Armor Systems

Another thread about counters which was originally about hard vs soft counters, having no arbitrary damage types, and general unit design has mutated into a discussion about armor which I think deserves a reboot in a fresh thread of its own.

There are multiple armor systems being discussed, with some confusion and muddling between them. All of these approaches assume there will be no arbitrary armor or damage types. I will describe each of the proposed armor systems in order of complexity, from simplest to most complicated.


HP Only

The simplest system. All units have a single number representing their durability. Units with heavier armor will simply have a higher HP number, allowing them to take more damage before they are destroyed. This system is perfectly capable of producing a very interesting RTS game, and this simple, purist system has largely been adopted in TA, SupCom, and Zero-K (with some damage types and mitigation, but minimally used). This is the default, as it has been a highly successful system in the past, and any proposed system has to meet this bar for functionality and depth.


Flat Armor

Flat armor is a simplistic armor system, which adds a small degree of complexity to the HP-only system. Each unit will have an HP and an Armor value, although its armor may be zero. All incoming damage is reduced by the unit's armor value.

A minimum damage rule may need to be in effect. A minimum damage of some small number means even against units with high armor, all weapons will do at least that much damage. Still, a minimum damage of zero is workable if the game is designed with the possibility of unit A being completely unable to harm unit B is taken into consideration. Generally this situation should be minimized.

This means that a unit with 10 armor which is dealt a 20 damage shot will have its HP reduced by 10, a 50% reduction. By contrast, a 200 damage shot will reduce the unit's HP by 190, only a 5% reduction. The important gameplay function of this system is that small shots are less effective against units with heavier armor, while heavier weapons that deal more damage are proportionally more effective.

Cheap, small, numerous units with light weapons will be naturally less effective against units with more armor, and can be very efficient sources of HP and DPS. It also enables heavier units to have less HP, making them squishier against heavy weapons to encourage smart play and positioning rather than brute force HP stacking.

Units that have very large amounts of HP become extremely difficult or time-consuming for any type of weapon to kill under a flat HP system, giving such a unit relative impunity as long as its HP holds out, which is undesirable. The enemy should be scary for every unit, prompting players to play as intelligently as possible. Even the most effectively durable unit should have to worry about whether an enemy weapon that can kill it is just out of vision range.


Proportional Armor

Proportional armor is similar to flat armor, only instead of a subtraction, incoming damage is reduced by a fraction. This system was used more extensively in TA and Zero-K than a flat armor system, especially for pop-up turrets and closed solar panels.

Reducing all damage by a percent has literally the same effect as increasing the unit's HP by that percentage. However it makes sense for certain units, like pop-up turrets or the Crabe, which switch between modes with differing levels of durability desired by the designers. Changing the unit's HP is a silly solution, so giving the unit proportional armor in its more durable form accomplishes the same goal.


Directional Armor

In some games, especially more detailed tactical games like Company of Heroes, have armored units with varying armor on a unit's front, side, rear, and even top. Typically a tank's front armor is its strongest, with weaker side, and quite weak rear armor. Projectiles striking the rear armor will do the most damage.

The damage reduction system used by directional armor should be either flat or proportional. It makes sense to use either, but their effects differ significantly. Flat armor is more effective against weaker projectiles, whereas proportional applies to the same degree against all incoming damage. Flat armor might reduce a very weak projectile's damage to zero, which has some weird effects. And proportional armor makes the unit more durable, but the most efficient DPS sources remain the most efficient after mitigation.

In my opinion directional armor is overcomplicated, and provides relatively little depth. Especially with large numbers of units involved, whether a unit's front or side armor is hit is largely a function of luck. It does make unit interactions more intricate and less deterministic, but does so in an opaque manner that is difficult for a player to intentionally leverage.


Destructible Armor

Destructible armor is a rarely used system, however it is innovative enough that I believe it deserves a mention. Gratuitous Space Battles utilizes this system, where armor mitigates damage, and is itself damaged in the process, reducing its effective mitigation.

Note that the method of destructible armor's mitigation is likely going to be flat or proportional, as the armor's effectiveness degrades with damage, reducing its protection.

It is theoretically possible to have an armor system that is both destructible and directional, but this is by far the most complicated armor system and in my opinion the player would seldom ever actually use such a system for effect, in either strategic or tactical terms.


Conclusion

In my opinion the HP only system is the gold standard of armor systems. Its simplicity and versatility made TA radically different and greatly more innovative and emergent than other RPS-centric strategy titles over the last ten years. However that is more a condemnation of rock-paper-scissors balancing than about armor itself. There may be a role for armor to play, as it allows a second degree of flexibility in unit design, allowing units to have less HP while being more durable in some contexts, provided the rule is universal.

