So, this whole armor thing

[chemical_art]

Just leaving armor as is doesn't really hurt the game. It's just a bit of a drag on the overall quality I feel.

I disagree with this. In situations that are meant to rely on armor including armor ships and anti armor units, they simply feel inferior compared to their non-armored versions.

And when attacking AI HW, when enemy ships have MK V armor while attacking ships are often averaging MK ~II-III then it beomes miserable depending on ship compensations.

To put it another way, armor by simply existing (and having other ships rely on it) causes those units that depend on it to be hurt. And there are plenty of edge causes that I don't think enhances gameplay but hinders it.

[LaughingThesaurus]

You seem to be getting at "yeah, the current armor system DOES have problems" and that's what I was saying as well. It does have problems, it doesn't really kill the game, but it brings the game down. Because the system causes the problems that it does cause, it needs either a solution that fixes all the problems, or it should be removed so it isn't hurting the game at all.

[TechSY730]

If armor is removed entirely (and that is a big if), instead of trying to find a new role for every fleet ship type designed around armor or countering it, the ones who have "obvious" repurposings can get those, and the rest that can't find a good, non-silly type "repurposing" or "gimmick" to take their place just "go away". They wouldn't really go away truly; save games where you have one of the removed types unlocked will remain unlocked, and existing ships of those types will still be there too. However, they wouldn't be availble as options to players or humans for subsequent games or unlocks. Basically, it would be like that neinzil carrier type bonus unit that was scrapped during the beta.

I know that removing a bonus ship type that has had a history is a pretty big move, but TBH, the game is already on the edge of the point where the sheer variety of bonus ships is starting to hurt the quality of the game rather than be an asset, so I don't think the "metagame" will care very much if 1 or 2 less popular ship types go away.

The only ones I can think of that don't seem to lend themselves to an "obvious" re-purposing are the anti-armor ship, the zenith polarizer, and the human hardened forcefields.
Either the polarizer or the anti-armor ship I guess could be repurposed to have implosion artillery ammo*, whichever seems more thematically appropriate. The other one can be "phased out" as described above.
As I mentioned previously, the human hardened forcefield generators would be removed, and either the regular human forcefield cap doubled or the regular human forcefield HP doubled (I would lean towards the doubling of the cap), but not both obviously.

The autocannon minipod could just get a moderate DPS buff or a small DPS buff + small HP buff.

The armor-rotter, as suggested by others, could gain a new DoT mechanic.

The armor fleetship could just gain more HP.

The space tank could be repurposed to be a "slow-mover" like the armor fleetship is, but with focus on DPS instead of HP. (Or the armor fleetship could be sped up to match the space-tank's speed)

The armored golem and heavily armored AI superweapons would just get a simple HP buff.

The planetary armor booster could instead do something like reduce all direct damage taken by friendly ships in the system by 25% or something like that ("special effect" damage such as attritioners would be unaffected)

The planetary armor inhibitor could instead do something like increase all direct damage taken by enemy ships in the system by 20% or something like that ("special effect" damage such as attritioners would be unaffected). Not sure if it should keep the enemy forcefield disabling properties or not. It's a great side benefit, but a rather unintuitive one as it is a remnant of when they were called planetary shield inhibitors.

The neinzul youngling nanoswarm could just have the other "debuffs" buffed slightly, or it could be given a pretty darn small DoT (the DoT would need to be stackable in this case, and there probably should be a max-DoT rate as well proportional to the max-HP of the affected ship. Same would go for the armor rotter too)

The raid starship is a little trickier, but a moderate HP buff, but let it keep its radar dampening, hopefully would keep its dynamics mostly intact.''

Any other armor-centric ship types I missed?

*BTW, I still think that ships with an implosion artillery ammo type need a hard cap on min and max damage, regardless of percentages, though those bounds can vary based on ship-type to ship-type of those with implosion artillery ammo type)

[Martyn van Buren]

Hey, I feel like walking into a hornet's nest and I have a very simple suggestion which could be totally irrelevant. At any rate, I am not great on the numbers side of the game and am increasingly leaning toward the option of just getting rid of armor. But I have an idea for a new gimmick to replace the armor/armor piercing dynamics of ships that rely on it. It has very little to do with the current system but it would add a new dynamic to replace the one we would lose if we scrap armor.

It seems to me like the interaction between marks is one of the more interesting things that the current armor system does, creating some large ships that can shrug off a lot of damage from smaller/lower mark ones.  How about replacing this with an ability, called "armored" or something else, that greatly reduces damage from lower-mark ships, perhaps by 10-15% per mark advantage? It would want to be applied only to a very few ships, probably those that have high armor as part of their current gimmick.

Meanwhile, I think armor piercing would want to become a damage bonus against higher-mark ships, probably flat (30%? 50%?), both because it seems deeply counterintuitive to have your damage bonus go up as you face tougher ships, and because having it scale would perversely make Mk I ships of this type the most powerful. Alternatively, it could just scale based on mark regardless of the mark level of the attacking ship. At any rate, I think this would create a pretty interesting dynamic, a select group of seriously fearsome ships optimized for killing your treasures.

