Air v. AA v. Ground balance

This problem of balancing air units v. AA v. ground units is a problem for every combined arm game because of the following:

1. Modern air power is simply the most powerful military branch for its cost. It can deliver precise heavy damage at very long range, with very high speed, and with very high mobility, at a relatively very low cost. Zero-K artificially deals with this by making air do unrealistically low damage, at unrealistically slow speeds, and with unrealistically low mobility - for its cost, against strong AA.

2. In practical terms, a modern army that controls the air just wins outright. Rockets/Missiles are just an extension of this power.

3. Developers have no choice but to artificially reduce air's power to "balance" it with ground units, because so many people want to play with ground units.

4. This is often done by artificially improving AA. However, this makes AA so powerful that it has to be artificially banned from use against other ground units to avoid AA, in turn, dominating the ground game. AA units in Zero-K like the Artemis realistically are just guided rockets with extreme range, damage, and accuracy - an air unit. However, it makes no sense that an anti-raider version would not also exist. It is much easier for a rocket to hit a unit that is mostly moving in only 2 dimensions instead of 3. It is an example of developers trying to artificially prevent air from dominating the game. Such a unit would quickly eliminate raiders from use. Unless a team first dominated the air, the tactic so loved of overwhelming a sector of the battlefield with an army or sneaking raiders by the front would never ever work. Of course, if a team dominated the air, then air would already have wiped out the enemy anyhow.

5. Cloaking units is really just a form of sensor jamming and that is what modern armies do. Modern air war is fought between the jamming and anti-jamming/stealth forces before either side fires a shot. The side that loses the jamming war loses the air war and therefore loses the war. In games this is just boring and often frustrating. Zero-K allows ground forces to cloak themselves, but does not allow high-speed air forces to do the same because obviously this would simply allow air to eliminate Zero-K's powerful AA.

There really is no good solution. Once developers introduce air units into a game they naturally will tend to dominate because of number 1 & 2 above. Because of 3, conflict arises between air players and ground players. Since ground players always out-number air players, air play becomes an unrealistic mess with air players (like me) stilted and ground players making all kinds of convoluted irrational arguments about why air play must be restricted to "balance" it for realism.

The fact of the matter is that air power by its very nature cannot be "balanced" without making it hyper unrealistic. Eventually air players will develop a tactic that dominates a favored ground tactic and the developers will favor the ground players simply because there are more such players. This dynamic will lead to players like me leaving the game.

Developers need to control expectations by stating up front that air play will never be allowed to reach its real potential, in their game, simply because it is not fun when ground play is realistically dominated by air play. It needs to be stated that if air play begins to dominate, then it will be nerfed so that it is always a sideshow niche part of the game, but that developers will endeavor to keep air play from being artificially restrained to the point where it is actually bad to build air at all.

Unfortunately, Zero-K is at the point where the only cost-effective use for air play in lobpot is for owls, swift scouting, 1-2 thunders, and enough bombers to compel the other team to spend more on AA/swifts/raptors than the bombers' value - not even possible on some maps (plus there are other ways to compel a team to ineffectively spend its metal so building bombers for this purpose just feels bad). In 1v1 air is useless, unless it can be used by surprise.

Edit: It is just the nature of air (i.e. a unit that moves in another dimension) that makes this hard - it has nothing to do how much or how little realism a game tries to incorporate.

I would admittedly like it if Artemis' fire rate was halved... It would still be great area denial and good at stopping bomber swarms but would no longer completely obsolete Chainsaw and other lighter AA.

Some people look at the units and think rock paper sizers, but rock is slow and can not engage with sizers. Instead, think interns of area denial. One game I was playing air, I told my team, If you knock out this AA, I will kill his com. My team knocked it down and in came ~6 Ravens. The AA made air less effective while it was there. The same interaction exists with riot and raid units. Riots have a difficult time engaging with riots. One could say a raid unit counters a riot, due to scouting and brining the skirmishers into the area. Riot units don’t engage raid units, they create an area denial to protect other units. If you knock out enough riot units, your raid units become super effective. The thing with AA, is that it provides area denial for 2 whole factories. Would it be better to have 2 types of air, requiring 2 types of AA? It’s almost there. Flails are overkill to smaller air, while other AA is better at the smaller air.

