One of the main obstacles to getting better at ZK is being able to choose from the enormous number of possible construction options. This includes economic structures/units, mobile units, static defenses, and commander load-outs.
Not every option in the game is created equal. While most options have some sort of scenario in which they shine, the number of these scenarios and predictability/likelihood varies enormously. A generalizable unit can be made with a high certainty it will come in useful even if little information is known of what an opponent is doing. A specialized unit has a small number of scenarios in which is excels and thus requires information whether from scouting or inference to be effective.
Unit generality:
[Spoiler]
This is a common theme in many games, essentially there is a trade-off between generality and speciality. Most units that are specialised at something sacrifice their general power to do so. Examples in ZK include:
- All anti-air.
- All statics.
- All riots.
- All melee units.
- All support units (venom, placeholder etc.)
In fact, you can go so far as to say every unit in the game trades power for specialisation. Any unit that is both highly specialised while having few weaknesses is a strong contender for being blatantly OP, and as such, is selected against by developers who want to have a balanced game. But the way in which they trade raw power for specialisation varies, and so some units, despite being a specialised raider or what have you, are more applicable in the absence of information than others.
Examples of units that are highly general:
- Glaive
- Rapier
- Defender
- Solar
- Mex
- Constructor
- Scalpel
- Swift
- Wyvern
- Krow
- flea
You’ll note that every single one of these units can be countered very hard in some sense. Glaives are one shotted by all sorts of things. Defenders can be killed with ease by assaults or artillery. Scalpels die to bombers so quickly. This is not a list of units that are invincible or should be used in every scenario. But it is a list of units that are applicable in many scenarios.
I will give you some explanations of what makes these units so general as opposed to what makes other units specialist, despite their stats not immediately intuitively implying as much.
The glaive is basically a more specialised bandit, and yet, it has a higher general power level than bandit. Why? Well, they’re both in the raider role, and glaive is better at that role – while bandit sacrifices some speed and firepower for range and durability, glaives are balls-to-the-wall glass cannon raiders, having many more counters than bandit. So how could it make any sense that glaive is a more generalist unit? Well, the role of raider specifically impacts on the game in a very consistent and generalizable way. Glaive is specialised at being generalizable!
They slow down your opponent’s expansion by threatening every front. They force expenditure both in reducing the speed of opponent’s economic growth, by actual raiding, by forcing non-raider counters that pose no threat to your own economy, and by being fast enough to plug any holes in your defense if ever you need to. I trust you’re familiar with the comicbook logic cliché whereby a timetraveler goes back in time to kill someone’s parents before they were born, which is a way to kill even a god. Well, that’s what glaives do. They kill your parent metal before you’re born. Glaives beat even their hardest counters in this way, and are thus extremely general units. While bandits can do this, they aren’t as good at it.
An example of a non-generalist unit is a mace. Doesn’t sound intuitive does it? Mace has decent range (enough to often hit gunships), can move on water, has intense dps that is highly precise and staggered, and thus can break both mobs of small quick units and single targets alike. So it basically counters anything it can shoot at. How is this not highly general?! Well, it creates very little pressure on the opponent to slow his expansion. It’s slow, and while it has good stats it can’t go toe-to-toe with statics for cost. It has good range for a riot but it’s still not enough to ever shoot a skirm. If a mace attacks you, it will kill a scout (say, one glaive for 65 metal), then it will head to your base and because you know it was coming you’ve got skirms or statics ready. Let’s reiterate:
1) Opponent makes mace. You build one glaive and spend the rest on eco.
2) Opponent kills your glaive. You react to incoming mace by erecting defences.
3) You opponent cannot kill you, and even if they don’t lose the mace you’re up in metal (far more than 65). If they lose the mace they lose completely.
So while the mace can kill glaives all day, the glaive enables counterplay to the mace, and thus the actual glaive itself is only a small component of the net increase in your overall position, which is the true measure of a units overall power. Despite its ‘narrow’ role and stat block, the glaive makes your play super adaptive. While the stats of the mace and its intended victims are quite general, they force you to commit to a particular playstyle which in turn makes it specialised in making you commit to being unadaptive.
So what stats predict the general ability of a unit? Well… Everything.
[Spoiler]
- Cost
- Speed
- Terrain type
- Dps
- Hp
- What it can hit
- What can hit it
Cost is probably the most direct correlate. Small cost allows a unit to impact on the game much earlier, which allows more opportunity for limiting opposing economy, but also allows you to flexibly make it without committing to certain play-styles.
Speed and terrain type are also a big deal. Mobility literally allows you to only fight the things that you want to fight (with a few exceptions). It doesn’t matter if warrior can kill scorchers for cost if they cannot ever fight them. It also allows you to initiate the fights that you want. It matters a huge deal that scorchers are able to catch fight unguarded mex. Terrain type is essentially a map specific modifier to speed.
DPS matters, or more importantly, the ‘do’ of a unit. Fleas are ultra-mobile and ultra-cheap, but if they didn’t have any means of doing damage their application would be limited to scouting uncontested areas. Mex don’t do damage, but they generate metal wherever they are. Damage and resource generation are two things a units can do. If something does nothing, its application is literally zero and therefore it’s not adaptive in any scenario. It’s worth noting that damage mitigation is a form of ‘do’, but a limited one.
