New armor system?

Well it feels like this is the final nail in coffin for a goal that were set long ago; to avoid arbitrary target restrictions and damage modifiers. Once you have done it once, it seems like a slippery slope were you might continue to do it more and more while the restrictions and modifiers are very hard to remove. It feels like a 'cheap' solution.
I have some ideas of some all-encompassing theoretical designs that can avoid target restrictions and damage modifiers but I struggle in vain to find a game that actually have very fast airplanes, no target restrictions and no damage modifiers.
I think that it is good to have a framework in place for mods but I hope that the current damage modifiers are just hotfixes to current balance issues like moving transports from basic air factories to advanced air factories.

That's kind of lame. It was kind of cool that missiles that missed would end up going somewhere.

It was very very bad that fighters would continually fire missiles at invalid targets though (like themselves)

No, I'm happy with this.

This actually makes a lot of "extra" units possible, that weren't previously so.

That wasn't a 'hotfix'. If you read my reason as to why it was moved, it has little to do with balance and much more to do with game flow.

So, I'm curious. How would you fix the issue of fighters firing at a flying unit coming off a factory, it hitting that unit, the projectile does area damage and damages the factory?

I actually can tackle this question, with no offence to you Scathis. Issuing this as a fix earlier rather than later itself is appreciated, nobody wants to wait the few weeks for a "proper" one when they can get a working fix right now.

-One can prevent fighters and their weapons from attempting to target "low", anything below themselves, they can target just low enough to target other fighters. I think there are weapon attribute hooks in the unit configs showing the area of swivel a unit's weapon has. Then again, this also prevents fighters from shooting landed fighters (which really is a defender's air buff anyway)
-One can reduce air aoe on fighters missiles to 0. Then again, the stray missiles will still kill the factory if there are enough fighters.

Then again, if that first fix I described is possible, I would love it as it at least lets the fighter get off the assembly line before being destroyed. Here is another idea, can you give units mid-lathing a different type of armor, that protects them from damage, and their armor changes as they become a finished unit? That sounds like it gets more complex though.

Lastly, right now gil-es can sweep a base killing all of it's units and structures uncontested because they kill everything so fast and out of range of retaliation Do you consider this a possible fix for them, reducing their building damage by 80% and their tank damage by 40%?

It's not an issue.


You have no anti-air. The enemy has fighters over your air factories. The fighters are free to strafe your factories. The fact that you're still launching air units means you are mentally deficient. The fact that the missile hits the factory encourages you *not* to build air units that are going to certain death, but instead to protect your air factories with AA so they never end up being camped by enemy fighters.

How do you deal with these edge cases:

• Planes don't fly on a single plane. They go up and down, especially gunships. By limiting their pitch they might not fire on something that looks like they can be fired upon.
• We have round planets. Smaller planets need large pitch ranges to be able to target enemies.

Fighters need to take out swarms of bombers.
How do you deal with this edge case:

• If you raise the damage high enough to kill a bomber in 1 shot, and they still aren't effective enough, you need to fire more missiles. More missiles cause server simulation, bandwidth, and graphical performance issues.

So, use the armor system more? Isn't that what you are advocating avoiding?
If you do this, there will still be a point in which they can be fired upon and all you're doing is delaying the inevitable.

Walls are the counter to sniper bots. They don't fire on an arc and therefore can't fire through walls.
However, they are on my radar to look at, as is most combat balance.
Changing their damage via buildings and tanks doesn't make much sense. The idea of the Sniper Bot is to kill a unit one at a time with a slow rate of fire. So if I did use it, I would just change it's structure damage. That said, I don't plan on using the armor system more than it's currently used.

Decreasing the speed of the transport, the acceleration of the transport and increasing the time to drop-off units also greatly change the game flow and the viability of hotdropping units in the enemy base for example. It also increases the risk of using transports and gives the enemy more time to intercept and longer time for ground AA to fire at the transports.

I agree with your concept of pacing where more options and gameplay opens up as the game progress. However as a matter of expansion I don't think transports matter that much. Mexes are so cheap that the big bottleneck is energy at the start and you can easily get to t2 really fast to start making t2 mexes which means that broad expansion isn't that important early on.

