Stealth Carriers Over-Powered?

For what's not in 'Top Priority Game Design'. Post your ideas, visions, suggestions for the game, rules, modifications, etc.

Moderator: Oberlus

Post Reply
Message
Author
Telos
Space Squid
Posts: 50
Joined: Thu May 09, 2013 4:46 am

Stealth Carriers Over-Powered?

#1 Post by Telos »

I've fiddled around a bit with stealth-carriers and they definitely seem overpowered. A simple early design is a Symbiotic hull with cloak, fighter hanger, and two launch bays. Later you can get automatic upgrades from fighter weapons techs, and improved carriers with better cloaking devices, or moving up to larger Protoplasmic, Endosymbiotic or Sentient hulls.

Currently, so long as your stealth level is higher than enemy detection levels, stealth carriers have zero risk in combat -- they can just launch their fighters/bombers and do some free damage, and so long as your stealth carrier begins military occupation of a system first, you count as "blockading" it, so no enemy ships can ever pass through. There are a few drawbacks that are pretty easy to play around. (1) You don't interact at all with planets, so that means you won't wear down planetary defenses to open them for invasion, but it also means you can camp atop a planet without losing fighters, thereby blockading visible enemy ships from ever leaving that planet. (2) You have to go back to your own supply lines to replace lost fighters, meaning that you often need to hover at the edge of your supply, jumping in to launch a swarm of fighters, then jumping back (often with all but one ship to maintain the blockade) to replace fighters, lather rinse repeat... (3) It's hard to fit extra fuel alongside your cloak and hangers, so you're usually restricted to doing this near your own supply areas. Fortunately, parking a stealth carrier in any system, even one with planets, shuts down enemy supply lines, enabling your own supply lines to extend further into enemy territory, so this restriction isn't as great as it seems. With a fleet of these cheap stealth carriers, you can easily lock down enemy movement, disrupt enemy supply lines, and slowly wear down all enemy ships, making enemy ships effectively not even be a part of the game.

Currently the game offers a few potential responses, though unfortunately the AI seems not to use any of them. (1) Often best is just to research the next sensor tech, though that gets expensive, and apparently the AI never thinks to do it. (2) Lighthouses reduce stealth of all ships that start the turn in a system, which can prevent stealth ships from camping in a system, but does little to prevent them from dipping in and out of the system, which is often what they need to do to replace lost fighters anyway. At best, this usually just makes them dip back out on a more convoluted path to avoid getting caught somewhere nearby at a system with their cloaks down. Also, the AI apparently never builds them. (3) Distortion modulators are a ship module with a similar local-decloaking effect to the lighthouse. I haven't tested them, but my understanding of FOCS scripting is again that this will only affect ships that start the turn in the same system, so again it won't do much versus a dip-in-and-out approach, though it probably would open the door to breaking blockades (which require a ship to remain in the system to maintain the blockade, which will be hard to do if that ship gets destealthed). This is also gated behind the very expensive Theory of Everything, making it be too late as well as too little.

I would say that one design goal should be to make ships more interactive and counterable than stealth carriers currently are. Here are a few potential solutions.

(1) The simplest solution would be to view launching fighters as a form of weapons fire that therefore automatically makes a ship be detectable in later rounds. Unfortunately, I think this would require a change to game code, though probably a very simple one, and one that shouldn't require any changes to the AI decision-making. A lighter nerf to stealth carriers would be to make launching fighters cause a decrease in stealth, which could enable a very stealthy carrier to still be invulnerable against low-tech scanners. Another lighter nerf would be to make carriers visible any round after they launch fighters, but then go invisible again if they don't launch fighters the next round.

(2) Another simple/crude solution would be to make launch bays have a stealth-reducing effect on their ships. This would be easily scriptable and would have the effect of making stealth carriers much harder to implement and abuse, especially with multiple bays. Unfortunately this has the effect of removing from the game the possibility of having (most multi-bay) carriers be stealthy *at*all*, thereby reducing the strategic depth of the game, which might be good from the perspective of keeping things simple for the AI, but is bad from the perspective of wanting a strategically rich game with a variety of moves and counter-moves.

