How Nations Aim Their Nukes: Smarter Targeting, No Wasted Blasts

Direct answer. The Nations are getting a smarter nuclear brain: before they fire, a Nation now works out what share of the blast would actually hit enemy land, and it rejects any target below a hard floor (atom 25%, hydrogen 40%). It also refuses to fire at a spot where another player’s nuke is already on the way down, and it will not choose a target whose blast would clip your or a teammate’s land. The short version for your decisions: Nations stop wasting nukes on water or on a spot that already has a bomb coming, so the two old exploits that let you bait them or hide behind water no longer work at Medium and above. You now have to deny their best target rather than rely on the AI making an inefficient shot. Everything here is an unreleased main-branch behavior change, so treat the exact numbers as the current upstream state until the change ships in a release.

The reason this matters is that the old targeting was exploitable in two specific ways, and both are now closed. A Nation used to be able to double-nuke the same tile — drop two bombs on the exact same city cluster and let the second one re-hit land the first had already scorch — and it could commit a whole bomb to a coastal target where most of the footprint was ocean. Both behaviors made the nuclear threat feel random and made some maps feel like they could not be defended. The new logic removes that randomness: the Nation now measures the target and only fires when the shot is efficient. That is a fundamentally different opponent to plan against, and it changes when you nuke, where you spread your cities, and what you do at the shoreline. Read the rest of this guide as the exact numbers and the decision framework that turns the new targeting into something you can beat, not just something you have to survive.

There are four specific decision moments where the new targeting changes what you do: the moment you fire a nuke yourself, because now you can predict whether the Nations will answer it or hold theirs; the moment you place a City or Port, because density is now the variable the Nations score against; the moment you disperse or concentrate your economy, because the anti-double-nuke rule changes how much a single cluster is worth; and the moment you pick a lobby, because map water now tilts the nuclear exchange toward the defender. Those four moments are the whole guide, and the sections below walk through the numbers behind each of them so you can check the position before you build or fire rather than after the blast has landed.

What actually changed and where it applies

The commit that drives this guide rewrites how a Nation scores a nuke candidate. It touches the candidate-scoring path in NationNukeBehavior.ts, and there are four linked effects, all sitting behind the same difficulty gate: above Easy, a Nation skips a candidate where another player’s nuke is already landing, it computes the blast coverage, it rejects a candidate below the weapon’s minimum, and it then filters the final winner if that blast would touch friendly land. Easy Nations get none of these filters. The gate is the single most important boundary in the whole change, so before anything else: if you play single-player or private lobbies on Easy, the Nations keep the old, less careful behavior, and everything in this guide about the coverage gate and the anti-double-nuke rule does not apply to you. The new targeting is live on Medium, Hard and Impossible, which is where the Nations are the main nuclear threat and where the change is felt.

Why the gate is there is worth stating plainly, because it is a deliberate design choice and not a bug. The Nations are the AI opponent, and the developers wanted them to stop looking stupid at high difficulty. A Nation that wastes a 100-tile hydrogen blast on the middle of the ocean, or that drops two atoms on the same city, reads as broken to a player. The coverage gate and the double-nuke check make the AI’s nuclear behavior look competent, which is exactly the point of the higher difficulty tiers. The side effect is that the AI is now a better nuclear opponent, so the difficulty curve is steeper precisely on the thing that used to be exploitable. That is the whole reason this change is a decision-relevant one and not a cosmetic one: it removes the two specific openings the community had been using, and it does so at the difficulty levels where those openings actually mattered. If you have been beating Hard or Impossible by baiting the Nations into inefficient shots, or by stacking a single dense cluster and hoping the Nations would not double-hit it, the next section shows exactly which of those plays the new code kills. The two plays it kills are the shoreline bait, where you relied on the Nations wasting a hydrogen over water, and the double-stack, where you relied on the Nations not re-striking a tile that already had a bomb coming. Both of those relied on the AI making a specific mistake, and the new code removes the mistake rather than changing the difficulty, which is a cleaner fix for the developers but a harder problem for you, because you can no longer plan around a flaw that no longer exists.