1. Firstly I think that it could be fine if fighters can hit ground targets.

2. Secondly there is very little connection between ground DPS and AA DPS. You can basically set air plane HP and anti-air DPS regardless of everything else. I can only think of the commander and the Firefly which uses a weapon that fires at both air and ground targets. This means you can decrease the HP of all airplanes by 10 times for example while also decreasing AA DPS by 10 times and the balance of the game remains almost unchanged. Suddenly fighters deal 10 times less damage to factories.

3. Thirdly, I think that fighters should have much slower acceleration and should preferably bleed off speed when they turn. This makes air combat much more tactical as retreating requires planning and dedication where you just can't turn on a dime and retreat. Also if the fighters are encircling an air factory or attacking some ground forces, they will have to sacrifice speed to do so which gives the defending player more time to react and intercept the enemy fighters which are basically sitting ducks as they are making low speed runs against ground targets.

4. Personally I think it is a problem that the hitbox of air factories extend above the hitbox of the airplane under construction. If the hitbox of the air factory is smaller so that the airplane isn't protected by the air factory hitbox as it leaves the factory it also makes it less likely that the factory will be hit.

5. Spinal weapons. This is mostly because I like dogfigthing. If some fighters only have non-leading forward facing weapons they have to fly directly at their target which puts them at risk as they dive closer to AA and have to stay on the tail of enemy aircraft. You can also differentiate fighters that are fitted with missiles and interceptors without missiles. Of course this opens a can of worms in how you should handle targeting, dogfighting and such.

6. High flying aircraft. This would be another all-encompassing change that could allow any ground or air unit to attack each other.
If planes fly at high altitude at high speed they will be out of reach of most ground units. If they also use spinal weapons, slow down when they turn and cost a lot of resources, attacking ground targets puts them at great risk as they become highly vulnerable, low and slow flying targets that even ground units can hit.

For less engrossing changes, I would propose 2(decrease air health and AA DPS), 3(slower acceleration for fighters and bleeding of speed when they turn) and 4(reducing the size of air factories hitboxes) where fighters lack the ability to attack ground targets which I guess could be implemented without much engine overhaul and balance changes.

Edit:The Firefly can also attack ground and air targets.

I don't quite understand. Do you propose that PA should use damage modifiers or that it is good that mods can use damage modifiers?

Actually, that's exactly the point. I am in no obligatory "camp" in this idea, what works works. I praise this fix being both simple and quick.

If it works, it works. I do not mind it. Generally, it's rather nice of you to not consider using it in widespread damage diversification though, more people rather it that way I suppose.

I mean, if you use it literally to prevent shots from doing damage at all where they shouldn't, then it's not so much armor as it is "stray-shot-prevention", it literally is only used to keep it from damaging what it already shouldn't.

I mentioned still fixing it differently later, because them attempting to fire upon unbuilt fighters still in the factory does look odd, especially when the factories actually aren't dying from it. You fixed 99% of the balance flaw from the mechanic this patch. Really, it is up to the player to build missile-flak to keep fighters off their air factory, and that is an entirely reasonable requisite for a player to get fighters back into the skies.

As far as sniper bots, it seems their damage in groups picks through walls so easily though. I am sure you are right though, they can have other numbers tweaked. Right now, I use them, and I am not afraid to advocate their nerf, they are like slammers that don't dwindle in numbers because they kill from out of harm's way. Generally, there are plenty of ways to play with the numbers, and a lot of things won't be balanced until the numbers are tweaked very specifically.

But the summary to all of this is, yes armor was a very useful engine hook and a very simple bug fix in this situation. I don't doubt it's use in player mods later or bug fixes later even.

I think it is good that PA uses damage modifiers.

T2 air health can be scaled back down to be more similar to t1. E.g. t2 bombers are scaled so they have thicker armor against land based AA, but are still vulnerable to t1 fighters.

In either vanilla or modding, a Gatling Gun vehicle - an anti-bot vehicle. You can adjust it's dps to be specifically good against light armored units (like bots, light vehicles like the spinner, etc.) but not be good against heavy armored vehicles or structured, and you don't have to adjust a whole lot of other values when you add it into the game.

In other words, it's not just "maximise DPS for metal, maximise health for metal, build two units only". You actually have to think strategically about what performs well against what and build mixed unit armies. "If I just build tanks, and then they run into gunships.... No, I want my tanks to be able to shoot gunships too" is just lazy player spamming unit because best.

I know people hate the rock-paper-scissors idea, but rock-paper-scissors is one way to get unit variety, and is in fact already present in PA.