(3) Distortion modulators should be available earlier, perhaps gated behind Lighthouses and Force-Field Harmonics, rather than behind Theory of Everything. This is an easily scriptable change, but isn't enough by itself, since distortion modulators only reveal ships after they've already arrived and done their damage, and (unless there are enough survivors to blockade them) will be free to duck out.

(4) Lighthouses and distortion modulators could increase local detection strength rather than decreasing local stealth, enabling them to detect ships that just dipped into a system this turn. Unfortunately, the game currently treats detection strength as a single empire-wide value, so this sort of change would require a change to game code, not just scripting. This solution would probably also require changing the AI to place higher priority on building lighthouses and distortion modulators.

(5) Lighthouses and distortion modulators could decrease stealth over a significant distance, rather than just in-system. This should be easily scriptable, but might seem over-powered. To capture ships that haven't yet dipped in, this would need to reach out far enough to see them on approach. If the range of decloaking field matches the speed of enemy ships they'll have no chance of getting in (though if the modulator is moving, then the combined speed of modulator + cloaked ship could allow it to get in undetected). If the decloaking field has shorter range, it could still decloak ships that approach by certain routes, though not ones that are careful to approach in a way that makes their last step in be longer than the range of the lighthouse/modulator, which could be an interesting layer of strategy, or perhaps just a royal pain in the backside.

(6) Planets should do more against fighters. At the very least, they should shoot their big guns at fighters, if no other target is available. Better yet, planetary defenses should include some flak weapons to help shoot fighters out of the sky and prevent perpetual free blockades by stealthed carriers.

(7) The AI should be triggered by stealth attacks to make appropriate responses, like researching sensor techs. I'm not sure how much ability there is to make AI research priorities be contingent upon triggers, but this response is one that is clearly needed in any sort of rudimentary artificial "intelligence".

(8) The game's "blockade" rules should be loosened to give fleets an opportunity to break a blockade without detecting every last stealth carrier. A simple version of this solution would make blockades last only 1 turn, whereas fancier versions might take fleet size or power into account in determining who, if anyone, gets to count as blockading the other(s) in drawn-out conflicts. A complementary solution would require that a carrier actually have surviving fighters in its hangers to maintain a blockade (something that's not even currently required), making blockades a bit harder to impose, and offering an option to break them by flakking away the fighters.

[Updated to clarify hull options, note role of blockades, add a couple more options, and improve discussion of others.]
Last edited by Telos on Wed Jan 02, 2019 8:24 pm, edited 2 times in total.

User avatar
EricF
Space Dragon
Posts: 357
Joined: Fri Jul 29, 2016 10:12 am

Re: Stealth Carriers Over-Powered?

#2 Post by EricF »

Like Ground Troops and Weapons, give all planets Fighters. Maybe one flotilla of each
Interceptor, Fighter and Bomber. More for more developed planets.
(Once the tech is researched of course)
I personally like this idea, because I still think planetary defenses even maxed out
are a bit weak. They present no real threat to a properly prepared fleet.

User avatar
Vezzra
Release Manager, Design
Posts: 6095
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: Stealth Carriers Over-Powered?

#3 Post by Vezzra »

EricF wrote: Mon Dec 31, 2018 4:30 amI still think planetary defenses even maxed out
are a bit weak. They present no real threat to a properly prepared fleet.
I agree. The main problem IMO is that planets get only one (although massively powerful) shot per combat turn. Unless we fix that, we can increase the planetary defense meter all we want, a properly composed fleet will be able to take down planets easily.

A single meter for planetary defenses isn't going to cut it in the long run. We need to "flesh out" planetary defense mechanics, giving planets fighters (or something similar) is definitely one of the things that could/should be added.

