push is committed, not after it has failed. The most common failure is the reset: the push breaks a front, holds it for a turn, and then the enemy pulls a fresh mass in from an adjacent district and the front is back where it started, with your push’s spent force now in the wrong place. The recovery for a reset is to identify the regrouping district and hit it with a nuke, because the reset is only as good as the district feeding it, and a bomb on the feeding district buys the front a turn to re-form with the reserve. The second common failure is the overcommit: the push commits the whole economy, and a bad first turn leaves the player with no economy left to re-form the front or to defend the rear. The recovery for an overcommit is to fall back to the nuclear posture on the fronts that broke and rebuild the ground front one district at a time, which is slower but is the only recovery available when the economy is gone. The third failure is the timing error: the push landed with the enemy’s full port behind it, and the push met a fresh wave and burned. The recovery for a timing error is to re-run the five questions, because the timing error means one of the questions came back yes when it should have come back no, and the fix is to find which question was wrong and fix the board so it comes back the right way next turn.

The meta-lesson of these recoveries is that a failed push is not a lost match as long as the posture that let you push is still intact, and the posture is the ground front plus the income control plus the nuke stockpile. If the push spends the ground front, the match is over; if the push only spends the wave, the front is still there and the next push can be re-run. So the recovery plan is always about protecting the posture, not about winning the specific push. The table below is the recovery cheat-sheet, and it is the thing you should have in front of you before the push is committed, because the recovery is only as good as the plan that was made when the front was still intact.

The player who plans the push to be spendable and the posture to be unspendable is the player who can take a failed push and come back with the next one, and that is the difference between a transition that is a strategy and a push that is a gamble. The recovery table is the cheapest insurance in the endgame, because it is written before the push, it costs nothing until the push fails, and it tells you exactly what to do in the two turns after the failure, which is the only window where a recovery is still possible.

Mode and map adjustments

The transition looks different in different modes and on different maps, and the same five questions read differently depending on which board you are playing. In team modes the win threshold is shared across the team, so the threshold test is not “how many tiles do I own” but “how many tiles does the team own,” and the push is committed when the team’s combined front can break, not when your individual front can. That means in a team mode the push is often committed earlier than the individual reading of the threshold suggests, because the team’s income is clustered and the team’s front is the thing that holds, and a single player’s ground advantage is worth more in the team context than it is alone. In FFA the threshold is per-player and the regrouping paths are multiple, because every other player is a potential reset, so the push has to be committed against the one player who is the real threat and the reserve has to cover the players who are not, which is a harder shape than the two-player endgame. The map size matters the same way: on a small map the Overtime window opens relatively early in tile terms and the income clusters fast, so the push is committed sooner and the conversion is cheaper, while on a large map the window opens later and the clusters are slower, so the nuclear posture holds longer and the push is a later, bigger commitment.

The map layout also decides the port path. On a map where the ports are clustered in one district, the push is a single direction, and the conversion is clean. On a map where the ports are spread around the edge, the push has to take them in sequence, and the conversion is a series of smaller pushes that each take one edge port, with the nukes holding the gap between them so the enemy cannot reset the edge you just took. The disciplined player reads the map before the endgame starts, knows where the ports are and how they cluster, and plans the push’s path on the map, not on the scoreboard. The player who does not read the map is the player who pushes into the wrong district first, spends the wave on a front that does not take the income, and has to re-run the whole conversion from a worse position. Reading the map is the cheapest part of the transition and the part that saves the most, because a push that takes the right ports first is a push that wins, and a push that takes the wrong ports first is a push that the enemy resets before it ever reaches the income that would have won it.

Where to go next

The nuke-to-ground transition sits between the nuclear decision pages and the endgame win-condition pages, and the two adjacent topics that complete the picture are the nuclear salvo sizing page and the post-nuclear port sustain page. The salvo sizing page answers the question of how much nuclear force to commit to a specific target, which is the input to the front test in the framework, because the front test needs to know whether a bomb will actually break the target it is aimed at or just reset the enemy’s clock. You can open the sizing decision at Nuclear Salvo Sizing, which walks through the Atom, Hydrogen, and MIRV trade-off and the blast-footprint math that tells you when a single weapon is enough and when a spread is the cheaper option. The post-nuclear port sustain page answers the question of what to do with the ports after a nuclear strike, which is the income test in the framework, because the income test needs to know whether the enemy’s refill is broken or still coming. You can follow that at Post-Nuclear Port Sustain, which covers how to hold a taken port against the enemy’s counter-push and how to convert a bombed port into a lasting income district rather than a temporary deletion.

Together, the three pages cover the full nuclear-to-ground arc: how to size the bomb, how to take the port, and when to stop nuking and start pushing, and reading them in that order is the path from a player who can win fights with bombs to a player who can win the match with the ground push that the bombs open. The which-nuke-commit page answers the related question of which specific weapon to use for a given target, which feeds the salvo sizing, and the nuclear-stalemate-breaker page answers the question of what to do when the nuke-only posture has already run out of value and the enemy is holding, which is the failure mode this page’s framework is designed to catch. The overtime win-condition pages answer the threshold test directly, because the falling Overtime threshold is the number the push is chasing, and the win-condition pages walk through how the threshold drops minute by minute and what a tile grind does to it as the match runs long. Working the three pages in order, salvo sizing, port sustain, and this transition, plus the overtime pages for the threshold math, is the complete endgame arc, and the player who has read them in order is the player who can run the transition without the scoreboard being the only thing that is going up. The arc is the cheapest endgame education there is, because every page in it is a decision you will face in the last twenty percent of the match, and the transition is the single decision that separates the players who sit on a high score from the players who actually take the map.