All zero-k units are based on reality to a degree, that is the starting point for all games. It's just a fact like 1+1 = 2. The reason air is so difficult to balance in a game is because it moves in another dimension than ground units and that gives it an inherent advantage.

These statements are quite frustrating, but I predicted them:

"It is, by design, a support factory. The main game in ZK does [must] happen on the ground."

"Air is certainly not useless in 1v1, but I don't want air to be so powerful as to be a mandatory switch. Watch some games with air as a second factory. It isn't just there for the surprise factor."

I've watched games where players use air without surprise in 1v1 and the only reason they have any success is because the other player does not use the cheap counters or the economy already is unbalanced.

I would like to see all factories have some air units while keeping the current air factories as optional support factories. After all, every single factory has dedicated AA units, it makes sense to for them all to also have air units. Landing pads are cheap enough or just add a solo pad to every factory. Likho could be added to the strider factory instead of any ground factory. If the idea has any traction, then I will post suggestions for each factory.

A lot of people (myself included) don't see a future with all out, do whatever it takes, warfare. Modern weapons are just so powerful and destructive.

Imo modern warfare is more of a 4x game type of thing.

I think air in ZK is balanced and that's the most that you can ask for out of a RTS unit and faction.

I also don't feel air is weak at all.

If you've ever played a game where an air specialist like Sab or Jummy hits their stride, you'll know that they can single-handedly turn the game. The well-known Likho Shower is an antidote to everything but the toughest units.

I'm not an air specialist by any means but running an air fac felt pretty potent in this game (with lathe assist from Loubouchan it must be added)

http://zero-k.info/Battles/Detail/1286272

To channel my inner sirlin~~~ and map RTS to fighting games:

--------

---

In most RTS is like fighting games with only a single character, while Zk is like a tag team fighting game.

Most RTS games is like fighting games where the character gains completely new, cost free, better moves when you super meter fills up, often changing gameplay dramatically going from super armor throw to half screen pokes. As a result a lot of strategy is about charging up the super bar like turtling and spamming keep away. In some game states both sides with both spam keep away to power up the bar or do some strange block string to power up instead of hurting the opponent. The players of those games tell us that the precise input and strategy to win the power up minigame is valid form of skill and balance and should not be removed.

Zero-K is like most normal fighting games, where you get a set of moves that remain valid most of the game and gain only a few at late game stages. All the initial selection provides a diverse set of options that span most game situations, with the more gimmicky choices merely having nonstandard but still valid responses.

The exception is of course, the Greater Spider Factory (known is GS), which have +150 range and +40% speed bonus and is totally overpowered. The solution to balancing the game is to give every other factory a special anti-GS "move" that is only works against it, has +100% range, +20% speed, +50% damage and overpowers the GS if you spam it. This makes the GS only useful when the opponent is pinned by a blockstring or combo normally....

Now as an aside, a few result from the development of fighting games:
1. Extremely simple games can take good amount of skill to master. Meme "one button" fighting game "dive kick" can have real player skill differentiation. Very simplified games like "Footsies" or Fantasy strike can also have large amount of player skill. RTS games are far more complicated than 1 dimension movement, 2 attack move game footsies and optimal play is impossible, the problem with bad games is that skill needed is unfun, for example:
I. Computing nash equilibrium response to of dozens of possible cheeses after obtaining the resource count of 3 opponent resource patches
II. Compute linear programming optimization problem on multi-resource investment after impact of random event (enemy action or else)
III. Optimizing harvester pathing and build queues
and so on...

2. It is possible to have pretty balanced games that fit an aesthetic. In fact in the fighting game domain, pretty balanced games are common (balanced well enough to be irrelevant to 99.99% of the player base: one can always get better and beat everyone out of the 0.01% by pure skill) and games compete on giving players what they want and thus generating subgenres like anime fighters, different movement systems and so on. Other design criteria like ease of learning is valid, thus things like "auto-combos" which reduce the pain of low rank players a lot.