As a result, I am partial to the flat armor idea for creating an emergent soft/hard target paradigm in PA. Much like how single-target weapons with long cooldowns are generally weak against swarms of cheap units, it makes sense for small weapons to be weak against armored enemies. Small weapons, however, should be the most DPS-efficient, which theoretically makes them the most effective against everything in the game, including high HP enemies unless short range or some other factor negates this high efficiency. Range creep in the late game may result from relying on the HP only system, and no damage types.

I vote for the HP only system, but have my eye on the flat armor system to see if we can get it to be superior.

Destructible. Failing that, bog-standard HP only.

No. 90% damage reduction increases a unit's health 10 fold. It gets pretty damn effective at high values.

Flat HP can serve a vast majority of needs for PA, as long as every unit works on the same page. It is no good if one theater starts using its own scale.

TA also had hunker bonuses in the form of solar panels and pop up turrets. The former makes the structure strong against sustained attack, but weak to surprise burst fire. The latter makes a structure strong against surprise damage and units it can not attack. The solar panel bonus was nice against pee wees, while the toaster bonus helped quite a bit against bombers.

Damage types should be used as little as possible, and preferably not at all. The physics engine already works as a method to create good counters. Let projectile behavior take care of things. If a unit absolutely totally needs damage resistance, then it should be applied directly to the unit and be part of its design.

You are correct on the math of proportional armor bobucles. I wasn't thinking about the numbers when I wrote that. The point is that it is identical to simply increasing the unit's HP.

None of these systems are conditional, and none of these systems are damage types. They behave universally, and would obviously be an integral part of a unit's design.

I'm going to go ahead and say I'd like to see a Flat Armor system at least TRIED.

There is something innately intuitive about a piddly little gun on a weak scout bot that is unable to scratch the armor of a hulking tank. And it should be fairly easy to visually show "You are being ineffective" with some sparks and ricochet effects when a light weapon hits a high armor target.

In addition, it expands the factor of different weapon types being more effective against one unit than another. In the HP only method, this was expressed purely as overkill - how much of the shot's damage is wasted by shooting at a unit with less HP than the shot's damage. However, overkill is limited and only gives you one option - weapon power > unit health. So you can make a high damage weapon less effective against weak units (Eg, Annihilator shooting peewees) but you cannot make the low damage weapon less effective against strong units (Peewee vs Goliath). You also cannot have a unit that is weak to high damage but strong against low damage (eg, a bunker defense that shrugs off small arms fire but gets blasted by a tactical missile).

In the Flat Armor method, we open up the possibility of making different weapons be more effective against different units, without making things too complicated. It's not too hard to visualize that lots of bullets (Peewee) is less effective against armor than a single rocket (Rocko). This allows for the tank to have similar HP to a Rocko, but due to armor it fares better against Peewees than a Rocko would. So Peewee > Rocko > Tank > Peewee purely based on weapon damage and armor values, as opposed to the TA method which is to make the economic cost of each be the balancing factor.

This is beneficial, because as your economy grows, the economic balance factor shrinks and you begin to choose units not for their economic impact but for their battlefield prowess. Peewees are no longer built, but mass Rockos are, because the more expensive Rockos perform better in large groups than the Peewee.


Anyways, there are quite a few more factors in balance than just weapon damage / armor values, but hopefully you all can understand why I support expanding the field of unit balance just a bit more to help keep all units useful all the time.

Flat damage reduction blows up when units start doing damage close to the armor values.

Exactly. See "Marines vs. Ultralisk" in starcraft. It's a bloodbath if the ultras have upgraded armor.

I think flat armor can create great games, but it's not what I want for PA. Directional seems interesting to me, simply because it makes positioning important. Maybe others don't want that.

I also want defensive buildings to have directional armor (some of them, at least)

Care to elaborate? I'm not familiar with marines or ultralisks.

You forgot my favourite armour system: Directional, destructible components, proportional.

I've a softspot for directional proportional armour systems, just because it makes flanking so much more effective. Also tank destroyers with high frontal armour just somehow feel right to me.

Holding an enemy advance with TDs while moving a couple of dozen light tanks around the flank who then can easily shred through the rear armour of the enemy heavies... its something you can't really do when units just have hp as a defense mechanism. (Its not micro when I'm ordering 50 tanks to flank the enemy. )

(I'd actually like an armour model with armour thickness and armour penetrations of weapons plus destructable armour plates... but that would definitely take it too far. I've played too much World of Tanks in the last months...)