Telos
Space Squid
Posts: 50
Joined: Thu May 09, 2013 4:46 am

Re: Stealth Carriers Over-Powered?

#4 Post by Telos »

So... yet another way in which stealth-carriers are over-powered is that you can park one at a choke-point in deep space and this counts as a "blockade" preventing visible enemy ships from coming through. It's not just that the enemies have to stop and fight off an ambush once as they're passing, which would be understandable. Then they get the little blockaded ring around them with the info that the only way they can leave is to go back the way they came. So my little 150PP carrier, now empty of its 4 bombers, is single-handedly stopping a fleet of 25 much more expensive ships from advancing into my territory, with absolutely no risk to itself!

User avatar
EricF
Space Dragon
Posts: 357
Joined: Fri Jul 29, 2016 10:12 am

Re: Stealth Carriers Over-Powered?

#5 Post by EricF »

Telos wrote: Wed Jan 02, 2019 2:49 am So... yet another way in which stealth-carriers are over-powered is that you can park one at a choke-point in deep space and this counts as a "blockade" preventing visible enemy ships from coming through. It's not just that the enemies have to stop and fight off an ambush once as they're passing, which would be understandable. Then they get the little blockaded ring around them with the info that the only way they can leave is to go back the way they came. So my little 150PP carrier, now empty of its 4 bombers, is single-handedly stopping a fleet of 25 much more expensive ships from advancing into my territory, with absolutely no risk to itself!
Emphasis mine

This problem exposes a flaw in both Supply and Blockade.
But maybe it is really a problem with Fighters.
A ship with a Zero combat value should not be able to block Supply and should not be able to Blockade.
I can't remember offhand, but I don't think a Scout with no weapons can Block Supply or Blockade.

Telos
Space Squid
Posts: 50
Joined: Thu May 09, 2013 4:46 am

Re: Stealth Carriers Over-Powered?

#6 Post by Telos »

Even if the carrier had kept a fighter, it's still pretty ridiculous that a single puny ship (too chicken to reveal itself, so really just a single fighter) could blockade a whole fleet. I'm not sure how the game decides who gets to count as blockading whom. As far as I can tell, it's whoever parks there first gets to count as blockading so long as they keep ship parked there with the aggressive stance each turn (not necessarily the same ship, and I'm not sure whether it even needs to have weapons, but if so, then apparently launch bays count as weapons, regardless of fighter count). I guess you can sort of think of it as being like a gatekeeper who has locked all the gates, except the one the enemy opened themselves on the way in, and the only way the enemy can open the outgoing gates is by clearing the defenders out of the way. Except it's pretty silly to think that one lurking carrier, perhaps with a few 1-hitpoint fighters, should be enough to make the difference in preventing a whole fleet from opening the gate it wants to go through.

So probably the game needs some other mechanism for deciding whether a blockade continues, e.g., one that breaks a blockade if the other side brings an overwhelming advantage of numbers, or perhaps of estimated ship power.

Of course, this wouldn't be nearly so much of a problem if the carrier became visible when it launched fighters, so it couldn't just lurk there risk-free turn after turn. There might still be other silly shenanigans, like rotating through 1000-hitpoint mature krakens to keep a gate closed without actually doing much, but this wouldn't be quite so vexing as being able to permanently lock everything up with a completely invulnerable ship.

User avatar
Oberlus
Cosmic Dragon
Posts: 5714
Joined: Mon Apr 10, 2017 4:25 pm

Re: Stealth Carriers Over-Powered?

#7 Post by Oberlus »

I would much preffer to fix the fleet/supply blockade system on its own rather than adjust stealth carriers with that in mind. See for example this suggestion.