--------

---

To respond to this:
https://zero-k.info/Forum/Thread/34717?postID=247421

quote: I'm not sure if I'm even looking to promote meaningful micro between air and AA, or more than exists. The skill test for the player making the AA is a combination of valuation of what to defend and a bit of 'tower defense'. The skill of air is to judge what they can kill for what sorts of losses, and to decide whether its worth it. The upshot is, if air units start looking too much like floating ground units, then it becomes hard to justify AA as a class of target restrictions... Sometimes a big blob of AA is like a weak anti-nuke. To use air in the area you've got to use other forces to destroy it.

I think the big part of the appeal of Zero-K in general is that it has the least "strategic" level consideration out of RTS games. It is in practice intensively tactical with no tech tree and the most barebones economic model, even simpler than explicitly tactical 'map capture' games like CoH. Now Zero-K has shiny and super stuff, but they are game enders after long tactical fights, and they are mostly buildings and "not stuff you are suppose to fight with."

When people choose a factory, there is more to it than choosing the strategic optimal move, like how people choose characters in fighting games. No, the aesthetic and style within the design is important and people want to play the game a certain way and have a path to grow into strength.

In just about all games there players complain about balance not because the top performance is easy, but a style of play the game advertises (or otherwise desired) is much weaker, dumb or otherwise disappointing and they are forced into unattractive "OP tactic". Airplanes and gunships are disappointing compared to land factories in that they fit in narrow niches without full spectrum gameplay for both sides.

Another "land factory" with the same level of uniqueness delta as spiders is compared to tanks would be more attractive than the existing gunship factory as it means a completely new and large set of complex tactical interactions. Which is to say, a fast hover-spider factory that is fought using normal units and not AA units sounds more fun.

quote: * Confers all-terrain pathing and the ability to (mostly) phase through units. Changes which weapons are allowed to shoot at it. Makes the unit use flying armour/weapon upgrades. We already have all-terrain units of average speed (spiders) so I don't think we need this with the addition of floating. Also, all-terrain is less important in ZK as the maps are much more open, so a flying unit of average speed would behave a lot like just another ground unit.

The most important part of air units in the standard scale RTS is stacking, which can be degenerate in enabling infinite density. Special AA unit with splash may be needed to prevent Air scaling but not much more. In any case, in a force density dependent game this variable has to be exploited, where as armor and weapon class doesn't matter that much on the grand scale.

quote: Basically, I'm not looking to promote this: quote: Similar ranges between anti-ground and anti-air weapon also promotes a lot of air-ground combined arms forces in which air hovers over ground. Compositions like Mutalisk Zerging, Marine Science Vessal, VoldRay-Ground, Raven-Mech and so on. Compositions like this are what I'm calling tactical uses of air units. These are armies of units with somewhat similar speed that move around in a blob and fight. The air units contribute the same sorts of things as ground units: firepower, abilities, and being in a target category. I imagine Muta-Zergling is more like the air/ground combination that I'm going for as this seems like a raiding army rather than a fighting army. But, in particular, Voidray-Ground just seems like a blob that strengthens the army it's in by adding another target restriction class.

Many of the Zero-K air units is more like Oracle or Banshee: there is a timing window for you to make a raid against base defense, not fight the army, and the window close and you just about never make significant again outside of special tactical situations. The interaction is mostly unit-static base defense and completely forgetable.

Muta-Zergling is a proper raider army composition: in that it can raid, but it is also the thing holding the line to prevent the opponent army from walking into your base. It can not win against land armies directly, but exploiting terrain it can inflict even attrition, hold bases with help of structures and chokes, and win base raid/trades in specific conditions and thus pinning the opponent army like all fast army compositions. There is a huge number of tactics and micro for both sides of the struggle, from controlling of mutalisks itself, baiting marines into Zergling surround, marines microed into line formation to focus fire a specific mutalisk in a stack, actually ambushing mutalisks with marine flank, hp regen management both both sides. Then one add additional units to mix up the tactical interaction: baneling over run, widow mines, tanks and it gets really complicated and fun quickly without too many arbitrary rules.