Hummingbirds work just fine against kestrels and hornets. The problem is that peregrines are invalidating for them and ground AA is crap.

When I have 50 tanks and a commander marching towards an undefended commander that isn't being guarded by any bots or tanks or defense towers and have 10 fighters standing by along with ground AA and the enemy can just stop it all by using 4 kestrels and 3 peregrines to kill the com which is in uber cannon range of the enemy com and was previously at around 1/5th more health than it it's ridiculous.

Idea: Make comsplosions deal 8000 damage. Should fix part of the problem.

Yes they do.


The problem is that Peregrines have to do massive damage against the high health air units, and the high health air units have high health precisely so they can tank damage from AA. Which makes it harder for hummingbirds to kill peregrines.

However if high health air units tank damage from AA due to damage mitigation (i.e. they have normal health), you can make all air craft have normal levels of health again. T2 fighters being more resistant to ground based AA, but still vulnerable to t1 fighters.

Bombers are already very vulnerable to fighters since they don't fire back so why would this be needed?

Why should a Gatling Gun vehicle do more damage to light armor? This is future robotic warfare. What applies to infantry today might not apply to robots. Also how do you know if a unit is lightly armored? How would the player determine the armor type without reading a label or checking a unit database? It might be intuitive for you but that might be because of your preconceived perceptions. What you think is logical might be illogical for someone else.
Also if a unit have low health that already means that the unit will die quickly even from light weapons.

If you want to see rock-paper-scissors design without any arbitrary damage modifiers, you should try Zero-K. You don't need damage modifiers to enforce RPS balance. There are plenty of other stats from which unit counter-relationships can emerge.

This is the worst use of damage modifiers. I can support broad armor types and damage modifiers if they are consistent and easy to discern from visuals. However applying them willy nilly to fix specific balance issues is inelegant, arbitrary, increases the complexity and breaks immersion. When one of the most powerful weapons like Overcharge in FA or the Ubercannon in PA is used against the ACU or a commander and does almost no damage, that breaks the immersion and goes against the intuition and logic of the player.

If the special damages are applied from case to case basis it makes it much harder for the player to know how their units will perform.

Uniform damage is elegant in that the player can abstract unit interactions between different unit types, they can trust their intuition and logic and do internal calculations and expect how their units will perform. Special damages breaks that if it is applied inconsistently.

It's almost like planes, gunships and bombers, through intrinsic characteristics, aren't useful on small moons and asteroids. You claim this is an edge case problem. I don't; I would ask why aircraft should be 'viable' in all situations regardless of terrain, planet size and type?

First, it's interesting that you consider massed bombers an 'edge case'. Second, it is also interesting that your only suggestion to an 'edge case' overpowered unit swarm is that, due to engine limitations, you must counter it with another overpowered unit (which you are also likely to swarm).

Question: At what point do you consider a number of bombers to be an edge case swarm? At what point during gameplay is it possible to achieve this singularity and what causes this number of bombers to become something more than just 'some bombers'?

Are you positive that fighters are the correct answer to a swarm of bombers?
Should fighters be the singular answer? Are there no other options to explore?

You have previously mentioned that it is important to check your premises. I would be grateful if you could elucidate on what your premise is for this situation.

Oh God, this argument is the dumbest. This is a game with bots. BOTS! A unit which would never really exist because it's inferior in every way to the design of a tank. They'd be sinking into the dirt on most planets and never get moving at all because their weight distribution is ****.

Because shifting goalposts like the hitting angle of an aircraft is inevitably going to cause problems down the road, somewhere else. Say, an aircraft moving over a hill and for a second, it behaves as if on a small planet and the system throws bugs.

Armor systems ensure you get a nice predictable result throughout. It's symptom fixing, but in this case fixing the symptoms is a lot easier than fixing the problem.

Fixing the problem is a time-consuming, zero-guarantee attempt that may or may not be fruitful, and may cause a ton of crap elsewhere.

From an engineering perspective, it's simply best to pursue the armor system and concede that tiny bit of accuracy.

Just think why it is that t1 bombers are outclassed so much by t2 bombers, and the same for fighters?

T2 bombers are very vulnerable to fighters. Peregrines. Not hummingbirds. And they are very not vulnerable to t1 antiair

Suggest how you would make it so that t1 air isn't invalidated, and likewise t1 anti-air is still relevant, after your opponent has stepped up to advanced air?