Regarding Stealth Carriers, blockades aside, I don't think they are that much overpowered.
Good research species can rush detection techs to be always on par with the detection techs of their stealth-focused enemies and that is enough to prevent any ship to stay stealth (because any hull good for stealth carriers requieres to have at least one more detection tech than the potential victim, since detection tiern N is a prerequisite for stealth ship parts tier N).
Good industry species can stack up ships to ignore the enemy stealth carriers: yes, you get some losses on round 2 from the fighters, but you also have flaks and fighters so that on round 3 all enemy fighters are gone and you just ignore the empty carriers and focus on what really matters, invading colonies.
Plus stealth carriers can't invade any planet, have no anti-planet weaponry.

Playing against underpowered AIs that will neglect detection (because they are already dragging behind on other, more important tech branches), yes, stealth carriers are overwhelming. But playing against human players or against AIs that had a good start, they are not that much of a definitive weapon.

Telos
Space Squid
Posts: 50
Joined: Thu May 09, 2013 4:46 am

Re: Stealth Carriers Over-Powered?

#8 Post by Telos »

I agree that improved blockade rules would make stealth carriers less over-powered, though they're still not very interactive and still provide an extremely cheap way to inflict heavy losses without much risk, so I'd still count them as "over-powered". Relative to the current not-very-intelligent AI, they're extremely over-powered.
Oberlus wrote: Wed Jan 02, 2019 9:26 am Good research species can rush detection techs to be always on par with the detection techs of their stealth-focused enemies and that is enough to prevent any ship to stay stealth (because any hull good for stealth carriers requieres to have at least one more detection tech than the potential victim, since detection tier N is a prerequisite for stealth ship parts tier N).
No, you're off by at least a tier. The living organic hulls (symbiotic, protoplasmic, endosymbiotic, bio-adaptive, and sentient) each start with at least one free tier (15+) of innate stealth. As long as you install the cloak unlocked at sensor tier N, everyone else won't be able to see them until they research the much more expensive tier N+1. The AI rarely researches Neutron Scanners (400RP), so that means you can take most of the map with the first cloak unlocked at Radar (100RP). E.g., in a recent game with 12 AI's (where I handicapped myself by not doing anything for first 20 turns), at turn 175 only 2 AI's had managed to even research Neutron Scanners (400RP), and I've never seen AI research Sensors (1K RP), so my carriers with Neutron-level cloaking (200RP) are completely invulnerable to everything I've ever see the AI produce. Things get even worse when you research Bio-Adaptive and then Sentient hulls, who both get 2 (!) free tiers of innate stealth (35+), and the latter of which also provide a stealth boost to their fleet-mates. With mere neutron-level cloaks, the only way these can be detected is for the enemy not just to match your Neutron Scanners (400RP), but also to raise you by both Sensors (1K RP) and Omni-Scanners (2K RP), which costs more total RP than Sentient hulls did (and much much more than Bio-adaptive), and doesn't come with useful population techs along the way.
Oberlus wrote: Wed Jan 02, 2019 9:26 am underpowered AIs will neglect detection (because they are already dragging behind on other, more important tech branches)
Out of curiousity, what are the "more important tech branches" that you think AI's should focus upon rather than detection tech? When an AI is on the ropes from getting hit repeatedly by stealth attacks, it seems to me that detection tech has pretty much got to be their *most*important* priority!

User avatar
Oberlus
Cosmic Dragon
Posts: 5714
Joined: Mon Apr 10, 2017 4:25 pm

Re: Stealth Carriers Over-Powered?

#9 Post by Oberlus »

Telos wrote: Wed Jan 02, 2019 9:30 pmNo, you're off by at least a tier. [...]
You are so right.
Digging a bit more, with bio-adaptive and small camouflage asteroid (both stealth 35), you can do without researching any stealth tech (plus not needed, since this hulls only have one inner slot, for the fighters) and enemy will have to get the two first detection techs to detect.
E.g., in a recent game with 12 AI's (where I handicapped myself by not doing anything for first 20 turns), at turn 175 only 2 AI's had managed to even research Neutron Scanners (400RP), and I've never seen AI research Sensors (1K RP)
I've seen AIs getting omni-scanner (and not because they had depleted the whole tech tree). But this only happens on games in which at least one AI gets clearly ahead of the others (otherwise they have an ugly tendency to get stuck in deadlocks).
To increase my chances of fighting against good AIs, I resign from any game where no AI is rocketing in number of colonies and at least one of research or production after turn 100.