The coverage gate: atom 25%, hydrogen 40%

The coverage gate is the core of the change, so it is worth the full math. When a Nation considers a candidate tile, it grid-samples a disk around that tile: it steps through the blast area in increments (the step is the outer radius divided by ten, at least one), and for every sample point it checks who owns that tile. Water, fallout and unowned land all count as nothing; only tiles owned by a non-friendly player count as enemy land. The inner disk — the region the weapon fully destroys, the inner radius — is weighted double, because the outer ring is only partially hit while the inner ring is total loss. The result is a ratio: the share of the blast’s sampled area that is enemy land. That ratio is compared to the weapon’s floor, and if it is below the floor the candidate is skipped entirely. The floors are an atom bomb at 25% and a hydrogen bomb at 40%, stored as 0.25 and 0.40 in the source.

The hydrogen’s higher bar is not an accident and it is the number you should internalize. A hydrogen blast is enormous, inner 80 and outer 100, so a 40% enemy-land share over a 100-tile radius is a much stricter test than an atom’s 25% over a 30-tile radius. In practice this means the hydrogen is far harder to deny with water or fallout than the atom is. If you can keep a Nation’s prospective hydrogen target below 40% enemy land, the Nation simply will not commit that bomb and will move on to a different candidate or wait. The atom is the more forgiving weapon, and the one the Nations will lean on when the hydrogen’s gate cannot be met. Two properties make this a live system rather than a fixed table. First, the coverage number is recomputed on every decision tick, as the map changes around the Nation: your border moves, water becomes fallout after a blast, a city is destroyed, a new port is built. The same shoreline target can be below the gate one tick and clear it the next, so you are managing a moving number, not a static line. Second, because water and fallout count as zero coverage, a map that is mostly water makes it structurally harder for the Nations to clear the floor. That is the reason island and archipelago lobbies get a real defensive edge from this change: the Nations’ own math works against them on water.

The final row matters because the MIRV warheads are what the Nation drops once it commits a MIRV, and each warhead inherits the coverage logic of its parent target. The coverage bonus column is the scoring term, covered in the next section: the weapon that lands on more enemy land is preferred, even against a nominally more valuable structure, and that is the lever you can actually pull.

The new anti-double-nuke and friendly-fire rules

The second half of the change is about where the Nations will and will not fire, and it has two parts that are easy to conflate. The first is the anti-double-nuke rule. A Nation now checks every nuke currently in flight — an atom, a hydrogen or a MIRV warhead from any player, not just a teammate’s — and if any of them is already aimed at a tile close enough to the candidate, the candidate is skipped. “Close enough” is defined by the inner radii: the game adds your weapon’s inner radius to the incoming nuke’s inner radius, and if the distance between the candidate and that nuke’s target is less than that sum, the candidate is treated as already covered and dropped. So an atom (inner 12) will not target a tile within 24 of another atom, or within 92 of an incoming hydrogen. This is the generalization the old code did not have: the previous behavior only avoided double-striking your own team’s nukes in team modes, but the new one covers any player’s nuke anywhere on the map, including your own. The practical effect is that a Nation will not stack a second bomb on a tile that already has a bomb coming, so the old “double-nuke the same city” pattern is gone at Medium and above.

The second part is the friendly-land filter, and it is the one that answers the “allies nuking allies” complaints. After the Nation scores all of its candidates and picks the winner by highest score, it runs one final check: if that winning tile’s blast would touch any land owned by the Nation itself, by a teammate, or by an ally (including an ally who is currently AFK), the candidate is rejected and the next-best clean tile is used instead. This filter looks at the full blast, not just the inner disk, which is a stricter test than the coverage gate. The reason it exists is the player experience: the worst nuclear moments are not the ones where a Nation nukes you efficiently, they are the ones where a nuke meant for an enemy lands on your port because the blast clipped your coast. By making the Nations refuse to pick a target that hurts their own side, the change removes the accidental friendly-fire grief that players have been reporting in team games. For you, this means the Nations are now safe to ally with on the nuclear front in a way they were not before: a Nation will not nuke through a spot that would damage its own allies, so your alliance positions become more predictable and the “who is going to get caught in the blast” problem largely goes away.