Imo, taking some cues from WW2 for tanks is a good idea, as with naval warfare or space battles for Brittain. )

If you'd add an armour system it should be fairly simple and generalized, since armour usually is something you can't really see at a glance. Ie. all armoured units would have no armour at back, half at their sides and normal amount (whatever that would be) at their front. Super armoured units, like tank destroyers, would have reflected this in their unit model, ie. "blast shields" or whatever in their front. So a tank would take 50% less damage from the front, 25% less from the sides and normal damage from the rear. A tank destroyer would take 75% less damage from the front, but full damage everywhere else. (One just averages the values for any hits at an angle.)

While you can make tank destroyers with just hp (low hp, high rate of fire, high damage, low turret turn rate or no turret at all, low tank turn rate), it would imo not be the same. Since any high alpha enemy could kill it (if it gets in range) from any angle, but with a proportional armour system, such an unit could last against a frontal assault for a while while attacks from the side or back would almost immediatly kill it.


It just feels to me, that when we have a very complex weapon modelling system (weapon hits depend on physics which is more complex then 99% of the usual damage systems in RTS) a bit more complex defense system couldn't hurt too.

On the other hand, an armour system with discrete values (ie. flat damage reduction) would imo be too complex for the scale PA wants to have. The issue is that flat damage reduction numbers are invisible and if they vary between units you have no way of knowing them beforehand. Also if damage absorbtion values and damage done are too close together you can easily get invincible units against certain attacks which is too close to hard counters to me. Even the hardiest of unit should be swarmable in some way. (Even though it may not be cost effective at all.)

Proportional damage reduction has the same issue when you get in the higher percentages (anything above 90% and you're getting in the exponential part of it) but if you limit it its not that bad and most important its very consistent. The annoyance of armour systems in RTS like StarCraft or AoE is that often different armour types works so different that its impossible to say at a glance wheter an unit will work against another unit or not without learning it by heart.

A limited directional proportional armour system could be reflected in unit design and wouldn't be exceedingly complex. Ie. my example of 50,25,0% (front, sides, back) for tanks and 75,0,0% for tank destroyers. This would also further differentiate tanks from KBots, since this was a discussion in another topic.

Certain base buildings could also incorporate armour for more exotic designs. A defensive turred with heavy frontal armour is a given. But how about a forward battle factory which has a 80,75,0% armour to build units directly on the front line? The back would be defined as the area the built units leave of course.


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


Where I think something else then hp would really shine is with large and super units, especially with capital ships and super robots if there are any. Huge units in a system with only hp are usually bullet sponges that are effective at 100% till their last hp, then they die.

Having a number of destroyable components for huge units could make them more manageable for smaller units and make them less of a kill asap or die target. It shouldn't be too complex of a system of course.

Battleship for example. It would have: 3xTurrets, Engine/Rudder, Bridge. Lets say each component has a fifth of the units hp. Damage done to a component is applied to the hp of the unit (even if the component is destroyed). So doing the full hp of the ship in damage would still kill it as normal. But if you do enough damage to an individual turret (the huge ones on the battleship) that one wouldn't shoot anymore. Bridge destroyed, no more vision. Rudder destroyed (via submarines), no more steering, etc.

Or with a Monkeylord, each leg is destroyable. -1 leg, half the speed, -2 a quarter of its speed, -3 legs, immobilized (but still can turn around with only one leg).

Since the numbers of huge units are usually limited (and if there are a tons of them, why do you even care about disabling small parts of them...) the added micro to target certain parts would be imo acceptable.

Doublepost 'cause I didn't want to add to the wall above...

Iirc it goes something like this:

An upgraded Ultralisk has an armour value of 6 with 500hp. A Marine does 6 damage per shot (9 if upgraded). So depending on their upgrades, a Marine will do between 0-3 damage. (Not sure if minimum damage done is 1 or 0.)

So Marines can shoot all day long on an Ultralisk, doing barely any damage at all. And with 500 hp they take forever to kill. (Ultras also do splash damage, so massed marines don't help much.)


To reiterate my dislike of flat armour systems: They're very opaque (where to you get the values from ingame?) and imo lead to bad balance with hard counters (high armour against low damage units).

It imo also doesn't add that much to the game. Directional armour at least increases the importance of positioning which is imo much more interesting and also much easier to understand.

Reminds me of StarCrafts #25! xD

Mike

I <3 directional armor because my two favorite maneuvers are flanking and bait+switch. It kinda pissed me off that Firebases couldn't be flanked in SupCom, as that is the primary weakness of a firebase, getting behind its defenses.