[LaughingThesaurus]

I think the ships that rely on armor gimmicks could just get an ability added that gives them their gimmick, though. So, for instance, give the Hardened Force Field this: "Takes 20% damage from all sources" Anti-armor ships can get some special ability that lets them tear through high HP ships a bit more easily (Possibly: Does 0.2% of the ship's max HP in damage in addition to its regular damage)

[Aeson]

1. I think removing (or changing) the current armor system is going to take a lot more work than most of you seem to think it will. I created a spreadsheet which contains the percentage damage which gets through armor for about 70 fleetships and starships (all at Mark I, and none of the modular ships were included) targeting one another. You can get the spreadsheet from here:

    https://www.dropbox.com/s/6wpm6epbobj7z1w/AIWarArmorEffectiveness.xlsx

You will need something that can view Excel 2007 format spreadsheets. Triangle ships are highlighted in blue on the target-side and are the first three columns on the attacker side. Percentages where you have bonus damage involved use red text instead of black (though please note that I used the bonus value listed in the exported *.csv file, which only contains the highest bonus; therefore, any ship which has varied damage multipliers may not have had armor correctly accounted for in the percentage of the damage which penetrates armor against things which it has a bonus against - the percentage can only go down, not up, if something like that happened).


2. I don't think that the current armor system is particularly bad. However, if you really want to remove it, which ship are you going to use for your baseline? Fighters? The average damage percentage dealt by a fighter in my spreadsheet is 95%, one of the highest in the table. What about Autocannon Minipods? The average damage percentage for those is only 52%. If you use the Fighter as the basis for scaling health after getting rid of armor, Autocannons essentially get a 100% damage increase (more, really, since it's down at 20% damage a lot of the time).

Then you'll ask "what if we used the average damage percentage that gets through to a ship instead of the average damage percentage dealt by a ship" for balancing. Well, we have Polycrystal targets that range from 75% average damage getting through up to 100% average damage getting through. As an example, 100% of fighter damage gets through Bomber Starship armor, but on average only 84% of damage from all sources checked does. How much should fighters be affected by this? Should they be pulled towards the 84% effective average? Get better multipliers to keep it closer to 100% current, bearing in mind that increasing the multiplier will affect all the other Polycrystal targets differently? Bear in mind, also, that we currently have ships that run the full range from 20% of potential damage all the way up to 100% of potential damage getting through the armor of most of the ships in the spreadsheet, and I also didn't include all ships or all marks in the comparison.


3. The stuff about armor being replaced by abilities: I really don't like that. I have no desire to have an 80% damage reduction or something like that hidden inside the abilities list.

[Histidine]

If you just remove armor because you are too damned lazy to actually put down factual details, that's on you, but you will piss off people who actually want to see this done right.

You didn't even bother acknowledging Tridus's response to this point.

No amount of data points has any meaning if the formula/algorithm they go into or come out of doesn't.

[Histidine]

_{double post because I think it works better this way}

    https://www.dropbox.com/s/6wpm6epbobj7z1w/AIW

Hey, thanks for this! I'm having a lot of fun playing with it, even if I have no idea how to interpret the results 

Some random facts I've picked up:

• The average standard deviation of the %-of-nominal-damage-dealt value is 14%. However, it ranges from 2.4% (the smallest value that isn't zero or Spire Starship) for Zenith Devastator to 38.7% for Autocannon Minipod. Other notable high and low scorers are Youngling Shrike (34.6%), Raider and Armor Ship (33.5%), Grenade Launcher (30.3%), Plasma Siege Starship (2.6%), Spire Mini Ram (5.3%), and Bomber Starship (7.7%).
  Fighter, Bomber and Missile Frigate are 14.7%, 18.6% and 18.0% respectively.
• Thanks to the way the math works, only the top 4 most-affected-by-armor units (AC Minipod, Z Mirror, Z Paralyzer, Yng Shrike) have any targets to which they do more than 1 sigma of damage above mean (29 each, from a list of 74 ship types). Of course, against unarmored targets this translates to up to these ships doing up to 1.65x their mean damage.
  
  ...That would be a lot in any other game, but in AI War we've needed 6x multipliers just to get our RPS triangles to work as intended.
• Of the 74 ships listed, only 17 do more than the minimum 20% damage to Raid Starships. Twelve of these ignore armor anyway.

2. I don't think that the current armor system is particularly bad. However, if you really want to remove it, which ship are you going to use for your baseline? Fighters? The average damage percentage dealt by a fighter in my spreadsheet is 95%, one of the highest in the table. What about Autocannon Minipods? The average damage percentage for those is only 52%. If you use the Fighter as the basis for scaling health after getting rid of armor, Autocannons essentially get a 100% damage increase (more, really, since it's down at 20% damage a lot of the time).