What the scoring means for you

The coverage gate decides whether a target is even allowed, but the scoring decides which allowed target the Nation actually picks, and the new code makes coverage a huge part of that score. The base tile score already rewards structures: Cities, Ports and defense posts are worth more than empty land, and higher-level structures are worth even more. The new term added on top is coverage times 50,000, so a target whose blast is 70% enemy land gets a 35,000 bonus, while a target that is only 30% enemy land gets a 15,000 bonus. That is a 20,000-point swing purely from the geometry of the blast, and it can outweigh the difference between two structures. The consequence is that the Nations will now prefer a target where their blast lands on a lot of enemy land over a target that is a more valuable structure but sits over water or fallout, even if the structure is nominally more worth hitting. This is the lever you can pull: the Nation’s target is the tile that maximizes structure value plus blast coverage, and if you can change either of those two numbers, you change which tile wins.

There is also a difficulty layer on top of the scoring that changes what the Nations will commit at all. On Impossible, the Nations only fire at targets with positive structure value, meaning they will not nuke an empty frontier or a bare coast, and they also avoid any trajectory that an enemy SAM would intercept. On Hard and Impossible together, the SAM-avoidance applies, so a well-placed SAM ring can make a target un-launchable even if its coverage is perfect. This is the interaction to understand when you are defending: the SAMs and the coverage gate are two independent denials, and a target that fails either one is not nuked. The scoring also has a crown-awareness term, where if the top player holds more than half the map the Nations will concentrate on the crown holder, so in a late game the Nations effectively become a nuclear ally against whoever is winning. All of this means the Nation’s nuke is not a random shot at your biggest city; it is a computed choice that you can influence by changing structure density, blast geometry and SAM coverage. The next two sections put this into concrete scenarios, because the numbers only become decisions when you can see a specific map and a specific position.

Scenario A: defending a mainland coastline on Medium

Imagine a Medium FFA on a map with a fair amount of water. A Nation has a silo near your mainland border and has been quietly accumulating. You have a coastal city cluster: three Cities and a Port built within a short distance of each other right at the shoreline, and your inland cities are spread out. The Nation reaches for a nuke and starts scoring candidates. Under the old behavior, a hydrogen aimed at your coastal cluster could have fired even though most of the 100-tile blast was ocean, and a second atom could have re-struck the same cluster while the first was still on its way down. That is the “wasted blast” and the “double nuke” the change is specifically killing.

Under the new behavior, here is the decision you should be making. The Nation computes the coverage of the hydrogen on your coastal cluster. Because the cluster sits at the waterline, a large share of the 100-tile disk is ocean, and the coverage is probably well below 40%. The Nation skips that target. It then evaluates the interior candidates, where your inland cities are spread out. If those inland targets read above 40% enemy land, the Nation fires there instead, and you now have to answer a hydrogen inland where you have your regrow and your Cities. The decision the coverage gate forces: you either deny the hydrogen by keeping your coastal cluster below 40% enemy-land coverage, or you accept that the hydrogen will hit an inland cluster and you defend there. The concrete move is to space your cities so the shoreline cluster reads as low coverage — spread the coastal cities out, keep the Port at the water but the Cities a bit back — so the Nation’s hydrogen is denied at the coast and the Nation is forced inland, where you can choose where it lands by where your Cities and defense posts are. If you stack everything at the shoreline, you are handing the Nation a target that fails the gate, which sounds like a defense but is actually a trap: you are denying the hydrogen but you are also telling the Nation exactly which inland cluster to hit instead. The coverage gate turns your shoreline from a weakness into a dial you control, and the best position is the one where the Nations’ hydrogen has no coastal target that clears 40%. The trade-off is that the coverage gate does not help your inland positions at all, so while you are denying the coastal hydrogen you are also telling the Nations exactly which inland cluster to commit to next, and the defense only works if your inland clusters are also spaced below the atom’s 25% floor.