If in a game I'm pumping out bio-adaptive carriers before any enemy has sensors, I'm out :D

Regarding Sentient hulls and bio-adaptive, the total number of minimum turns is so high that good research empires can get omni-scanner soon enough to counter if you are not overwhelming them in production.
Telos wrote: Wed Jan 02, 2019 9:30 pm
Oberlus wrote: Wed Jan 02, 2019 9:26 am underpowered AIs will neglect detection (because they are already dragging behind on other, more important tech branches)
Out of curiousity, what are the "more important tech branches" that you think AI's should focus upon rather than detection tech? When an AI is on the ropes from getting hit repeatedly by stealth attacks, it seems to me that detection tech has pretty much got to be their *most*important* priority!
I completely agree with you. I used the wrong words. I meant "more prioritised tech branches". I certainly think AIs tend to use suboptimal tech orders.

User avatar
Vezzra
Release Manager, Design
Posts: 6095
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: Stealth Carriers Over-Powered?

#10 Post by Vezzra »

Telos wrote: Wed Jan 02, 2019 2:49 amSo my little 150PP carrier, now empty of its 4 bombers, is single-handedly stopping a fleet of 25 much more expensive ships from advancing into my territory, with absolutely no risk to itself!
This is a design flaw/bug, plain and simple. Only armed ships can enforce a blockade, and a carrier which has its fighter wing depleted and no other armament should not be consodered "armed" by the engine.

Opened an issue on github: https://github.com/freeorion/freeorion/issues/2343

Ophiuchus
Programmer
Posts: 3433
Joined: Tue Sep 30, 2014 10:01 am
Location: Wall IV

Re: Stealth Carriers Over-Powered?

#11 Post by Ophiuchus »

Vezzra wrote: Sat Jan 05, 2019 4:50 pm
Telos wrote: Wed Jan 02, 2019 2:49 amSo my little 150PP carrier, now empty of its 4 bombers, is single-handedly stopping a fleet of 25 much more expensive ships from advancing into my territory, with absolutely no risk to itself!
This is a design flaw/bug, plain and simple. Only armed ships can enforce a blockade, and a carrier which has its fighter wing depleted and no other armament should not be consodered "armed" by the engine.

Opened an issue on github: https://github.com/freeorion/freeorion/issues/2343
I found that this topic is not a simple bug but needs a design decision/discussion; the Armed condition does not seem suitable for the blockade distinction:
Ophiuchus wrote:i .. tested using stealth bombers against some monsters. At the end of the combat i had no fighters left, so with the new implementation it was Not Armed. But actually it gets resupplied and hits hard.

I think the implementation (which prevents empty carriers from blocking enemies)solves the blockade problem.
But i think if a ship is able to do damage it should match the Armed condition (that includes empty carriers resupplied with fighters).
Vezzra wrote:Hm, an empty carrier which got resupplied with fighters isn't empty anymore... I guess we need to address the case when a carrier lost all its fighter in a battle, but is within supply, so will get its fighters reinforced immediately.

Which means we need to decide if a carrier should be considered "armed" in that case or not. Probably yes I guess?
Ophiuchus wrote:Basically we need to differentiate between different types of armed in different cases; something like Military (contains offensive parts like weapons, launchers, troops), Weaponized (contains weapons or launchers), and Blocking (was already at here last turn, has still some weapons operable/fighters left).

