Expanding on Nukes

I generaly prefer to win and have fun while doing so ...
The ammount of units doesnt matter but how they are used
However i didnt say that destroying things should be a way to balance things (bad wording on my part?)
But what you as a player want to do is denying your enemy access to any ressources and production capabilities as much as possible ... you dont want to get nuked? Go scout find were the nukelauncher is and destroy it ... you dont want your enemy to built anything? Go destroy his energy and metal ... he will probably try to do the same ...
Defencses are there to deny your enemy gaining more territory ... but the best thing you do is denying him the tools to gain territory to begin with...

Preventive strategy only works if you have the power to do so.

And no matter how many times people insist that an enemy building a nuke would give the the economic advantage, it still persists that the player with the nuke has a much greater advantage then the player who could squeeze out 30 more tanks.

If the only way to say, hypothetically, destroy enemy tanks was to destroy his factory's and prevent any from being built, doesn't the player trying to build the all or nothing device only have to stall you to ensure victory?

And if you could muscle through and take out the nuke, then you already had the power to outright win by killing the commander.

So why wouldn't you have?

I am just saying that if the counter to something is to kill it before it finishes building, then the player who is building it isn't going to build it in some field for you to find and destroy. They will rush towards it, and go all or nothing in a last ditch to pull of the construction that once finished guaranties victory because now you cannot stop it without realistically also having done the same with the counter, if you built it in the right place.

Which leads to a cat and mouse spiral:

Nuke but no anti? Win
No nuke but anti? win
Nuke and anti? lose
No nuke and no anti? draw

Its just so bad on many levels to have a device that does that to a game.

Only got to page 3 so far but although it hasn't been discussed any further than in the OP I must say I kinda like the idea of the anti missile gun network, although I'd probably prefer it as a laser beam tower defence network. The more you have targeting a missile, the more effective they are at taking it down. Their range is about 2x that of an anti nuke.
At about 2.5k metal cost, a couple can keep up with 1x catapult fire, 6 or 7 could take down nukes. They would obviously have to prioritise targets.
They require power to fire.
They cannot fire at ground units or aircrafts.

The fact this counter is analogue could make for some pretty nerve racking missile approaches.
This should work fairly well against early nukes, but for nuke barrages you'd still want anti nuke silos for more instantaneous take downs.

Oh just a quick question:
I haven't yet thought of trying this our but does area attack work when you select a bunch of nuke launchers? Ideally for this use we'd like to be able to draw the area and not just rely on a circle of varying diameter.
Saving individual nuke launchers to hotkeys for IP nuke barrages is time consuming and breaks up all your previously assigned hotkeys.

risk/reward ... you cant always go for the save route as it may lead to stalemates

And then get consistently nuked every-time as a result of 'playing it safe'.

So firstly, I would like to say that I quite like the changes proposed in this thread. The key contention that many players have with the nuclear game at the moment is that it is dull, and potentially has a very low skill requirement, in terms of both strategic thought and micromanagement. The issue is that some people have drawn the conclusion from this that because of it's lack of interest, the nuclear game should be nerfed, to discourage it's use. This is somewhat silly.

Don't nerf the boring part of the game to make it less important. Make that part of the game interesting. This is the whole point of this thread. Ideally, the nuclear game should be just as interesting and dependant upon player cunning as any other part of the game. I want to play a game where one player can be using WW1 Maginot lines, minefields and overwhelming infantry (bot) pushes, another can be using WW2 combined arms with tanks and blitzkrieg air-support, yet another is fighting with insurgency-like stealth and ambushes, while the fourth player is using cold-war missile tactics. Most importantly all of these play styles, and many more, should be just as interesting to play and no more likely to prevail if deployed incorrectly. That is a varied and interesting strategy game. It's also a pretty tall order, and not one we shall see for some time, but I do thing it is within this game's grasp.

Now, onto the actual proposals themselves. Firstly, I like the idea of having a cheaper and more spammable mechanism of nuclear attack and defence, as that is essentially what the "tactical" nukes and point defences bring to this discussion. However, the idea that they should be non assist-able feels like unnecessary special casing. One of the fundamental mechanics of this genre is resource allocation, in that a player who allocates more resource to a particular avenue of attack will be more likely win it, whether it is more tanks, or more nukes. An unassistable structure just feels like an unnecessary bottleneck in resource allocation. I could understand this if defences were expensive and late game, as a less flexible deployment time for nukes would give time for opponents to build an expensive defence. However with cheaper defences, this is unnecessary. Furthermore, assuming there is a delay between the firing of successive missiles, the ability of a player to overwhelm another's defences is dictated more by how many missiles they can launch at once (i.e. how many silos) rather than how quickly they can replace spent missiles. It's true that a player assisting missile production can reach the point of being able to overwhelm their opponent quicker, but in doing so they have spent additional resources on gaining such capacity, meaning they have paid for it with weaknesses in other areas. For example, they might not have as many tanks, so when they see a tank army approaching, they have to spend some of their rushed missiles to eliminate it, meaning they have less missiles now, and parity is restored. This certainly highlights the need for mobile defences as well.