Scenario B: a MIRV barrage and the anti-double-nuke rule

Now a second scenario, one that shows the anti-double-nuke rule in action. It is an Impossible FFA on a world map, and a Nation has committed a MIRV at you. The MIRV splits into multiple warheads, and under the old behavior the Nations could concentrate several warheads on your single most valuable city cluster, letting the second and third warheads re-hit land the first had already scorch. Your dense cluster was a double-saturated target. Under the new behavior, each warhead is a separate nuke with its own inner radius (12), and the anti-double-nuke check means the Nation will not aim a second warhead at a tile that already has a warhead coming within the sum of the inner radii (24 for two MIRV warheads). So the MIRV warheads spread out across your positions rather than stacking on one spot.

The decision this creates is about how you arrange your cities, and it is the opposite of what you might expect. You do not want one dense cluster that reads as 90% enemy-land coverage within 100 tiles, because that is the single target every warhead and every other Nation will score highest and hit first. You want your economy spread across several medium-density clusters, each one reading below the point where it is the obvious winner. The anti-double-nuke rule is what makes this work: because the Nation will not double-hit a single tile, it has to spread its warheads, and if your clusters are spread too, the Nation’s barrage is diluted across many targets instead of concentrated on one. The concrete decision: keep no single cluster above roughly 50% enemy-land coverage within a hydrogen’s 100-tile reach, and put your Ports and your best Cities in at least two separate medium-density groups so no one target wins the scoreboard. This is a real change to how you build on water-heavy or MIRV-heavy maps, and it is the single most useful takeaway in this guide for a high-difficulty FFA. The Nations now see your density, and density is what they target, so the player who spreads their economic weight is the one who survives the barrage. The concrete test is simple: if any one of your clusters would be the obvious first target a Nation picks, you are one MIRV away from losing your whole economy, and the fix is to break that cluster into two smaller ones that are each worth less and neither of which wins the scoreboard.

The failure mode and the counter

The failure mode this change creates is specific, and it is the one you should watch for. The old failure was that the Nations wasted nukes, which you could exploit by luring them into a bad shot. The new failure is that you bait a target and the Nations simply do not fire, and you mistake their non-firing for a win. Here is the concrete trap: you deliberately concentrate a lot of your structure at the shoreline, hoping the Nation’s hydrogen will fire at a coastal target where the blast is mostly water, and you count on the bomb being wasted. Under the old behavior that sometimes worked. Under the new behavior it does not work the way you hope, because the Nation now sees the low coverage and skips that target entirely. The bomb is not wasted — it is redirected to your next-best target, which is your inland cluster, and that is the target you wanted to keep safe. So the bait backfires: you spent real structure density to create a fake target, and the Nation moved on to the real one. The counter is to understand that you cannot waste a Nation’s nuke the way you used to, and to stop spending density on decoy positions that will not hold.

The real counter, and the one that is still available, is to change which target is the best one, not to hide targets. Because the Nation picks the tile that maximizes structure value plus blast coverage, you win by making your best target a worse target for them, which is almost always by dispersing. The two moves that still work: first, disperse your Cities and Ports so no single tile reads as high coverage and high value at the same time, which lowers the score of every candidate and forces the Nation to spread or wait. Second, use the SAM ring on Hard and Impossible, because the SAM-avoidance is a separate gate that can make a target un-launchable even if its coverage is perfect. The combination is the whole defense: disperse to lower the score, ring the remaining high-value clusters with SAMs to deny the launch, and accept that the Nations will now hit your best cluster, so you make that cluster survivable rather than trying to make it invisible. If you are losing on the new targeting, the almost-certain cause is that you are still playing the old game — concentrating and baiting — against a Nation that no longer makes the mistake you are baiting for.

There is one more failure mode worth naming, because it is the subtlest: the failure of treating the Nations as a fixed enemy you can pre-solve. The old AI was a constant, so a player could memorize its habits and run the same build every game. The new AI is a function of the live map, recomputed every tick, so the same build that worked last game is the wrong build this game. The fix is to stop building a single “meta build” and instead build a small decision procedure you run in every position: compute which of your clusters the Nations would score highest, change that one, and repeat. It is slower to think about and much harder for the Nations to solve, because it turns your defense into the same optimization problem the Nations are running and lets you win it. The players who lose on the new targeting are the ones still running the memorized build against a target that is no longer fixed.

