Strategy games present players with an evolving set of problems and a selection of tools with which to solve them. The problems in RTS tend to boil down to the fact that the enemy has units, so a lot of the tools offer ways to damage and destroy units. But using the same type of tool over and over can feel repetitive, so many games also give players ways to weaken enemy units without killing them. These non-lethal tools are known as status effects, or debuffs, and they add variety by letting players weave non-destructive actions into their battle plans.\n\nSome games do without status effects, but the goals of Zero-K demand a large toolbox, so ignoring status effects was not an option. The first status effect, known as either paralysis or stun, was present in our earlier ancestor, Total Annihilation , although it made little use of it. In fact TA had two sources of stun: the Spider and a global-range EMP missile. The Spider could disable a single target by continually firing at it, while the EMP missile disabled Core units in an area. This faction specificity was due to Core being a faction of robots, while Arm units are piloted, which made them immune to EMP. Core struck back though with the Arm-killing Neutron missile.\n\n\n\nModders went to town on TA and added a few more stun units - but haphazardly. I suspect paralysis was a niche mechanic since TA barely told players how the mechanic worked, it did not even indicate whether a unit was stunned or not. Spring (the engine behind Zero-K) did a lot better in this regard, but there was still not much more by the time Complete Annihilation entered the scene. Arm had gained a few EMP death explosions and some stun sidearms, while Core had a Gnat-like cheap stun drone. The faction-specific effect of the EMP and Neutron missiles were also removed at some point. From here we expanded the arsenal and fixed issues with the underlying systems.\n\nOne of the early goals of Complete Annihilation was to overhaul the weapon visuals, and part of this involved consistency between the visuals and the underlying mechanics. This means that dissimilar weapons should look different, and also that similar weapons should avoid implying a mechanical difference by arbitrarily varying their visuals. This was bad news for the plain-looking paralysis lasers, but the new idea also highlighted the fancy lightning weapons as a problem, since they were little more than pulse lasers mechanically. These problems solved each other when we gave all lightning weapons the ability to stun and reskinned the existing stun effects to shoot lightning.\n\nGiving EMP damage a unique weapon type was a start, but we also had to balance EMP around the removal of armour classes . Several units were resistant or immune to EMP damage, including commanders, units that deal EMP damage, and striders. Strider EMP immunity was perhaps the most egregious, since disabling a single huge unit is a great application of EMP. Players should be rewarded for playing around their tools to come up with something powerful, not butt up against arbitrary constraints. Most things can be balanced around, so removing something for being \"too OP\" is a last resort, as hidden restrictions punish unaware players and make them distrust the tools the game gave them in the process.\n\n\n\nLightning weapons stun units by dealing EMP damage that accumulates on their target. This alternate damage type sits alongside health and is displayed under the main health bar. A unit becomes stunned when the EMP damage it has taken exceeds its current health. Being stunned completely disables the unit, but the effect is temporary since EMP damage decays over time. The decay rate is Health/40 per second, which is why it takes 40 seconds for a unit to drop from 100% EMP to 0%. Fractions of current health are unwieldy though, so it is time to stun some Jacks.\n\nA Jack has 6000 health, which means it heals 150 EMP damage per second. A Jack with 6600 EMP damage takes four seconds to decay down to 6000, so 6600 EMP damage corresponds to four seconds of stun. Note that most of the EMP damage required to stun a unit is spent reaching the stun threashold, with only a small amount needed to push the stun duration higher. If EMP damage could rise indefinitely then most units would end up with very long stuns, which is why each weapon has a limit on how much EMP damage it can deal over the health of its target, called the stun timer. Knight has a stun timer of 1 second, so cannot send a full health Jack beyond 6150 EMP damage.\n\nIn fact, Knights cannot deal EMP damage to anything on full health, simply because they deal both regular and EMP damage, and the regular damage is applied first. The old lightning weapons retained some of their regular damage when they gained EMP. The difference is shown by colour, with blue lightning dealing both damage types, while yellow lightning is pure EMP. Venom, the descendant of the TA spider tank, eventually gained a small amount of regular damage, so has some blue lightning mixed in. The order in which the damages are applied is important due to the interaction between EMP and health.\n\n\n\nThe inbuilt EMP systems offered by Spring have issues. The default stun threshold is based on maximum health, not current health. This works fine, but making EMP and regular damage completely independent wastes the potential of the system. There is a sweet spot where the disparate systems of a game exist in a loose web of interaction, not so tightly bound that the synergies become obvious, but just enough for people to find creative combinations. So in 2009 the CA developer SirMaverick added a way to use current health as the stun threshold.\n\nUsing current health as the stun threshold works well until you consider what happens when stunned units lose health. Say we have a healthy Jack on 6600 EMP damage, which corresponds to a four second stun. Then some Glaives come in to kill the Jack before it recovered, but they need not worry, since damaging the Jack would cause it to be stunned for longer. Indeed, if the Jack goes down to 4500 health in those four seconds, then it would still have at least 6000 EMP damage, giving it a stun of at least 13 seconds. This was a problem: we wanted some synergy between EMP and regular damage, not for EMP to become redundant after the first shot.