One issue I can see with this system is the potential confusion between the different types of defence. Let us suppose that I have an array of gun defences, and a single ABM silo. My opponent launches a swarm of tactical nukes, and a single strategic nuke, with the aim of overwhelming my lighter defences with cannon fodder while the larger nuke slips through. It would be in my interest to reserve my ABM silo for use only on the more serious threat of the strategic nuke, while risking the chance of tactical nuke penetration by letting the point defences handle them. How is it proposed that I can control my defence network in order to achieve this? This is not a criticism, merely an honest question.

Finally, I think that at some point we would still need to face the spectre of adding more units. Although the proposed changes go some way towards alleviating the present scenario, the nuke game will always be a fraction of the interest of the rest of the game, as long as it has a fraction of the units. The current number is one offensive missile, one factory, and one defensive structure. This proposal adds another missile, another defensive structure and another factory. Options are discussed for mobile missile launchers (bombers, ships, subs), but even that only alters the mobility of the missile launcher. That would help in terms of allowing players to attack from different angles, and potentially circumvent defences, but it doesn't alter the main effect of a missile getting through. It just makes it more likely if used correctly.

I would also like to see additional types of missile, either in the strategic or tactical capacity. An EMP weapon is an obvious possibility, as well as some sort of MIRV system, which has greater defence penetration power in exchange for less overall damage. There has also been a discussion of nukes as planet sculpting tools. The ability to alter the playing-field is an extremely powerful one, and doesn't make sense as a normal consequence of nukes for the immense capability it would bring. However I could potentially see an expensive missile which has a dedicated role as punching small craters, or blasting holes in mesas, to either block or allow travel. One could even go as far as missiles which have cameras used purely for intelligence gathering, or radar jamming chaff to deny Intel. I would also be a fan of a defence system which diverts and jams missile guidance, perhaps as some sort of orbital unit. I like this idea because the "missiles get through/don't get through" dichotomy is now mixed up with a third mechanic of "they hit the wrong place". I would personally find it entertaining, and it could be an interesting idea if it behaved in a semi-predictable manner.

Overall, I like the ideas presented, however, I think we will need to go further in order for this segment of the game to stand up well against the others. I suspect that the devs won't have time to do this kind of overhaul for a while (if at all), and we will have to be limited with smaller tweaks to the existing system. However, I'm optimistic in the long term for more development, or possibly modding hooks to enable this kind of system.

I as a rule always build anti-nukes as par for the course in any base, outpost, my reasoning here is that it is a defensive tactic, much like turrets and artillery. Oh and make sure they have a range-overlap, almost Venn diagram in appearance on the Range overlay, as that better safe than sorry tactic.

The main reason here is I used to play against TA and SupComm against a hardcore nuke spammer... so even though I don't use nukes often, if anything I use it to open up the way for a unit seige and as Arty support. So I'm safe from nuclear retaliation when the enemy realises it's doooomed and tries to nuke me.

Anti-nuke in my opinion is more important than nukes, and it's an investment for the end-game strategy.
Much like spreading to a second planet via teleport, unit cannon or Astraeus is an investment against the enemy throwing rocks at your base.

Target prioritisation, all defenses automatically and always turn towards the biggest threat (BM) until it is down then move on to smaller threats (SM). Some SM might have gotten through but you prevented the biggest devastation from happening.

That's fine for defences with the same characteristics, but what about when we have a long range defence (the ABM) and a short range defence (the gun based CIWS)? The first defence to see a target will probably be the ABM, meaning it will spend it's missiles before the gun CIWSs see the missiles. However, the case may well be that we want the ABM to not spend it's missiles unless there is a credible threat, such as having it ignore tactical nukes, and wait for a strategic nuke. However, if there is a very large number of tactical nukes, I may want to spend the ABMs in order to thin them down a bit for the guns to finish them off.

Knowing whether the threat is credible is not an easy task that can be simply solved by a target prioritisation list. This means that player input is required. Given that attacks will be swift, and this is not a micro oriented game, it is unreasonable to expect the player to manually target incoming missiles, or be forced to reactively control parts of their missile defence network. This means that there would need to be some way of giving orders to the missile network for when the shots actually start being fired. The result is that we either need more complicated unit stances available or some other UI technique to allow this control. Neither of these seems easy, and I don't know how they could be done without adding ludicrous amounts of complexity, or handing over a lot of control to some sort of AI heuristic to estimate the positional capacity and loading limits of a missile defence network.

