Zero-K Patch Notes — June 7, 2026
Aggregated from Steam, cross-tracked with Battle.net coverage on GamePatchNote.
The basic reclaim mechanic is as follows: units leave wrecks when they die, and these wrecks can be harvested by constructors for extra metal. This sounds like a simple enough mechanic, but the more I think about it, the more I feel like Zero-K would not work without it. Reclaim is common to most games inspired by Total Annihilation, although its importance varies depending on how lucrative it is. As always, I am only intimately familiar with the design of one of these games, so while the ideas that follow are likely to be widely applicable, I focus on how and why reclaim works the way it does in Zero-K.
The main role of reclaim is to dynamically create areas of interest around the map. Some units die in a skirmish, leaving a few hundred metal on the ground, which makes holding the area slightly more enticing. This leads to more battles that drop even more wreckage, which can even snowball into thousands of metal worth of incentive to keep fighting in that area. Eventually the feedback loop is broken by one side gaining enough control to deplete the whole field. The pacing of this dynamic is great: the stakes keep rising while the outcome is uncertain, then quickly dissipate once one side wins, releasing the tension.
A single wreck field, that everyone pours their armies into, would get a bit dull after a while. Luckily, there are often multiple fields scattered around the map. Reclaim essentially generates multiple shifting king-of-the-hill style objectives, but in a natural way that targets the most contentious parts of the map. Players shift their focus around to secure wrecks, which encourages the use of mobile armies that can make these kinds of strategic moves. The side that focuses more units on an area has a better chance at winning, but leaves themselves open to attacks elsewhere, which can trigger the creation of a new field of wrecks to fight over. Sometimes one side wins on all fronts simultaneously, but when this happens the reclaim lets them close out the game quickly, which is preferable to keeping the game going long after it is over.
Contestable wreck fields only form if players have a reason to be out on the map. Metal spots fulfil this role generally, but many maps also kickstart the process by starting with reclaimable map features. The classics are rocks and trees, but sometimes a map will contain old machinery or magical purple crystals. Trees tend to be energy-only, which can lead to constructors being more efficient than Solar Collectors early on, while rocks contain metal. A lot can be done with small amounts of reclaim, as every bit counts early on.
In addition to changing the value of an area, wrecks change the terrain itself. Most wrecks block movement and projectiles, to the point that an open plain can be transformed into a maze of passages and cover. Large units can crush smaller wrecks, at the cost of reducing the amount of metal that can be reclaimed, and not all wreckage blocks. Hovercraft and ship wrecks do not block movement because wrecks on the sea floor used to block surface movement. A smattering of light units also drops wrecks that do not block movement, presumably because the wrecks are so small that not being able to walk over them would look silly.
It can be tempting to think of reclaim as a comeback mechanism, and it is in some situations, but it is more useful to think of it as a mechanic that increases variance. It can certainly provide the metal needed to rebuild after a raid on your base. However, if the enemy knew that you would come out ahead, then they would not waste their units on diving your base in the first place. So reclaim softens the blow of effective assaults and deters marginal ones, but a similar effect could be achieved by buffing defences. It is hard to call this a particularly strong comeback mechanism, in theory at least, because players should avoid making raids that just feed metal. In practice mistakes can be made, but the advantaged team is not obliged to make them at any greater rate than an underdog.
Things get more interesting when we consider wrecks that have a reasonable chance of being claimed by either side. The long-term success of an assault often depends on who gets to reclaim the resulting wrecks, which introduces swinginess by widening the range of potential outcomes. So even if a player attacks with the expectation of a positive outcome, there is often a good chance that the defender will rally and take the reclaim for themselves, or at least take a sizeable chunk. This is the stronger form of comeback mechanism offered by reclaim, as uncertainty benefits the underdog. It has to, because if the game were to just plod along predictably, we would expect the disadvantaged team to lose. Base raids have less of this type of variance than fights out on the map because the defender is destined to get the reclaim.
The importance of reclaim depends on how much metal can be gathered from wrecks. Too much metal and the first bit of reclaim will snowball into a win, too little and the system becomes useless. In Zero-K units leave wrecks with 40% of their value when they die, unless they died to a massive burst of damage, in which case they leave 20% of their value. This value was picked very early on in Complete Annihilation, so perhaps the precise value is unimportant, although I suspect there is some leeway offered by tweaking the design of the rest of the game around a simple round number.
The mechanics of reclaim are also important, although I might not think so if we did not start with quite a poor implementation. The default system offered by the engine has reclaimers fill a progress bar over their target wreck, which yields a lump of metal when filled. This was an ugly system for many reasons, as lump sums require extra storage to process, and being hit by chunks of metal makes your economy needlessly difficult to manage. It gets worse when you consider the arguments and jank caused by multiple players trying to reclaim the same large wreck, as only the constructor that contributed the final tick of progress was rewarded.
Reclaim in Zero-K works by constructors gradually sucking the metal from wrecks. All constructors can reclaim, and they all do so at a rate equal to their build power, as any other choice would be arbitrary. This refusal to apply an arbitrary multiplier makes constructors quite bad at reclaiming, when you consider how long it takes to reclaim thousands of metal with five build power. The upshot is that reclaim can be taken during a battle, but that doing so at any significant speed requires a lot of investment, which increases your risk of losing the battle. A popular technique is to bury construction towers, or to put them behind walls, which works nicely as a way to encourage players to leave their mark on the map with terraform.
Reclaim has another whole set of mechanics for reclaiming live units, and in this case the default offered by the engine is even worse. The old live reclaim system essentially armed every constructor with a reverse repair weapon, one where "killing" the unit resulted in a lump sum payment. The lump sum was bad enough, but the worse part was that it could target enemies. Units would not automatically reclaim enemies, so this ability was hidden behind a fight with the UI. We could have automated offensive reclaim, but instead decided that arming every constructor was too much of a nerf to raiders, so made (fully built) enemies unreclaimable.
The old live reclaim system had another issue: it could be used to avoid explosions. Units killed by reclaim skipped their death explosion, which meant that a Singularity Reactor could be nestled in your base, provided there were enough constructors around to safely dismantle it when it came under fire. This was a neat trick, but would remove the point of explosions if automated, so the CA developer det added the reverse-build live unit reclaim system. This method of reclaim turns units back into nanoframes, at which point the unit can be un-built for an 80% metal return. The final step was to make nanoframes explode when over 80% completion, which was easy enough to do with a gadget.
I shouldn't complain too much about the defaults offered by the engine, as they were mostly inherited from Total Annihilation, and the engine was easy enough to extend. Some improvements even predate Complete Annihilation, such as area commands. Wrecks are much harder to reclaim without area commands, as the options are to either issue a patrol command, or to click on each wreck individually. Patrol was risky though, as constructors might decide to repair instead, or wander into turret range. Area reclaim gives you much more control over how far your constructors venture, and there are even keyboard modifiers to make them ignore trees and other metal-free reclaimables.
Precise control of your reclaiming constructors is vital because players are encouraged to put their constructors on the front line, reclaiming at the edges of the battle. This is great in that it gives the enemy targets of opportunity to snipe, but losing constructors this way would be frustrating if they were not where you meant them to be. Constructors also have other frontline tasks, including repairing, construction, and some can even resurrect units from wrecks. A large part of the game is judging how many constructors to build, and where to deploy them, but that will have to be a topic for another time.