I think it would be unbelievably cool to have flankable Tank Destroyers and Tanks, as well as emplacements. But that's for other games, PA is far to strategic and futuristic to be dabbling in anything like this. If you're going to build a gun turret, it's going to be heavily armored on all sides, because why not pitch in that little extra bit of mass for invuln to flanking maneuvers?

In TA-derived games, the advantage to a flanking maneuver isn't that you do straight-up more damage, but rather all of the other benefits it confers. AoE weapons are not as effective, direct fire weapons with *any* degree of inaccuracy are less effective (read: anything but beams and hitscans), forces get some redundancy, walls and terrain aren't as debilitating, your enemy CANNOT retreat without taking heavy losses, you can reinforce much more easily from multiple directions, you can trap enemy units if you flank from all sides, and micro becomes exponentially harder for your enemy, and only slightly harder for you.

Ok, I can understand this may be irritatingly broken, but is armor really the reason it is broken? Or are the other factors (cost, efficiency, damage) simply multiplied by the armor and thereby making the unit broken?

In isolation, the idea of armor is that unit A does less damage to unit B than it does to unit C. In the case of the Ultralisk, if the marine is designed as the counter for the Ultralisk, then obviously it's broken. But if there were some kind of higher damage, lower ROF unit other than a marine that could take on the Ultralisk, it wouldn't be nearly as broken right?

I'm not sure how a proportional directional system would be any less opaque. How do you know that a tank killer has much stronger armor in the front? Where are the numbers? How do you know that unit takes 50% damage from the front, and that one takes 90%?
Unless you suggest a global armor system where ALL units take 50% damage from the front, 75% from the side and 100% from the back.

As for positioning and flanking, a much simpler and intuitive way to promote positioning is to limit turret turn rate and arc of fire - especially for defensive emplacements, if you had a fixed arc to fire in, it would make things rather interesting. Units coming in on the flank when everything is pointing forward get an automatic bonus while the turrets turn to me the threat.

This marine/ultralisk situation is being held up as a case where a flat armor system is dysfunctional is completely wrong.

Starcraft uses a weapon/armor upgrade system, where there are three levels of upgrades. Upgrading weapons increases your damage by +1, typically, and upgrading armor increases your armor by +1.

Ultralisks have 1 armor by default. Marines deal 6 damage by default. An ultralisk with +3 armor will have a total of 4 armor, and they also have a special Ultralisk armor upgrade which adds +2 armor to the Ultralisk. This means that, with every upgrade possible, ultralisks have a maximum armor of 6.

However, marines should have weapons upgrades at the point in the game where ultralisks should have armor upgrades. So instead of dealing 6 damage, marines would have +3 weapons, dealing 9 damage.

Additionally, there is a minimum of 0.5 damage in Starcraft. So even if you had unupgraded marines against fully upgraded ultralisks (ridiculous) then each marine is doing 0.5 damage per shot.


In practice, marines are actually fairly good against ultralisks in Starcraft II, despite the ultralisks' armor.

Nah, the Ultralisks are more of a counter to Marines then vice versa I think. I guess the complaints are more that Marines just can't roflstomp everything. (But what do I know. I play Starcraft for the story. Its a horrible story and badly executed... but there aren't any other RTS with campaigns nowadays. )

But honestly, I just don't like the fact flat armour numbers would effectively just hit any weapon that is low damage and fast fire. I like my PeeWees killing big tanks.

Basically yes. All numbers are made up of course, but it would be a simple system. Units which look like they're well armoured (ie. graphically) would have tank armour and units which look like they have a big frontal armour would have that type of armour. And finished, no more variations, since because its a hidden stat, true, it needs to be as simple as possible. It wouls also only apply to tanks. KBots wouldn't have it, neither would airplanes or whatever.

I agree here with you. Imo the point of an directional armour system would be more to allow low/medium damage units to do significant damage against highly armoured units when flanked. Ie. usually a heavy tank has high hp, high damage but is slow and slow to bring its weapons around. You can flank that one with a faster unit and eventually kill it, which would be slow though since you still need to eat through all the hp.

The benefit of directional armour here is that you can offload some of that hp into the proportional armour, so that attacks from its flanks or from behind kill it faster.

Lets design a heavy assault tank (for attacking fortified positions). High hp, medium speed, slow rotation, slow fire rate. (Yes, there's artillery and the like for attacking bases, a heavy assault tank is fun though. ) How do you counter such a tank?

