POLL: Balancing Air - Continued!

Remember what I said about making a mess? When an idea doesn't work so you add on more rules and more exceptions, that's called making a mess.

Supcom2 also tried this by making an umbrella shield that "only blocked artillery". Hahaha. It was pretty damn good defending a valley and attacking uphill last time I checked. It could block everything in the right situation, clearly not matching the intent of its design.

Keep it simple. An idea should do what it's supposed to do. If it doesn't throw it out, and get another idea. An RTS world has no shortage of solutions. For example, an EMP could paralyze a target's ammo regeneration. Only the bomber has energy, and without energy it can not continue to engage. The bomber deals its first attack(for a unit that excels in first strike this is OK, don't throw a fit over a unit succeeding at its role) but it is ultimately defeated in a convincing fashion.

Keep in mind that bombers are getting AI improvements to spread out more and optimize their damage. It's not like bombers (or any fixed wing craft) ever converged that well anyway, as they tend to attack in a LINE rather than any kind of blob. After the first strike they're all over the place, which is exactly the antithesis of how a flak defense excels.

Gunships are the air units that stack up into blobs. They are probably the most consistently stacked unit in the game, even if they happen to lose their stacking ability. Clearly flak is going to deal its best damage against gunships, so by buffing flak against bombers, what you really do is you crush any hope of using a gunship.

If an idea is doomed to not work the way it should, it should be left to die. Flak will never be the best choice against bombers, at least not without wrecking many other aspects of aerial warfare.

You picked that one out of its context quite nicely. It was a response to the statement that shields in SupCom were supposedly designed to defend against artillery strikes but were setup in such a way that they affected much more of the game than just artillery.
Since it would have been possible in the engine to setup shields without upsetting everything else not doing so must have been a deliberate design choice. So the artillery shield was the simple idea, and letting it blocking everything and its mum was adding in more rules - thus creating the ensuing mess.
I can't comment on SupCom2 besides that they might not have clearly distinguished between artillery and normal projectiles - again nothing to do with the concept in and of itself.

Hahahahaaa

Try not editing quotes to prove a point please. Wasn't making a point, was explaining MY view not speaking gospel.

Awh thanks man.

"Blocks all projectiles" is ONE rule. Why? Because the human mind remembers it as a single rule, with no exceptions.

"Blocks ABCD but not EFGH" is a pile of rules. Why? Because if you do not memorize all the weapon types that do or don't work, you will not know how it works. This is made even more obscure because the shield is a gigantic bubble, which is now blocking some weapons but not others. It is a visual paradox that makes the situation worse.

The Bomb Bouncer was just a directional shield, it didn't have any special rules.

Well it was supposed to be a shield that only protected it's self from above.

But that doesn't work in valleys.

It seems I didn't make myself clear. The point I wanted to raise is that the final shield design in SupCom was not due to technical limitations but deliberate design choices; not about the actual artillery shield on the xBox.

In regard to artillery defense shields being counterintuitive in concept: If shields were originally designed to deal with artillery only, it would have been possible to implement them as such. Furthermore, this could have been easily conveyed to the player by not letting their domes touch the ground but end in mid-air some 10-50 meters above ground.
If balancing out artillery fire was not their intended purpose my point is moot.
But taking the final state of shields in SupCom and saying that a different design would have been a mess and would have been difficult to wrap your head around because it differs from the one being used in the end is a no-comment - of course your point of view changes when you change your point of view.

See:

Bad UI feedback is just as bad, if not worse than a pile of rules.

So what happens? The shield blocks all the projectiles. It's a simple rule, with huge repercussions. Why? Because the shield blocks everything now. It can not be tweaked around any single variable, it can only be tweaked against the entire game. Is that a problem? Hell yes. A defense was supposed to do one simple thing, and it dragged the entire game into the crossfire. It made a MESS.

The umbrella shield TRIED to address this, by shaping the shield in a direction that only artillery can come from. Did it work? Noooooope. The shape of the shield wasn't the problem, it's more fundamental than that. The problem is that you tried to block artillery with a shield. It simply doesn't work.

Was it a mess? Yes. Did it have to be? No. I claim that shields could have been implemented in a way conforming to their supposed original design. You claim that the implementation was a mess. I don't see why you are so intent on disproving my statement by proving another, unrelated one.

The umbrella shield is another kettle of fish entirely, though. I don't know for sure but it seemed to be designed to explicitly block fire coming from above it - if the designers meant only artillery fire to come from that direction they should have taken every step to ensure this assumption holds. Because if someone finds a way to ensure that any fire directed at targets covered by the shield has to come from above - congratulations on making good use of the environment.

The umbrella shape did not target the weapons it was meant to target, and it could be used to block pretty much everything else. It was a failure of an idea because it didn't work as intended.

Is that idea "a bubble that only blocks specific projectiles"? If so see:

The closest idea to an anti-artillery weapon is to use a TMD style of defense. The TMD would excel against singular large hits, which is exactly what artillery is. Even if it did target everything and dip into other aspects of the game, the cost of a dox bullet is still FAR less than the cost of any artillery bullet. The TMD would still excel at the task it was meant to.

We seem to be going a little off-piste here guys. Stacked Air > Shields. If you want to talk about shields, which aren't even going to be in the game, please create a thread, or join one of the myriad threads on shields.

Sorry for getting sidetracked here, I intended the original comment to be on design decisions in general. I guess it is safe to say that it would be at least worth a try to balance bombers by exchanging their float-and-drop behaviour with something more interesting. Heck, this could even be used to make single-target bombers automatically turn tail after dropping a bomb to evade retaliating fire - sort of following an elliptic attack path. So many ways to differentiate units beyond unimaginative number juggling...

