So, this whole armor thing

[Aeson]

Subtractive armor and reality also have only a tenuous relationship with one another. It's just slightly less tenuous than the relationship between percentile armor and reality.

@Histidine: you may wish to revise your idea slightly, specifically with regards to the minimum armor value. Under your idea, if I have a zero armor ship, any hit at all could potentially destroy it (x/0 = infinity assuming x is a nonzero constant) or cause computer issues, depending on how the programmer or the software handles division by zero.

My main issue with it, aside from a question of what to do with zero armor, is that it makes no distinction between rapid weak attacks and slow powerful attacks given that both have the same armor penetration, which makes it just a different way of reaching a percentile armor system. I also wouldn't call it a quasi-percentile system; it is a percentile system as you are dividing any incoming damage by a number, it's just that you add in the armor penetration multiplicatively rather than additively, and instead of presenting me with an armor percentage (e.g. 90% armor) you present me with an armor number (e.g. 10). Telling me that I have 117 armor on my ship and making the default armor penetration be 100 just means that my ship has about 15.6% armor against 'default' opponents, and the bonus per point of additional armor penetration might look a little strange (and can vary wildly depending on ship type, which is also true for subtractive armor). Your way of determining the effect of armor penetration is a bit more complicated than simply subtracting the armor penetration from the armor percentage, since the effect varies depending on the armor value, but it's still just a percentage system.

Percentage armor isn't something that would stop me from playing, but it is something that I don't care for unless all the attacks which are affected by the armor happen to have a similar magnitude of power on the per-shot basis (for example, I don't expect the attack strength for various types of medieval heavy infantry as represented in a game to vary that significantly, so percentile armor makes a degree of sense there), but the indications we get for attack power in AI War all indicate that the attack power even within a given ammunition type varies significantly (for example, the lasers of a Regenerator Golem are three orders of magnitude stronger than those of a Mark I Laser Gatling, according to the wiki; something that can shrug off the lasers of the Regenerator Golem isn't likely to even notice Mark I Laser Gatling fire, but things which can shrug off Mark I Laser Gatling fire cannot necessarily ignore the Regenerator Golem).

If the reason for the armor re-work is just that it currently favors high-damage attacks over rapid attacks to too great an extent, I'd rather see mechanics brought in (or brought to significance) that help create a reason to favor the rapid attacks. Whether that means introducing a miss chance, making swarmers really swarm (but also drop like flies to weapons fire), making overkill matter more, or something else, I don't really care. But I do want to see the distinction between high rate of fire low damage attacks and high damage low rate of fire attacks preserved. Percentile systems, no matter the guise they come in, do not do this; rather, they just encourage the use of the highest DPS weapon available (granted, armor penetration values and the hull damage multiplier system currently in place will still affect ship selection with percentile armor systems in place, as will the durability of the weapon platform), because there's no difference between a 50 DPS weapon that fires 50 shots per second and a 50 DPS weapon that fires one shot per second under a percentile system, aside from overkill or too many targets. Since I don't think that AI War has too many situations where a faster-firing weapon actually allows you to kill a target significantly faster than a slower-firing weapon, I don't think that changes to the armor system would really affect that part of the balance equation too much (yes, the faster-firing weapons would become slightly less undesirable, since percentage armor affects all weapons equally, but the ships are still generally weaker and the over-time DPS curve for the high-cap ships is still generally more linear than the DPS curve for the low-cap and normal-cap ships).

[LaughingThesaurus]

I still don't like the idea of throwing formulas into the mix because it makes it harder to know exactly how to plan based on ship stats.

[TechSY730]

I still don't like the idea of throwing formulas into the mix because it makes it harder to know exactly how to plan based on ship stats.

Technically, any armor system will involve a formula. Even a "simple" subtractive armor system involves, well, subtraction, which is a formula. 
The only armor system which would have no formula is not having an armor system at all. 

I get what you are saying though; the math should be simple enough that you can at least guess correctly within an order of magnitude how much the reduction is by just glancing at the stats. So maybe crazy formulas involving, say, logs of ratios interpreted as percentages but with bounds, which might by statistically sound for balance, may not be the best choice from a game design perspective.

[Vacuity]

To me the most important goal should be relative transparency in terms of mechanics.  Realism is nice, but I have to confess that I'm not really asking for or expecting that in a game with Zenith Hydras and Neinzul Scapegoats.  The entertaining, interesting and challenging (frustrating?) gameplay is more important to me.

In terms of a system that is simple in all circumstances (different cap options, combat options and five marks of ships), the percentage-armour system seems a winner, despite not seeming particularly realistic.  It's easy to understand and easy to do calculations in your own head to work out roughly what's going on.

Following something like Diazo's baseline of 0/20/40% armour and armour piercing at different ship tiers does a reasonable job of "upping the stakes" when starships and superweapons are in the mix, and encourages a degree of strategy (which ship unlocks you want/need from ARSs and backup servers based on the AIs' unlocks), even if the tactics of engaging an armoured golem might not be much different.