Longer range weapons, outmanouvering it (which is hard when it rolls towards a base), higher dps, etc. Which are actually the things such an heavy assault tank is designed to fight against. And if you flank it, you still need to eat through its massive hp pool. With directional armour, once you outflank it, you need to do less damage to destroy it. Ie. once such a thing is in range of the things it wants to destroy, with normal hp mechanics, you can attack it just as well head on, since outmaneuvring is too late then, it will do its damage till its destroyed.


Of course, does the game need such an unit is another question. It just like the positional aspect it would add. I never have followed Supcom much, but from what I can tell (and I may be totally wrong here) it usually ended up with streams of units engaging head on most of the time and concentrating force (or firepower) in certain areas to make breakthroughs.

This would be mainly to make head on attacks with certain units better and flanking attacks against those stronger.


Edit:

Well, Stimmed and massed marines with Medivacs are good against pretty much everything in Starcraft II, aren't they?

(Banelings, PsiStorm, Tanks... that kills 'em fast of course. The other zerg thingie I don't remember, splash thingie.)

But thats mostly because of how many one can concentrate in a small area with their extremely fast rate of attack.

This reminds me, StarCraft II has some interesting positional play, since you usually move around with one giant deathball, having that deathball in the perfect position and not spread out is often paramount, unless there's aoe involved, then you should've spread out. But yeah, micro heavy that game.

Also, you turn around on a dime in it. Also its the game with the hardest "soft" counters I know of. Knowing which unit is good against which one is totally arcane and coded into their damage types and armour types. Your army can get melted in seconds if you had the wrong unit type.

Yes, one of the struggles with Supcom (and to a lesser extent, TA) was that the later battles really degenerated into giant blob vs giant blob. HP-only systems are prone to that. I'm not saying it wasn't fun, but I do think the overall experience could be improved even more by a slightly more complex defense mechanism.

I guess the disadvantage of armor is the same for any armor system. With regards to the 'hidden' value of armor, directional, proportional or flat - units should 'show' that they are armored. Bullets should ping off the armor, sparks should fly, and the unit should LOOK like it's got armor.

In a flat armor system though, to use your assault tank as an example, you'd have a tank with high armor but lower HP than a 'standard' tank. Rather than needing to wear down a high amount of HP over time, you would need to bring up some slow firing rocket kbots that have a higher damage output. These bots would be less effective against the higher HP 'standard' tank and yet be more effective at taking out the low HP, high armor assault tank.

Exactly like how a modern tank today can blitz through infantry, but needs to watch out for the occasional RPG launcher which can really mess up its day.

Given the amount of units in a sup com army i dont really see how the implementation of armour would effect the blobness, units turn pretty fast so directional armour on the flanking is negated within seconds. Normal armour just means that you need a proper balance of units in your army or you will get your *** handed to you (as all counters become harder). It also increases the potential benefits for micromanaging from relatively slim to potentially massive if you can target your forces appropriately. As skill levels progress the more homogenous armies (because you need to have an appropriate army balance, your opponent will also leading to a 'best' unit ratio) will require micromanagement in every battle because your opponent probably will be leading to consistent losses if you don't.

Flanking firebases can be beneficial due to the time it takes for artillery to turn and aim at a point 180 degrees from their usual heading. This also makes tactics like hitting from multiple directions worthwhile, without the need to have arbitrary damage multipliers/reductions based on direction of attack. Turret turning applies to mobile units also.

In this case, I'm a bit torn between what I would like and what I think is usable in the game.

The directional element, however you want to implement it, would give small scale tactical options and would offer additional distinction between fast and maneuverable bots and slow turning tanks.

On the other hand, to make use of it you need fine grained control over units on both sides if you think in terms of dozens of units. It would depend on the UI options and the low level unit AI to be more than just a statistical modifier to the battle.

Directional armor is already in the game due to the number of units; the front of your army will be short-ranged heavy tanks while the rear will automatically make up more vulnerable artillery guys.

In fact; it's even better than directional armor, because you can decide for yourself how much armor you want in each spot of your army, by simply moving some heavy tanks to the rear of the group to guard it.


As for no armor or flat rate armor; as long as it can be made really clear. I loved the way armor blocks were visible in Company of Heroes; misses or shells that couldn't penetrate armor would visibly bounce off the armor and spin into the air. I think flat armor could be really interesting, but I've also seen games without it work just fine.

Situational percentage armor is also a really cool trick that makes for some interesting units. Although I'd like to see it used in more interesting ways than "no move = armor" and "no shoot = armor", which are the only two flavors I've seen so far.