Maybe simply turn the power off the ABM until you actually scout a BM launcher? Not the most elegant I guess. I'm not even sure turning your ABM off currently prevents it from shooting its anti-nuke.

Also since you would need several (cheaper) gun/laser based CIWS units, one would be tempted to set those up on the base perimeter and cover all/many angles while the ABMs would sit in the central areas. The idea I exposed earlier of a longer range laser based CIWS could probably make that work fairly well ?

Also maybe the ABM should only be able to take down BM while the gun/laser based CIWS would be able to shoot both types down. That'd take care of your problem.

That still doesn't solve the issue, even if the ABM can be turned off. I still need to turn it on in order to use it, and its unreasonable to expect someone to do that in the middle of a missile attack. As soon as it get's turned on, it can waste missiles on trivial targets that the guns could handle.

I'm assuming that anyway. The only way to not make it waste missiles on trivial targets is to not have areas which are covered by both forms of missile defence, which defeats the point of a synergy between the two. As I have said, I'm not sure that there is a way to have the ABM act sensibly without relieng on either a lot of micro, or a fairly complex heuristic.

I can think of a way to do it in terms of how one would approach the problem, but it ain't pretty. I want to see if anyone can think of a more elegant solution.

To me that misses out on a key point, if you set everything to prioritise against the major target, well that forgets to account for things like the "major" target being a decoy for an inventive group of "minor" targets to get in and smash you to pieces because you "won" the battle and took your eyes off the "harmless" group.

Or the alternate, wiping out the minor group, but sacrificing your defences, meaning the major enemy, maybe near death itself, still has enough health left to make sure your victory is well and truly prryhic in nature.

Instead of things like priority listing, a smatter of everything everywhere to give you coverage against most methods of attack, which is expensive, except if you're notwilling to spend a smeg-tonne of money to defend your empire, how are you expecting to break their empire if all they need to do is sneeze to beat you?

Those are some really good points. I really like the idea of diversifying nukes but I see how a priority system of some sort would be needed to make it work. I guess the simplest way would be to have missile anti-nukes only shoot strategic nukes but that's far from prefect.

I think the most important difference between the tactical nuke and the strategic nuke is the tremendous difference in range and interplanetary capability of strategic nukes. If they are more destructive then a target priority system is fairly straightforward to implement which would make missile antinuke preferentially attack larger missiles.

Assuming there is a difference in yield which would make the player care about the order in which they are intercepted, the most elegant system would be to have ICBM's and tactical missiles be visibly different in size, and have antinuke prefer to shoot at ICBM's within range. However ICBM's could be harder to intercept because of their orbital trajectory.

The antinuke system absolutely must be completely automated. Expecting players to micromanage their nuke defense is an extremely bad idea. It would be the closest thing you would ever get to a hardcore quicktime event in an RTS game- "Press X to not die" only in this case it is "target fire the nuke to not die." The antinuke needs to do the best possible job given the available antinuke facilities without any supervision. However the player has to decide how much antinuke to put where ahead of time.

I think that having to micromanage your nuke defense wouldn't be such a bad idea if everything else wouldn't take so much micro already and if it's only in the case of the long-range anti-nuke missile. I mean about how much range are we talking here? You maybe imagined it completely different than me but I thought that those missiles had an interplanetary range too, just like the interplanetary-nukes themselves.

How do you automate them without having them intercept nukes that are directed at your opponent?

I agree that the missile defence network should not require active intervention to make it function efficiently. The only decision a player should have to make is where to build their defences. I also agree that we can assume that a strategic nuke getting through is almost always going to be a more serious threat then a tactical nuke, thus missile defences should prioritise them over tactical nukes. However, I'm not sure I'm explaining myself well enough as to how complex an issue efficient targeting is. It is not trivial task to figure out how to sensibly spend a player's ABMs efficiency. I think I will need to show some concrete examples:



I apologise for the image quality as imgur is being weird. The diagram above represents a simple scenario, where a defending player is under attack from a single tactical missile. The defending player has an integrated defence network, with both CIWS guns, and an ABM silo at the back. The engagement range of these are given in green and yellow respectively. For the sake of simplicity, let us assume that 1 missile flying over 1 CIWS for the distance of one engagement circle diameter will almost certainly be downed. In all these scenarios, the ABM silo only has one ABM in it's magazine.