Then there is the idea that zero-k gunship of all things are raiders. From what I've seen, the most common use of the factory is nimbus in an army role, stacked over ground forces. The second most common use is rev as early game unit snipes, third would be trident also stacked over ground forces, being not much faster than some ground AA and can not to reactive interceptions at all. It is actually a good thing gunships is weak because nimbus is such a boring unit, attack move auto-retreat mono-spam dominates the likes of chicken games.

The high cost and reasonable health of higher tier zero-k gunships plus its ease of retreat means it is actually pretty good army unit and pretty bad raiders as it lives off repair -> attack cycle with fairly slow speed, with accessible repair power generally focused around defenses or army. Gunships is actually not fast enough to flank against faster ground AA for backline attacks on accessible flatlands, as AA have similar speeds outside of locust, and airplanes are generally better unless against undefended positions.

If gunships are already army units, it can be made better. More distinct roles, more distinct tactics... away from uniform hp/cost and dps/cost.
-------------

As for airplanes, what you get is IMO:

1/3 the reaction time while demanding 3 times the situation awareness to play to good effect, compared to surface factories. Now in 1v1, air is a side show, maps are smaller and opponent are more understood with consistent interaction. In lobpot, information explodes in density.

In ZK, airplanes are suicidal and take significant and fast manual control to work. The required information for successful sortie is partially observable: it is possible to figure out what is happening with fast flyby, observation of radar dot movements and observation of unit hp/penetration depth before death: but none of that is easy.

I am reminded of aircraft carriers in world of warships, before they removed the whole RTS Carrier part. The problem with RTS carriers was that it has a brutal learning curve (difficult plane micro, learn AA drop off curve of all ships, need game flow awareness to maximize game impact). In that game they also have huge game impact as such, elite players are capable of things like 70% solo in a 15v15 random MM while the average player just get pwnt out of the game (match maker make both side have equal carriers), with the carrier player base shrinking over time.

ZK Air may not in the absolute sense be very powerful but I think it has greater skill - power curve than land facts in lobpots: any middling player can learn the lance ball or some robust porc structure and do okay, while air take a different skill set that require dedicated effort different from playing the game normally. Making air powerful is tricky if player skill-power scaling is higher since it would mean the best player should play air to carry, which is bad for fun.

The lack of direct counter play options for other players means that "uneconomic" air defense massing has to be done if one is to not get farmed repeatedly. This is really rather boring for both sides.

One may actually lower the skill (effort) - power curve to enable increases in absolute power. (how the average player experience air) This may not require game changes since good education resources can change effort - skill - power curve without changing the game. Alternative changes may include modification to interface, either better defaults and micro training or something more significant.

-------------------------------

quote: How powerful do you think air should be? What sort of impact should it have? Should it be so important that the best player on each team has to play it? What motivated this post?

Now from a personal greedy demand is infinite perspective:

Air should be, following the theme of the game:
1. Ground factories fight air units with tactics of maneuver and fire control not preparation of hard counter units
2. Air factories should have full tactics-filled game play against each other and ground factories
3. Ground factories fight air units with normal units generally, not exclusively using AA units.
4. Gunships should have full start to end game life cycle
5. Airplanes, as a support factory should scale mid to end game
6. Degenerate air deathball does not happen, or have reasonable counters
7. One does not have to go air to not suffer a very large disadvantage in most game-type/map combos
8. Air units should have good dynamics, feel good to control


Some ideas on how to do this:

Smaller Things:
1. Locust gets a short range AA that allows it to brawl down Trident. This can be build into raider Locust > 'Skirm" Trident, plus ground riot > Locus circle counter. Should add potential tactics than monospam to gunship mirror air superiority situations without impacting ground negatively. (eg. larva maps with land route terraformed out)
2. Revenant gets sonar and underwater targeting. Existing weapon already splash to underwater. Why leave one of two air factories completely unable to deal with half of sea?
3. Lower flying transport, either specifically or option select. Faster pick up and less drop damage on being shot down. Balance by animation if too much.
4. Add a low, avoidable damage for cost, high survivability airplane to scale to late game. Like I said before, if bertha exists there is a threshold where damage inflicted per cost is quite tolerable and not oppressive. Air player want to play air and that is why this thread exists.
5. Buff sparrow so airplanes is not so critical in non-tactical sense
6. Minelayer plane: drops persistent mine that costs low metal (relative to moving bombs). Should enable some creative play.
7. Blastwing gets self range radar jammer. Should allow it to be used to screen after long range AA is built up.
8. Strafing airplane as flying Halberd: high hp needed for multiple passes, low effective dps for tough and fast vehicle to balance. Maybe give its weapon napalm effect as a tough/fast anti-cloak/repair option since you can't let it kill much and its not much fun to have its weapon be just plain useless.

Larger things:
1. Shift AA to slow, emp and gravity in general, outside of anti-stack splash AA. This way AA can not complete the kill chain just by being placed, and instead normal ground armies have to move to fight it. This enable fast air strategically, but slow for tactical time frame interactions and to interact meaningfully with both ground armies.
2. Reduce ranges of air units and some AA to be more in line with normal units to reduce dependency of dedicated AA. This also adds ground force density constraints to AA, which interacts with terrain and ground force requirements to produce unique tactical situations. At common 800+ range with high cost/power density, existing AA is not surface area constrained much.
3. Bombers as anti-artillery/anti-force multiplier, not effective against line forces of assaults/riots. This will demand tactical scale flanking to avoid front line army units (and opponent counter micro), not focused on strategically flanking roles or computing long range aa density.

shin_getter that is an interesting post. I'm not sure what to say since, in the theoretical parts of the post, I tend to think one of the following:

• You're describing how ZK already works, so you seem to be solving solved problem, or
• There are good reasons why ZK shouldn't or can't try to work how you describe, and I don't see the solution working.

Sometimes I consider the problems solved on the design level, but the balance isn't quite there. I still don't see the need for design level solutions.

In the practical parts of the post I tend to think one of:

• That unit/change sounds a bit too janky to work, or
• Anyone is free t o try to make a polished version of that unit idea.

Maybe I'll have to start with a rough list.

Now for some specifics.

quote: The most important part of air units in the standard scale RTS is stacking, which can be degenerate in enabling infinite density. ...weapon class doesn't matter that much on the grand scale.

I covered this under "phasing through units" and I agree. I'm not keen on having air that can stack more than it does. Ideally bombers wouldn't stack the way they do now. Although I don't see how you can say weapon class eventually dosen't matter. Something like 2/3rds of starcraft units can't shoot up.

quote: Many of the Zero-K air units is more like Oracle or Banshee: there is a timing window for you to make a raid against base defense, not fight the army, and the window close and you just about never make significant again outside of special tactical situations.

I disagree on an objective level. Likho exists throughout the game and Thunderbird seems like it should too. Raven might with the changes. Nimbus and Trident have staying power. I see Krow appear late and used well. Transports are always a niche risk. Owl is extremely useful. You're correct about Locust, it has a timing. Locust is a raider though, that's how they work. You're not describing how I observe air to behave.

quote: The interaction is mostly unit-static base defense and completely forgetable.

I think you underplay the interactions of mobile AA and just making and controlling units in a way that isn't vulnerable to air in the first place (such as spreading vs Likho or Phoenix). However, on static AA, I'm not overly concerned by the defending player having no orders to give. There are already plenty of orders to give in ZK and I don't think requiring players to click on enemy aircraft will lead the game anywhere good. The defending player did their side of the interaction when they decided which AA to build, where to place it, and how to position all their other assets around it.