Mode and map adjustments

The change interacts with mode and map in four specific ways, and each one shifts the decision. First, by difficulty: on Easy the Nations ignore the coverage gate and the anti-double-nuke check entirely, so the old meta — including baiting and stacking a single cluster — still works there, and you can use Easy to practice the building patterns without the Nations punishing you. On Medium and above, the baiting is dead and the dispersal is the only reliable defense, so the difficulty setting is not just a strength dial but a change in which tactics are legal. Second, by map water: the more water on the map, the bigger the defensive edge the change gives you, because water and fallout count as zero coverage and make it harder for the Nations to clear their floors. Island and archipelago lobbies are now genuinely easier to defend against Nations than mainland-heavy lobbies, and that is a real shift in map selection if you play Nations on a specific map. Third, by game type: in team games, the friendly-land filter is the big change, because the Nations will not nuke a spot that would hit a teammate, so your alliance positions become safe from accidental friendly fire and you can build near a Nation’s border without the blast-risk fear. In FFA, there are no allies so the friendly-land filter does nothing for you, and the whole defense is the coverage gate and the SAM ring. Fourth, by endgame: on Impossible the Nations only nuke positive-value targets and avoid SAM-interceptable trajectories, so late-game the Nations become a disciplined nuclear force that you have to out-position rather than out-bait.

The map adjustment to internalize is the water advantage. If you are choosing a lobby and you know the Nations will be the nuclear threat, a water-heavy map is the better defensive choice, because the Nations’ own coverage math works against them there. If you are attacking a Nation, a water-heavy map is the worse choice for you, because the Nations will refuse to hit your water-adjacent positions and will instead concentrate on your inland clusters, where you have less room to disperse. The two-way implication is the same: water is now a defensive asset against Nations and a liability when you are the one trying to force a nuclear exchange. The decision to make when you pick a map against Nations is therefore whether you want to be the defender with the water edge or the attacker fighting across it, and the answer determines which map you should queue for. This is a small but real change to map selection, and it is one of the few places where the new targeting directly changes which lobby you want to be in before the first tick.

Putting it together: the decision framework

The whole change reduces to four decision rules you can check before every nuke and every build. Rule one, the coverage floor: before you commit a hydrogen, ask whether the Nation you face can clear 40% enemy-land coverage on your best cluster, because if it can, the hydrogen is coming and you should defend the target it will actually hit, not the decoy. Before an atom, the floor is 25%, so the atom is the one to expect on any map where the hydrogen is denied. Rule two, the anti-double-nuke: do not assume a Nation will stack a second bomb on a tile that already has one coming, because the new code stops that, so a dense single target is no longer a double-saturated target and your defense should spread rather than rely on the AI not double-hitting. Rule three, the friendly-land filter: in team games you can build near a Nation’s border with less fear of the blast clipping you, because the Nation will not pick a target that hurts its own side, so alliance positions are safer than they were. Rule four, the scoring: the Nation picks structure value plus coverage, so you win by lowering the score of your best target, which is almost always by dispersing your Cities and Ports and by ring SAMing the clusters you cannot afford to lose.

The single biggest decision this change forces is a change in how you think about density. The old advice was to concentrate your economy and let the Nations’ random nuclear shots be your problem. The new advice is that density is exactly what the Nations target, and the player who spreads their economic weight across several medium-density clusters is the player who survives the nuclear barrage, because the anti-double-nuke rule and the coverage gate both reward dispersion and punish concentration. The second biggest decision is the water choice: if you can pick the map, pick water if you are defending against Nations, because the coverage gate makes water a real defensive asset, and pick land if you are attacking, because the Nations will refuse to hit your water-adjacent positions. The third is the SAM ring on Hard and Impossible, where the SAM-avoidance is a separate gate that can deny the launch entirely, so the best position is a dispersed economy with a SAM ring around your most valuable cluster. If you internalize those four rules, the Nations’ smarter nukes become something you can plan against, and the guide is done. For the adjacent topics, the how Nations put you under pressure guide covers the troop side of the same difficulty, and the how to break a nuclear stalemate guide covers the offensive nuclear decision that this guide is the defense to.