Querying for Military and Weaponized could be composed of DesignHasPartClass condition, so such are not really necessary IMHO. Querying for Blocking is probably difficult in FOCS so that would make some sense to provide, but i dont know if there is a use for this.
What i plan and would like your need input for:
  • ensure empty carriers do not block enemies (even if they get resupplied) directly in the code which checks for blockade
  • ensure also resupplied carriers are considered dangerous
  • the Armed condition seems to rely more on the FOCS interpretation than on the cpp combat mechanics - so refactor the Armed condition uses to a Weaponized macro, check the uses if something else is intended
  • if there is a place in scripting which checks for blockades introduce a Blocking (or similar) macro; i am not convinced this should be defined in FOCS, but AFAICS no harm comes from this
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Look, ma... four combat bouts!

User avatar
Vezzra
Release Manager, Design
Posts: 6095
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: Stealth Carriers Over-Powered?

#12 Post by Vezzra »

Ophiuchus wrote: Wed Jan 09, 2019 11:28 amI found that this topic is not a simple bug but needs a design decision/discussion
Correct (which is why I've been talking about a design flaw/bug, not "just" some buggy implementation).
the Armed condition does not seem suitable for the blockade distinction
Without having thought too much about how to solve the problem (and therefore if we can get the "Armed" condition redesigned in a way that it might suffice for all cases), I think that your assessment is correct. We are probably going to need more detailed classifications of military ships, something along the lines what you're suggesting.
Basically we need to differentiate between different types of armed in different cases; something like Military (contains offensive parts like weapons, launchers, troops), Weaponized (contains weapons or launchers), and Blocking (was already at here last turn, has still some weapons operable/fighters left).
Introducing a multi-level classification of ships sounds like good/reasonable approach. The top level classification should probably be civilian/military, which further distinction on lower levels depending on the capabilites of the ship. We already have that in a rudimentary way with the "Armed" and "Has Fighters" conditions, but apparently that's not sufficient anymore.

The "Blocking" condition you mention however is something that tells what a ship is/has been doing. While we might require to track that, it's still something different than a classification of a ship based on its abilities, so shouldn't get mixed up with that. Just to keep design and implementation clean.
What i plan and would like your need input for
Well, I'll try to give some feedback, but keep in mind that I'm not familiar with the actual implementation, so I can't say much about that. I can only provide some input on the design level.
ensure empty carriers do not block enemies (even if they get resupplied) directly in the code which checks for blockade
So, if a carriers fighter wing gets depleted in a combat, but the carrier survives, is within supply, will get its fighter wing reinforced and can resume combat the next turn, it still shouldn't be able to keep up a blockade, do I understand correctly? While I have a slight preference for the other option (consider a carrier in that case as being able to keep up/continue the blockade), because it seems a bit easier and simpler to me, we certainly can take that road. The fluff explanation is easy enough: the arrival of the new fighters takes some time, during which the carrier can't project any offensive power, allowing present hostile forces to break the blockade.
ensure also resupplied carriers are considered dangerous
This I do not understand, what purpose should that serve? That sounds like an assessment the AI needs to make of course, but not something we need to keep track of by means of ship classification or conditions like "Armed".

As far as the question if the classification of ships/determination of conditions like "Armed" should happen in FOCS or the backend C++ code I'm probably not the first person to ask. Personally I think such basic/essential classifications/conditions should be determined by the engine and not rely on FOCS, but here I'd like to hear the input of the other devs - Geoff, Dilvish?

And we probably should move that discussion into a dedicated thread...

Ophiuchus
Programmer
Posts: 3433
Joined: Tue Sep 30, 2014 10:01 am
Location: Wall IV

Re: Stealth Carriers Over-Powered?

#13 Post by Ophiuchus »

Vezzra wrote: Sat Jan 12, 2019 7:48 pmAnd we probably should move that discussion into a dedicated thread...
Armed, Dangerous, Military, Weaponized
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Look, ma... four combat bouts!

Post Reply