Of course, as was pointed out nearer the beginning of the thread, the more factors there are in combat, the harder it gets to balance combat.  The simplest system of all is no armour system at all.  I'm not sure I want that system, but it would unquestionably make the game mechanics simpler and easier to balance.

[Ranakastrasz]

Yes, the intention is to simplify the armor system, so that, at the very least, the game developers can figure out how to balance it.

My post, actually, with the huge set of text, thoughts, possiblities, etc, I made a serious mistake. Namely trying to find a solution rather than stating all the problems.

~Damage currently has Four factors. It used to have one, back when shields caused a random factor. (which was, admittedly, complex, and even setting it to a linear damage reduction, would have been complex even then, if more reasonable)

Damage, Armor, Type Multipliers, Tier Multiplier
Damage and Armor, as well as Type Modifiers, were noticed.

Tier, however, wasn't, and actively sabotages any attempt to balance via using armor.

T1 to T2, damage and armor doubles. This retains the exact balance, tier to tier, which, ideally, is carefully crafted (and recrafted via feedback) until ships are all reasonably balanced, and behave as the developers want in combat.
However, Tier 1 ships attacking Tier 2 Ships, will be significantly less than 50% as effective, or indeed 25% as effective (due to doubled health AND damage) Because Armor, additionally, if on the target, cuts off more damage than expected. The reverse is true. High Tier ships attacking armored lower tier ships, well, Might as well pretend those lower tier *Armored* ships don't have any, as it is far to low for higher magnitude damages.

~I imagine balance calculations don't really take into account the massive difference between Tier 1 and Tier 2, which is where the Armor effect seems to disappear most.

Hence, Whatever is done, needs to take this into account, and ensure that any attempted balance is not just thrown out the window any time ships of different tiers fight. After all, I seem to recall that the developers wanted lower tier ships to still be viable, and required, unlike other games where you disregard inferior units once better ones are available.
At this point, unless fighting purely with, say, cloaked ships (which never have armor IIRC) The balance will generally get thrown out the window. Especially when the AI gets higher tiers, (Every game, homeworlds, say...) so you generally need armor piercing, and armored units are ineffective (as too much cost goes to armor negated by tier advantage)

Anyhow. Either having Armor work like, say, World of Warcraft, where Attacker Tier is taken into account (Defender armor multiplied by attacker tier, or similar) or using an armor system which disregards tiers all-together (%Based, with unarmored 0%, light at 20%, moderate at 50%, and heavy at 80-95%. Or any other system with similar tier independent properties)

~Having 0% armor at Tier 1, 20% at Tier 2, and 40% at Tier 3 is a horrid idea, because ships already gain doubled and tripled health, as is. Increasing that gap is part of the current problem, and this does not help.

Note that I cannot be certain about any of this, but It seems reasonable to me, at least.


~Core turrets seem to have the same problem, outclassing normal turrets almost universally (Unless T3, somewhat), rather than leaving lesser turrets a place in the game after aquisition, but that's a different story, and thread.

[TechSY730]

Not sure if ship "tiers" need to be formally recognized by the damage system. I would be happy to see ship "sizes" be formally recognized though, which would strongly (but not absolutely) correlate to ship "tiers". Even if it is not a pervasive in every calculation but be like hull type multipliers are (some units will get bonuses against some sizes of ships), that would be enough.

[Tridus]

I don't think the ship tiers are a huge problem. There's some standard scenarios, assuming Diazo's armor suggestion:

1. Normal fleetship A (T1) vs normal fleetship B (T2). Both of them have armor 0, nothing happens.
2. A vs armored fleetship C (T1). Assuming 50% armor, C has effectively doubled health against A and is probably going to win. The upside is that this is consistent for A being any type of ship, rather than being a huge problem for Gatlings while basically meaning nothing for ships with big guns.
3. A vs armored fleetship D (T2). They're both going to have the same percentage of armor, so their shots are being effected the same way. That is, they're both doing half the damage they should be. T2 is going to win pretty easily, which is the case right now anyway.

The reason why armor goes up with tiers right now is because damage goes up, and T1 ships are balanced against other T1 ships, while T3 is meant to be the same way. Giving T3 the same armor as T1 makes the armor essentially worthless as tiers go up. With a percentage system every tier can have the same armor and it's always effective against an equivalent tier ship (or any other tier).

Changing this will require some rebalancing anyway, so that's not something I'd be overly worried about.

[Ranakastrasz]

@Tridus.

Exactly why that solution is fine in my eyes. It solves the problem, and Simply.

[Toranth]

Subtractive armor and reality also have only a tenuous relationship with one another. It's just slightly less tenuous than the relationship between percentile armor and reality.

T1 makes the armor essentially worthless as tiers go up. With a percentage system every tier can have the same armor and it's always effective against an equivalent tier ship (or any other tier).