Are you kidding?

Armor is based on thickness.
It is easier to penetrate something that is not thick.
Doxen move faster, they accelerate faster.
F = mass * acceleration.
Doxen cost less metal.
Therefore, they have less mass.
Armor is made out of mass.
Less mass = thinner armor = less energy to penetrate.

All of those facts point to "light armor".

If it's got less health/is faster, that suggests it is light armor.

A Gatling gun does not do "less damage" to thicker armor. Armor is based on thickness - it's not an energy shield. If the armor is less thick, it requires less energy to penetrate, so damage mitigation is not as high. Armor is damage mitigation. That is how it differs to "health". The armor does not last as long before it is penetrated.

Before you ask why a million times, think about it from a materials engineering and a real world physics point of view.

Armor always will be based on hardness and thickness. It has been based on both of those for several hundred million years.


And I'm sorry, but if the player is illiterate and doesn't read the information in the game, but just blindly builds, they'll still figure out what it's good against/bad against by how well it performs, so it's rather a moot point.

There is no display of stats in PA. You know bots have less health than tanks because bots die 1:1 in bot duels, and 10:1 in tank duels. That's really obvious to an observer who plays the game.

It is a sake of example unit, but nonetheless, I don't want to spend hours and days and weeks adjusting every single health and damage value in Planetary Annihilation and checking balances, etc. just to mod in one unit.

Let's say my antibot vehicle kills bots faster than tanks (A tank kills a bot every 4 seconds).

To acheive that, I have to give my 'anti-bot unit' a DPS higher than 21 dps. By doing that, in PA's current system, I've also made an anti-bot vehicle which is a better anti-vehicle unit than a tank. It is universally true that players will maximise dps/cost. My 'anti-bot unit' can't be much more expensive than a tank, because of opportunity cost. Let's say dps is 30. It can't be more expensive than 215 metal, because at that point, you're just better off building tanks to fight bots. It has to be more expensive than 215 metal, because before that point you're better off building superior anti-bot unit.

Sure, I could f*ck around and adjust a whole load of values on it and every other unit/structure so it fits in exactly where I want.

At the end of the day, all of this is circumvented by simply saying "Does 25% damage to armored units and structures" in the code, writing a description that says: SK's AntiBot Vehicle: Anti-Bot rapid fire chaingun, and leaving it at that.

Armor is a realistic mechanic, and it is WYSIWYG. Zero-K is limiting its simulation. I can think of many more good RTS's that use damage modifiers and achieve good balance. Like Scathis said, and I don't often agree with him, but armor deals with the edge cases. Armor deals with the balance between units that can't be isolated from the rest of the system.

I'll close by saying: the anti-bot vehicle is an example only.

Kiwi, Since when was Planetary Annihilation about accurately simulating real-world physics?

Actually, that's factually incorrect. Legged walkers are an incredibly efficient way of moving over extremely rough and uneven terrain. There is a reason that millions of years of evolutionary processes have converged upon legs as a good idea and it's because tracks or wheels are extremely inefficient (both materially and from an energy standpoint) when moving over anything other than a flat surface. The reason we don't use them is that walking is actually a surprising technically difficult feat to accomplish. It's only very recently that we have had the kind of cheap processing power, servo accuracy and high performance materials to give proper adaptive walking systems a serious look at. There is a reason that the US military is putting large amounts of money into companies like Boston Dynamics in order to come up with actual walking robots. I would say that their results are both incredibly cool and slightly terrifying.

However, all of this is besides the point. Godde has made some interesting points that a lot of the behaviour we would intuitively expect can be developed from playing with the appropriate levers in the simulation system of projectile motion, unit size and unit health. I would be interested in seeing how far such things could be taken.

They're looking at legged vehicles to carry supplies over rough terrain. They're not building mechs because mechs are stupid. Mechs have more weak points. Mechs have inferior fields of fire. Mechs have inferior weight distribution leading to them requiring much stronger surfaces compared to a tank of similar weight.

A tank will bring more armour while being able to cross more surfaces than a mech. The tank would also be faster. Some tanks have better surface area to weight ratio than a standing human leading to them traversing terrains we could not.

Simply put, the idea of making arguments based on "realism" is incredibly stupid and everyone should stop doing so immediately.