As the tactical missile enters the range of the ABM silo, if the ABM is programmed to fire at the highest priority target it can see, then it will fire it's missile at the single attacker. I think that most people here could see that this would be sub optimal, as the CIWS would easily dispatch the enemy missile without having to squander an ABM. In this scenario, if we were manually controlling the ABM, the best thing to do would be to tell it not to fire except in the unlikely event that the incoming missile somehow made it through the CIWS line. Already, we can see this isn't an easy thing to communicate, as the ABM needs to know where the CIWS turrets are in order to efficiently engage enemy missiles. However it gets even more complex when we add more missiles in scenario 2:



Here, the attacker has launched multiple tactical nukes, and a single strategic warhead. given the spread of missiles, and the number of CIWS (6 vs 5), it's likely that one missile will get through, possibly even the strategic one. However, even if the ABM silo prioritises the bigger warheads, it will still waste its ABM. The first missile to enter its engagement range is a tactical warhead, at which point it becomes the highest priority target in range. In order to work effectively, the ABM needs to somehow know to wait for a bit, before trying to take out the strategic missile. This is a scenario where the difference between a simple prioritisation system and a prescient silo could make the difference between a strategic warhead going off or nothing getting through. One potential option is to give the ABM a greater targeting range than it's firing range, so that the bigger threat will hopefully become targeted before the ABM silo has spent its missile. However, all this does is delay the problem. As long as the attacking player launches the strategic weapon slightly later, the ABM will still target the wrong thing. We could also suggest that the ABM silo just ignores non-strategic warheads, but even that is open to abuse, as seen in scenario 3:



There are the same number of missiles before, only this time there is no strategic warhead. Also, the attacking player has tried to be cunning, and has directed more of his missiles at one location, hoping to locally overwhelm defences. If we had an intelligent human in charge of the ABM silo, then they would look at this scenario, and come to the estimation that the missiles will get through unless the ABM silo intervenes (and maybe not even then). If the ABM silo launches as soon as possible at the correct missile, it can lighten the load on the CIWS cannons, and hopefully prevent a breakthrough. Furthermore, the missile taken out is crucial. Destroying one of the two missiles coming down the sides is pointless, as these will likely be stopped by the less heavily loaded CIWS cannons anyway. The ABM silo needs to realise that the best chance for stopping a breach is to target one of the centre missiles as soon as they are in range.

I hope this demonstrates that target designation is not a trivial matter. In each of the above scenarios, the ABM must behave differently in order to maximise it's usage. To make the ABM behave well, there actually needs to be some clever heuristic as to determine whether or not the ABM should fire, and which missile the ABM should go for. I can think of one way to do it, but it is complex and inelegant. Essentially, each missile has a "threat" variable assigned to it on launch, dependant upon the type of missile. For example a tactical nuke may be 100 and a strategic nuke may be 500. The calculation is preformed by the ABM silo. The ABM silo "knows" the target of each missile. It then performs the calculation whereby the distance yet to be travelled through a CIWS engagement range is subtracted from the threat value, while an inverse function of the distance to the nearest n missiles is added. This would produce a dynamic variable for each missile in flight, and if the threat value is greater than the unaltered value for the least threatening missile, then the ABM would fire. The end result is that the ABM would prioritise large missiles, in dense swarms, which aren't going to fly through a CIWS engagement zone. This is not a perfect algorithm, and I can think of one or two edge cases where it could potentially make a bad decision. It also encapsulates enough parameters that it would need to be tuned to provide a decent coverage. I have no idea how hard this would be to implement, but I do not expect this or any other solution to be easy. In fact, I would be really curious to see how a dev would hypothetically handle this problem, as it's a pretty interesting challenge.

Of course they would, (like planet busters) but they would also have a **** ton of nukes. Just because you can build more sophisticated weapons of mass destruction doesn't mean you wouldn't still employ and use vast amounts of bog standard nuclear fission bombs.

In my opinion nukes should be cheaper, to reflect realism. Before someone jumps in shouts GAMEPLAY > REALISM, I might as well add that the game can be balanced so that cheap nukes won't be detrimental to gameplay, and I'm not just talking about cheaper anti-nukes. Ground/Sea/Air based armies would have to have a way of being less vulnerable to a defensive nuke. A mobile anti-nuke is the first thing that springs to mind.

Neat idea! That algorithm would work quite well if you added a way to calculate the load on CIWS and if the ABM should assist due to overload. Maybe reduce the amount you subtract for a CIWS based on how many missiles it is handling.

That should actually be emergent from the algorithm. If missiles are penetrating deeply through the CIWS area, then the threat variable in each one will rise until the point where the ABM steps in. If there are multiple CIWSs covering the same missile, then it is travelling through multiple engagement areas, and thus it's threat index will be lowered.

The problem is not the general behaviour of this system (aside from a couple of edge cases). The problem is that it involves assigning dynamic variables to individual units, each one with a computational overhead. I've seen RTS engines where this is a terrible idea, with a large computational overhead, and I've absolutely no idea whether this would be workable in PA.