There are two big reasons I dislike the percentage armor ideas.
1)  Percentage armor systems produce silly results.  A fighter with 10% armor blocks 400 damage from another fighter.  It blocks 10,000,000 damage from an Artillery Golem.  Sure, the Artillery Golem still killed it, but the armor was suddenly 25,000 times as effective!  Realism aside, this result is unbalanced.
2)  In the absence of Armor Piercing, percentage armor is absolutely no different from a percentage health increase.  This is solvable only by making armor piercing a common feature, which reduces armor back to almost no importance again.

Diazo's idea of making base Armor and AP values for fleetship/starship/superweapon classes isn't bad - it's almost identical to making Ship Size a multiplier type, an idea I very much like.  But the size multiplier is a much more versatile idea, with more possiblity for interesting ships.  Diazo's idea could be implemented very quickly, but unless it's a temporary measure until ship sizes are implemented, I don't think it's worth doing.


Finally, the tier argument.  Mark II ships are already more than twice as powerful as Mk I.  So why is it bad that armor has an increasingly dramatic effect as mark difference increases?  A Mk I ship shooting at a Mk V ship SHOULD have severe disadvantages.  A Mk V ship should be better than the Mk I version.  And if armor is that ship's gimmick, then it darned well better be improving its raison d'être!
In this discussion, neither fixed or percentage matters; indeed only the fixed vs mark-variable point matters.  And I am strongly of the opinion that making armor non-variable reduces the advantages of a Mark increase.

[LaughingThesaurus]

That's why I threw armor piercing in with a percentage armor idea. Stuff like armored golems would probably have a lot of armor piercing.

[chemical_art]

The more I think about it, the don't want armor at all.

[onyhow]

I still think that new armor need to have multiple systems, or the balance between the various ship tiers would make no sense, like the current armor...

[KDR_11k]

So what's the point of armor except allow the attribute "armor piercing"? The old shields had a clear effect: Effective ranges of enemy units were reduced. That's now pretty much covered by radar dampening. Right now HP seem to be effectively the unit's armor, I'd say make AP units get bonus damage against higher HP targets.

[Tridus]

There are two big reasons I dislike the percentage armor ideas.
1)  Percentage armor systems produce silly results.  A fighter with 10% armor blocks 400 damage from another fighter.  It blocks 10,000,000 damage from an Artillery Golem.  Sure, the Artillery Golem still killed it, but the armor was suddenly 25,000 times as effective!  Realism aside, this result is unbalanced.

You're not wrong, but that's hardly a new problem. Why does a Superfortress do 20 times more damage to a fighter than a bomber and *100* times more damage to a fighter than a scout? Why does a Zenith Hydra (with the same ammo) not do the same thing?

That result is no less goofy than artillery golem hitting armor, but it's not imbalanced. Since the artillery golem is still killing everything its shooting, I don't see it as a problem either.

(And on a similar but opposite direction: why can a mk IV Protector Starship block 4 million DPS from fortresses (more than a mk III can put out), but only 40,000 DPS from laser gatlings (1/4 of mk I cap DPS)?)

2)  In the absence of Armor Piercing, percentage armor is absolutely no different from a percentage health increase.  This is solvable only by making armor piercing a common feature, which reduces armor back to almost no importance again.

Also true.

Finally, the tier argument.  Mark II ships are already more than twice as powerful as Mk I.  So why is it bad that armor has an increasingly dramatic effect as mark difference increases?  A Mk I ship shooting at a Mk V ship SHOULD have severe disadvantages.  A Mk V ship should be better than the Mk I version.  And if armor is that ship's gimmick, then it darned well better be improving its raison d'être!
In this discussion, neither fixed or percentage matters; indeed only the fixed vs mark-variable point matters.  And I am strongly of the opinion that making armor non-variable reduces the advantages of a Mark increase.

Not really. Most of the ships with armor get very small increases per tier right now, vs doubling of the firepower of other ships that also gain a tier. It's not really an improvement. Just doubling the HP and leaving the armor at 50% or whatever means the armor is blocking twice as much damage before the ship dies, which isn't a lot different than what you get if you increase it by 500 a tier and the stuff shooting at it had its damage increase by 5000.

Aside from the complexity of making flat value armor work in a game with all the different modes this one has (and code simplicity is a big selling feature for whoever has to maintain it), it leads to armor being totally meaningless or totally crippling when different tiers of ships fight. Enough armor to do serious damage to a fighter's DPS isn't going to really slow down a bomber, let alone a starship, but it will massively hinder gatlings.

I'm not really sure how much better that is.

[Ranakastrasz]

The problem (I see) with Tier 1 being more ineffective than scale-up of damage and health alone would cause, is that, as said, the Developer wants, or at least wanted, all ships to be useful, even when higher tier versions are unlocked. With the dramatic scale-up causing lower ships to be significantly more obsolete than before, well, it seriously breaks that concept.

Also, a linear armor system is still unrealistic regardless. If health is how much internal damage a ship can manage, and armor is armor, then I would expect any and all blows to damage armor, eventually wearing it away, or else taking out, say, 20% and then coring the ship. This is complex and adds little to the game, and makes armor more of another layer of hitpoints. That is the other thing I liked about the old shields. It made more sense, or rather had less real-world correlation.