HP represents a large factor in whether the thing a unit does is doable in contested territory, which is a very good place to want to do things! Unless you have no intention of having a unit in harm’s way, reliability of a unit is strongly tied to hp.
What can a unit hit? Many units can’t hit both air and ground, or units that are moving quickly, or anything that’s more that’s 300 elmos away. This is a large limiter of the scope of what your unit can do.
What can hit your unit? It’s difficult to counter something you can’t hit. The number of unit that can effectively target your unit is akin to how often the situation arises in which your unit is invincible. Does it cloak? Does it have a shield? Does it have so much HP that burster units are impotent? Does is fly? Does it walk underneath the range of AA? These are all situations that greatly improve the efficiency of a unit if it happens that the opponent doesn’t have the right answers. While this sounds like a kind of specialisation (and it is), negating a particular element of your opponents force is crucial to almost any engagement. These attributes scale with mobility, because mobility allows you to pick the engagements in which these attributes will work in your favour.
You might notice that no units ticks every single box here. Rapier is the only unit that comes close, but it’s arguably not cheap, and its ‘do’ is limited by the nature of its damage delivery – it is a burster which struggles to take engagements that are in any way prolonged (most of them). It’s a duck taken flight that’s given a slow sidearm.
What does the monstrosity that is this so far 1400 word intro mean, and what does it want? Well, given the unintuitive nature of unit generality, I thought it would be a good idea to have a voluntary unlock scheme for players that want to interact with the more fundamental elements of the game: the units that shape what the meta is. This list will simplify the game down to the bare bones, so that you can build it again from the ground up. All the red-herrings that are overly specific will be thrown to the side so that you can introduce them once the basics are mastered. This guide is designed both for noobs and for players who are experienced but who have not grown as players in a long time.
[Spoiler]
Eco:
Level one: Metal extractors and solars. Can make tidals when land for solars is not available.
Level two: Wind gens are allowed if the elevation is over 200, or if they are needed to connect grids across rough terrain where solars will not go. Can make tidals whenever you like.
Level three: Can make geos (not mohos), and wind wherever you like. Can make caretakers.
Level four: Can make fusion
Level Five: can make singularity and moho. Can use pylons. Can use storage.
Statics:
Level one: Can make radar, lotus and defender. Can make razor if opponent is definitely air. Can make urchin on water maps.
Level two: Can make stardust and skydust.
Level three: Can make stinger and sneakypete/eraser.
Level four: Can make gauss if artillery is known to exist. Can make faraday around critical structures or anywhere that there is at least 300 metal’s worth of other firepower. Can make hacksaw when worried about immediate bomb threats, and cobra when you need a no-fly zone. Can make DDM if grid allows, but only if solars (or pylons if allowed) are linking it – if there is an energy structure providing 25 energy or more it must be quite far back. Can make protector if nuke is known to exist. Can make aegis if you’re unit balling or protecting critical structures from tacnuke etc.
Level five: Can make newton, annihilator, chainsaw, screamer, behemoth. Can make advanced radar and sonar. Can make tacnukes silo.
Factories:
Level one: cloaky, amph, light vehicle
Level two: shield, hover
Level three: planes, gunships, shipyard, spiders
Level four: heavy tanks, jumpjet
Level five: Athena, strider hub
Units:
Level one: raiders, land workers, swifts
Level two: skirmishers and assaults, and mobile anti-air. Can make cranes if
air superiority is secured. Vultures.
Level three: artillery, thunderbird, wyvern. Support units such as placeholder, racketeer, and venom. Dirtbag.
Level four: Riots, crawling bombs. Infiltrator. Raven, phoenix. Blackdawn. Felon. Air re-arm pad.
Level five: djinn, striders, surfboards, Valkyries, vindicators, gnats, blastwings. Hawk. Krow. Skuttle.
Endgame:
Level one: sharp-shooter, reaper
Level two: ultimatum, grizzly, wyvern, crabe, goliath, Reef
Level three: dante, krow, silencer, big bertha, funnelweb, catapult, warlord
Level four: Scorpion, bantha, sumo, tremor, tacnukes, leviathan
Level five: detriment, DRP, Zenith, starlight
Commanders:
Level one: riot cannon on strike or battle, HMG+ablative plating on recon
Level two: Nanolathes, radar module
Level three: Lazarus device, range module, siege com, SLAM, drones
Level four: Everything not mentioned in level five
Level five: slow beam, lightning gun, radar blocker
Commander unlocks will need updating later when I can be arsed :P
So this guide serves a few purposes:
- to break the feedback loop of poor unit choices exhibited by low elo players
- to offer a roadmap to improvement
- to clarify the distinction between a unit that can be made when you know little about what an opponent is doing, and one that is used to counter a specific thing.
It is not:
- me telling you how to play (this is a recommendation)
- me telling you what is actually better in any specific scenario
Start at level one in every category unless you consider that category mastered. Every ten games you win you go up a level in a category of your choice. This increment is arbitrary and you can choose whatever you want really.
Edit: I've got to go and it's likely that this will need refining, but I thought I may as well post it and see if there's any interest first.