\n\nThe escalating stun duration problem was solved by scaling EMP damage to changes in health. Consider a full health Jack with a four second stun. If the Jack suddenly dropped to 4500 health, then its EMP damage would be rescaled from 6600 to 4950. Stun decays at 112.5 per second at 4500, so the Jack retains its four second stun. This is why mixed regular and EMP damage weapons deal their regular damage first: to avoid instantly scaling down their EMP damage. The result is a nice push-your-luck interaction between EMP and regular damage. Do you spend your stun at the start of the battle to gain an early advantage, or risk waiting for some damage to be dealt first to go for a longer stun?\n\n\n\nTechnically Zero-K uses maximum health as the stun threshold and applies a multiplier to incoming EMP damage, but the result is mathematically equivalent, and a shifting threshold is easier to think about than a dynamic multiplier. I only realised this after writing a gadget to re-implement EMP damage from scratch, but the work was not wasted since the gadget was later used to implement disarm. This was a new status effect that worked like EMP, except it allowed units to move and self-destruct.\n\nDisarm was added in 2013 after the switch to one faction loosened the Arm monopoly on lightning. The intent was to create a weaker version of stun for use in cases where a full stun was too powerful. It was initially used by Thunderbird and Racketeer to prevent them being otherwise nerfed into the ground. Disarm allows movement because we specifically wanted to nerf these two units against mobile targets, and movement is ubiquitous enough for disarm to be meaningfully distinct from EMP. Disarm also allows units to self-destruct, which requires a more complex explanation.\n\nConsider a version of disarm that disallows self-destruction, and imagine a crawling bomb under its effect. To deal damage, this bomb needs to run into the middle of the enemy force and collide with a fatal projectile, supplied either by the enemy or a nearby ally. This is fiddly, but it is even worse for the other side. Units would rather not explode, all else being equal, so need to avoid shooting at the bomb. However, if there is something even more valuable deeper in the army, then the bomb should be shot to protect it. Whether to shoot at a disarmed bomb can be a strategically important decision, but its implementation involves rapidly flicking state toggles or clicking on small, fast, units. This would entail a terrible fight with the UI , so we removed the choice by letting bombs blow themselves up. Now destroying a bomb only ever deprives it of options, so killing one is hardly counterproductive . This extra allowance is only required for disarmed bombs, since stunned bombs are much easier to avoid.\n\n\n\nDisarm fleshes out our roster of status effects but risks being redundant with EMP damage. This is always a risk for self-reinforcing mechanics when investment in one trades off against investment in the other. To solve this we made the two damage types interact by having disarm damage \"float\" on top of EMP damage. With this system, a unit is disarmed if the sum of its EMP and disarm damage exceeds its health. So a Jack with 4000 EMP and 2600 disarm damage would be disarmed for four seconds. Disarm also only decays when no EMP is present. Technically this system is equivalent to giving disarm to all sources of EMP damage, and making the two damage types independent otherwise.\n\nThe disarm and EMP weapons of Zero-K can be sorted into two types: one-shot abilities and sustained fire. The former include exploding bombs, one-shot missiles, and the lightning bomb, and last anywhere from 16 to 45 seconds. These are used to disable antinukes and to kick off short games of protect the stunned unit. I quite like how stunning something expensive creates a mini-game around a temporary objective, with the outcome determining whether the metal spent to stun the unit pays off. Some one-shot stuns can be reused, but with a long cooldown, and the units that wield such stuns often take fatal risks to apply them.\n\n\n\nSustained fire stuns and disarms are wielded by more ordinary units, and often deal regular damage as well. These stuns last for one or two seconds so tend to act more as hindrances than full disables. Focusing fire for a full disable is often possible, but at the cost of taking more damage from other sources. The low stun time and high reload time of sustained fire weapons means that locking down a target depends on avoiding overlapping stuns. Consider three Knights with a 1s stun timer and 2.2s reload time. If one of them fires every 0.73s, then the target will be stunned forever. However, if they all fire at once, it will enjoy 1.2s of freedom every reload cycle. The latter situation feels like fighting the UI , since it would not take much for the Knights to pause momentarily to stagger their shots.\n\nWasting stun time is a longstanding unit AI issue. Overkill prevention does a great job for Racketeer, but we only just started experimenting with more general solutions in the 1.13.9.0 update released just last week. Instead of adding more unit AI, the current approach is to design the stupidity out of the system by reducing the cost of overlapping stuns. The new mechanic, called overstun, allows units to deal slightly more damage than their stun timer would allow, but only if the target is already close to the stun timer. For example, Knight has a one second stun and one second of overstun, so it can add a full second of stun to anything with at most a second of stun duration. For example, a Knight shooting at a 6100 EMP damage Jack would be able to push the Jack to 6250 EMP damage. This is a buff to sustained fire stun weapons, but should take us to a better place than balancing around the stupidity or trying to resolve it with unit AI.\n\n\n\nFull stuns run the risk of frustrating players by removing control of their units. This is mitigated by making most stuns brief, precise, or partial. Slow damage plays a part in reducing the lockdown load of our status effects because it can never fully disable a unit. In fact, there are 12 sources of slow damage, not including commanders, compared to 11 sources of EMP and 4 sources of disarm. However, an exploration of slow will have to wait as this post is already bursting at the seams. I only just managed to squeeze in disarm because it works the same way as EMP. Slow damage is like EMP damage, except different in almost every way.\n\nIndex of Cold Takes