So divide Autocannon damage by x, where x is a value from 2 to 5 

But it's actually not so simple, even if we disregard the whole mark level issue. 
For example, Raiders do an average of 69% damage, but that figure includes 20% to a bunch of bonus ships and 89% to Fighters and Missile Frigates. Now, those two triangle ships are easily way more common than the bonus ships, so being able to do a lot of damage to them is more important than to those bonus ships. As such I'd say x should be closer to 0.89^-1 than 0.69^-1 - but how much closer?

And what happens when you take into account the fact that one of the 20% targets is the Bomber, the third triangle ship and one of the most dangerous?

(And of course, all this assumes that the ships are balanced to begin with and things will turn out alright if we make sure the new system produces similar values to the old one. Hah!)

[Draco18s]

Thanks to the way the math work

So basically, leaving everything alone for 95% of units* and increasing HP by ~20%, would leave things almost exactly the same.  The outliers are on a small enough degree that its unlikely to change the win-loss ratios that much.

Per-mark isn't likely to change much either, as armor increases aren't going to change the tables between mis-matched marks units enough to be noticable (i.e. if the kill to death ratio between two units of the differing mark is x:y, removing armor is going to change that ratio by less than 2%, because the damage and HP increases from mark are so much more significant).

*Massive-amounts-of-armor units that have it as their Thing (hardened force fields, armor golem, etc.) being outliers that would be rebalanced separately.

[keith.lamothe]

While I don't think the intent of any of you is to filibuster on this, as a whole the community is doing so most effectively  

Which is fine: the old energy system sat wrong with a lot of us and we had occasional megathreads on the subject but we could never really agree on a solution.  Until we did, of course (and while the new energy system isn't massively better, it is better and deals with the main problems of the old one, imo).  But if we'd just thrashed around reworking the system with every one of those threads... I don't think that would have been good.

Here the only approach I see (personally) as being a big step forward (rather than a mostly-sideways step into other problems that are roughly as bad) is replacing armor and the current ship-type vs hull-type systems with a weapon-type vs hull-type+hull-size system.  Those multipliers would reflect an aggregate of the conceptual stuff like:
"how fast does this target maneuver"
"how fast does this gun track"
"how much ECM/ECCM is involved on each side"
"what is the hull made of, and how does the weapon interact with that"
"how big is the target"
"is this a light-speed weapon, a projectile weapon, or a guided weapon"
"does the weapon need to make contact with the target, or does it have a standoff range,  and does a proximity hit do less damage"
"how armored is the target, what is that armor made of, and how does the weapon interact with that"

And so on.  Lots of stuff that I'd want to model directly (to some degree) in a tactics game, but we can't model at AIWs scale.

Naturally that's like a rebalance of the entire game, but so it goes if we want to really move forward on this, I think.

[Draco18s]

Naturally that's like a rebalance of the entire game, but so it goes if we want to really move forward on this, I think.

The friend-of-mine-who-got-me-into-AI-War has been asking why they don't go the "bullet type" vs. "hull type" kind of system anyway.

Not saying I'm advocating for or against such a system.  There are benefits of it ("Flamewave can't hurt Polycrystal, ever") but also drawbacks (massive refactoring and rebalancing involved).

[Diazo]

While I don't think the intent of any of you is to filibuster on this, as a whole the community is doing so most effectively  

I think this is a symptom of the fact that a small change won't fix the issue so everyone has their own preferred small change with the problems that change has, and argues against other peoples suggestions based on the problems that that small change has.

Here the only approach I see (personally) as being a big step forward (rather than a mostly-sideways step into other problems that are roughly as bad) is replacing armor and the current ship-type vs hull-type systems with a weapon-type vs hull-type+hull-size system.  Those multipliers would reflect an aggregate of the conceptual stuff like:

And I would also like to see this type of rebalance go through, as I don't think anything smaller with actually fix the underlying issues brought up in this thread.

In the meantime, I feel other things need dev time first, such as the hacking mechanic that is currently being implemented and the AI HW's brutal picks being too brutal.

D.

[LaughingThesaurus]

Keith, I'm in support of that idea as long as you can work out a way that this system will be transparent to the player. A whole bunch of constantly changing damage multipliers based on some obscure factors and formulas will make it harder to pre-plan what you'll need to do.

[KDR_11k]

Weapon types vs hulls is the easiest to learn, the learning curve for AI War could use some reduction anyway.

My problem with all "bigger ships get more armor" ideas is that the current situation is that bigger ships are better anyway. Most high cap ships are considered UP while most of the OP ships are low cap. This gulf needs narrowing, not widening. There are already enough mechanisms that favor low caps over high caps. An evasion mechanism that makes swarmers less vulnerable to stronger weapons might be useful for the balancing but not something to weaken them. I think the innate overkill advantage of swarmers just isn't there as long as swarmer caps are like 2x triangle caps. Maybe they need to be simulated as some sort of composite unit where every swarm unit consists of ten ships or so and one attack can only destroy one ship (unless it has AOE).

[Draco18s]

My problem with all "bigger ships get more armor" ideas is that the current situation is that bigger ships are better anyway.

AND favors the bomber as the counter unit of choice.