Once we have formations for air I suspect this flying is a block will be gone anyway.. and based on the live streams they are really close to having a 1st pass on that.. perhaps even next patch

i just hope we can set formations for air to allow for true carpet bombing of a base or all in a line for single target stuff.

the thought of 15-20 bombers in sets of 5 in a V formation carpet bombing a base (even my own) would be thing to watch!

I agree formations will solve part of the issue and will look awesome. However if we have several formations overlapping, the stacked damage problem unfortunately remains.

Sorry raven but this post was the one that finally convinced me that no stacking is the way to go.
The problem is in a base flack will destroy any stack that get near, but everywhere else on the map stacked air beats everything, and once you can stack attack power ultimately you cannot balance it ( just stack more).

The how to converge over a target is what creates the difference between advanced and basic: basic is smaller so more bombers can get closer to the target faster, making the single bombs better at anti building or single unit. ( what I belive there purpose is)
Advanced is far larger, making the formations bigger, meaning they can't all attack at once ( less effective vs single unit), this firmly puts t2 in the realm of anti units as they can carpet bomb sections of large formations, one at a time and still do damage.

And gunships still hold the glass cannon, in that they can circle around a unit to get maximum attack damage but a) have to travel around and b) suffer from low health.
And if a full circle of gunships isn't enough attack damage we have other problems...

To be clear I think ground should need support to advance against air, but at the current stage a large bot force with anti air mixed in gets carpet bombed in one or two passes by stacked air, with no significant losses to the bomber force. With no air stacking the air will still win, but due to the bombers coming in slightly spread out anti air will take out a few and might get more then one shot.

Just with regard to gunships and fighters: As they can fly above one another and still hit targets (as they fire forward) they may not be as affected by removal of stacking. However obviously this 'stacking' by flying above eachother would be limited by vertical range (imagining the normal flat circular range of land/naval, projected into a sphere for these Air units capable of moving and engaging targets in three dimensions). That would mean that you would still reach a point where only so many gunships could occupy so much space, even if they were flying above each other in a circle around a single target.

that kinda of stacking is an entirely different issue, having fighters and bombers at different highest I am ok with as it prevents micro of fighters and bombers, having fighters or gunships self stack is kind of self defeatist (then you have problems with stacking fighters vs bombers) and becomes ugly edge cases (more so when not needed)

Anyway a stream of fighters flying overhead into a battle looks far better then one stack...

It's all well and good to say "stacking is bad", but I'm not seeing anyone actually describe what the behaviour of aircraft would be in order to achieve this. But firstly, some clarification:

Firstly, aircraft moving are easy to unstack. Even Total Annihilation was able to able to make units un-stack when going to land, so doing it in the air shouldn't be a great leap. I have no problem with this. Secondly, unless you want all aircraft to move like gunships and stop in mid air, please don't ask for aircraft to not clip at all. It's just not feasible, isn't worth the development time and has very little impact on the game. Being 3D doesn't change the fact there's still limited airspace, and potentially 1000's of aircraft that are moving. The best realistic outcome would be units that try not to clip, but if they need to, they do. This removes any wierdness with large blobs moving opposite directions through each other, or during large dogfights. Thirdly, taking the previous two points eliminates most issues except for one: Bombing runs. This will be the focus of the rest of the post.

Since a bomber's job is to hit a target, each bomber that is ordered to attack must have a chance of hitting the target during their run, assuming the target doesn't move (the exception being the bomber was too close to realign to target when given the order). This means there can't be bombers off to the sides that can't possibly hit the target, which reduces the behaviours to just two broad methods if you don't want them to stack:

A. Bombers form a line and make a bombing run together
B. Bombers take turns at making a bombing run

These, however, are overly broad. When giving an order, you should know how it will be carried out. So, keeping in mind this needs to work with arbitrary bomber units (eg. a range of speeds & sizes, potentially mixed):

• How do bombers determine what order they will be in?
• For A, where is the line formed?
• For B, what do the bombers do while waiting?
• What happens when different selections are given separate orders to attack something? What happens when additional bombers are "Added" to the attack later?
• What happens when bombers auto-attack due to patrolling, rather than a manual attack command?
• What about bombers who are told to move within range of a target, then manually told to attack it? One would assume is should be able to immediately drop it's bombs, but this is easily exploitable to cause stacking.
• What about bombers that are told to attack ground on top of/next to the target, or if several bombers have separate orders to attack ground units that are right next to each other?

(I can think of more, but these will be enough for now)

So to show you why I've been arguing against trying to prevent bomber stacking when making the bombing run, try the following - keeping the above in mind, describe how bombers act in the following situations. Remember, the behaviour should work in any circumstance, and someone giving the order should be able to see in their mind how the bombers will carry out the task. I should also not be able to think up other situations than those below where the behaviour wouldn't work. Boxes indicate different selections, and the cross is the target. Lines just indicate targeting, not movement:





If we get past of all this, we can talk about the balance impact of trying to attack in a single file. If you don't think it's a problem, try it with land units against ground defenses, and see how far you get

Since changing altitude might not be the be-all-end-all to the problem how about flocking behaviour? Or those wondrous flow-fields that are supposedly working their magic in all affairs pathfinding. Let them do some work. Supcom2 managed to solve the problem for ground units and those can't even change altitude.
Attack orders could be solved in such a way that no concentration of units at the target occurs. In SupCom bombers dropped their yield at quite some range (close to the static anti-air range) and then continued to fly in that direction. Instead of happily strolling towards their death they could loop back as soon they drop their bomb - once again following some flocking behaviour so they might fan out and converge again.
Another thought: You are flying a bunch of barely maneuverable units into a big furball - if they can't avoid collision why should they be rewarded for that? Let there be collision damage and people might refrain from headlessly throwing their units together into a big deathball.