I think we might have exhausted the Muta-Zergling analogy. I think it's fine for a mix of raiders and aircraft to be able to take fights. See Raven bombing riots or Thunderbird disarming an area. I've seen similar things happen with Harpy in the past. Mutas baiting Marines only to get killed by Zerglings is the same interaction as raiders picking off exposed AA. I don't see how ZK dramatically differs from what you describe, up to some reasonable balance tweaks, and within its context of escalation. I think Nimbus is fine, it encourages an AA interaction. It is the type of unit that could be mindlessly spammed in chicken, but I don't hold that against it.

quote: If gunships are already army units, it can be made better. More distinct roles, more distinct tactics... away from uniform hp/cost and dps/cost.

This is one of the points where I don't see where you're coming from on a theoretical level, but on a practical level I agree. Gunships could be more interesting. Revenant is a pepetual one trick pony. Locust perhaps needs to be moved away from its knife edge. Harpy has not adjusted to advances over the years. Krow was the first good strider but other striders have since become good.

quote: As for airplanes, what you get is IMO: 1/3 the reaction time while demanding 3 times the situation awareness to play to good effect, compared to surface factories. Now in 1v1, air is a side show, maps are smaller and opponent are more understood with consistent interaction. In lobpot, information explodes in density.

I agree, disagree, and don't see a problem. To me, the 1/3 reaction time and 3x situational awareness is the asethetic of planes, and GS to a lesser extent. In teamgames it's the ultimate support factory. It can't fight armies or even hold ground, but by overviewing the whole battle you can find the perfect spot to dive in and provide the push needed to make a push or shore up a front. This is what I want to be doing as the air player, striking swiftly and strategically, not just playing a land factory in the sky. On 1v1, "side show" is more emotive than specific, but I can say for sure that air is still a common escallation pathway. Planes and gunships are each more popular than all the other second factories combined. Knowing air is less important than knowing ground as not all games reach this point, and people can pursue airless styles. Both these things are fine by me.

Time for some quickfire:

• "In ZK, airplanes are suicidal..." - Likho and Nimbus seem to survive pretty well. Some air units are designed more for repair than others, just like ground units.
• "...and fast manual control to work" - Maybe you're talking about a bug or some maneuverability characteristic that can be tweaked? This doesn't seem like a point against the rest of the design.
• "The required information for ... but none of that is easy." - Yes, there is skill to using planes.
• "I am reminded of aircraft carriers ... brutal learning curve ..." - Air is different and perhaps harder, or at least suited to different playstyles, and balancing this against its power has been tricky. Air used to be more mandatory but that lead to problems. I think it is still good, but not an overwhelmingly necessary plop. Factory Plates exist in part to let someone plop air and switch into ground, and for other players to pick up some air if they feel the need.
• "ZK Air may not in the absolute sense be very powerful but I think it has greater skill - power curve than land facts in lobpots..." - Yes, see above.
• "Making air powerful is tricky if player skill-power scaling is higher since it would mean the best player should play air to carry, which is bad for fun." - Yes, luckily it seems like people who enjoy air get good at air so it somewhat cancels out. Also plates allow players to switch roles around more within a team.

quote: The lack of direct counter play options for other players means that "uneconomic" air defense massing has to be done if one is to not get farmed repeatedly. This is really rather boring for both sides.

I don't think there is a fundamental lack of sufficiently economic air defense for land players. Land players often can't ignore air forces coming at them, but forcing some AA is just an aspect of the power of air. I don't observe players being farmed repeatedly by air.

quote: One may actually lower the skill (effort) - power curve to enable increases in absolute power. (how the average player experience air) This may not require game changes since good education resources can change effort - skill - power curve without changing the game. Alternative changes may include modification to interface, either better defaults and micro training or something more significant.

Sure, sounds good. Work on some QoL or guides. This is the type of thing that has already been happening.

Well said, GoogleFrog, I agree, I like airplanes to be a little more different in that way (more attention needed, fast action, scaling, etc.) otherwise ZK would be boring if all the factories were the same.