Space battles are broken and it ruins the game

[Agathors]

Stellaris should have hoi3/4's naval system (more appropriate + gain pratical experience in developing naval mechanic that can then be implemented in other games, mainly hoi's serie.) In addition they could implement a hybrid of hoi4's land battle resolution with Stellaris's excavation mechanic to simulate naval battles. Add to that victoria 2's economic, CK's politic and Stellaris's excavation based tech system influenced by hoi2-3 and victoria 2 for a innovative experience.

 

[Pancakelord]

Stellaris should have hoi3/4's naval system

You mean what they added in MtG?

MTG was awful, its an extremely opaque implementation and it also fundamentally wouldn't work (even the land battle idea), as Hoi4 battles are 100% simulated, not taking location or turn-rates into account. Which stellaris has as it uses pseudo simulated battles (missiles and PD are fully simulated entities, whilst energy/lasers are raycast/hitscan).

If you add this you might as well just remove solar systems from the game entirely, as there wont be any point rendering them anymore.

 

[Agathors]

You mean what they added in MtG?

MTG was awful, its an extremely opaque implementation and it also fundamentally wouldn't work (even the land battle idea), as Hoi4 battles are 100% simulated, not taking location or turn-rates into account. Which stellaris has as it uses pseudo simulated battles (missiles and PD are fully simulated entities, whilst energy/lasers are raycast/hitscan).

If you add this you might as well just remove solar systems from the game entirely, as there wont be any point rendering them anymore.

1. I never said I want only a hoi4 MtG implementation but to use Paradox's best renowned knowledgeable concepts, which include HOI4 ( the HOI 3 of paradox 2.0 like the OOB) has it central tenet for it's shared design philosophy with Stelaris.
2. This will indeed change it's DNA like Megacorp have done (what should have been Apocalypse instead of a band-aid). This may be impossible to implement fully with the current state of developement but should definitly be considered for a hypotetic sequel for a 4X (should be a 5X like MOO3 was destined to be) GSG. Like the tiles system, the solar system should not be rendered unless Paradox want a 4x RTS game akin to SoaSE.

 

[-Marauder-]

Doomstacks are a problem because fleets are so expensive. At 3 alloys per metallurgist, a 100 alloy corvette is worth 34 pop months of production. Similarly, an 80 Corvette fleet is worth two centuries of production by a single pop. As long as ships are expensive, I predict that doomstacks would be the dominant strategy because players want to protect their investment and losing ships is incredibly expensive to the loser.

Make navies cheaper or scale alloy production up and I reckon it would make sense to split up your fleets.

Even with a Megashipyard fleets take fairly long to replace. And as you mentioned, they get very expensive later on. Which means you can't just "sacrifice some" to stall or distract. And as tactical decisions are almost non-existant the only way to assure you're going to win is by bringing the bigger stick to the fight. Losses hurt, a lot. Losing an entire fleet pretty much means the war is over.

Rather than doomstacks, the problem is how valuable ships and fleets really are.

 

[-Marauder-]

Even with a Megashipyard fleets take fairly long to replace. And as you mentioned, they get very expensive later on. Which means you can't just "sacrifice some" to stall or distract. And as tactical decisions are almost non-existant the only way to assure you're going to win is by bringing the bigger stick to the fight. Losses hurt, a lot. Losing an entire fleet pretty much means the war is over.

Rather than doomstacks, the problem is how valuable ships and fleets really are.

I agree with this really. But it should work both ways: if your entire fleet is in one system, then where it is not is guarding your home planets. Therefore, it should be possible, even easy, for someone to go and nuke those planets into barren wastelands. I'm pretty sure that if someone invents FTL tomorrow and we find a fanatical purifier is lurking a few systems away, we'll be hiding lots of nukes in deep space just in case they decide to do something stupid..

Also, since space is vast, running away from a fight you don't want to have should be straightforward. Ok, there is that mechanic with the retreat button, but it seems to take months to power up those escape drives, by which time your fleet is going to take a severe pounding - particularly since they always rush at the enemy, no matter how unlikely the odds...

What you are seeing in-game are the massed mobile fleet assets. Them rolling out to wreck havoc doesn't mean nobody is left at home. From Border stations to citadels, all the way to defense armies and military infrastructure and the personnel to man them still exists. If you want to introduce Relativistic Kill Vehicles into the game. Sure thing. We can do that.

Problem just is. There's almost no way whatsoever to defend against those. They're cheap, would render pretty much everything currently in-game moot, and would end up with almost the entire universe being turned into a lifeless husk within a few decades of the game start.

 

[CrazyJ]

honestly the doomstacking problem seems unfixable without adding utterly unrealistic penalties to the game. i can only think of a few ways that could work and they all involve debuffing fleets based on how many ships there are in a system and that is just plain unrealistic.

 

[Zoolimar]

If you want to introduce Relativistic Kill Vehicles into the game. Sure thing. We can do that.

Problem just is. There's almost no way whatsoever to defend against those. They're cheap, would render pretty much everything currently in-game moot, and would end up with almost the entire universe being turned into a lifeless husk within a few decades of the game start.

Well we just need to account for the fact that RKVs would take 40-100 years to go through 1 jump. 3-4 hyperlane jumps and you are looking at target hit 300-500 years since you launched the RKV. That's assuming ships with FTL can't ever intercept them. FTL unless it is incredibly limited changes the playing field completely. Things that move slower than light don't matter that much.

 

[sillyrobot]

Well we just need to account for the fact that RKVs would take 40-100 years to go through 1 jump. 3-4 hyperlane jumps and you are looking at target hit 300-500 years since you launched the RKV. That's assuming ships with FTL can't ever intercept them. FTL unless it is incredibly limited changes the playing field completely. Things that move slower than light don't matter that much.

No, you build the RKV out of Kuiper belt material in the target system. It costs a power system and manoeuvre drive. Or if you don't think you can hold off the desperate attempts to seize your construction site, you include a FTL drive on the RKV and start the acceleration immediately after arriving in system.

RKV fails if the system drive doesn't impart momentum on the vehicle. Short of that, they are pretty impossible to stop.

 

[Zoolimar]

Or if you don't think you can hold off the desperate attempts to seize your construction site, you include a FTL drive on the RKV and start the acceleration immediately after arriving in system.

So you either spend weeks to months building a large enough RKV to bust the target world in an enemy system or build it in your system and then send to another system. And either way it needs months to get up to speed. In fact until you have anti-matter torches you can't even get RKV required speeds inside a single star system without it doing incredibly large elliptic orbits for like half a year if not more.

And all this time your opponent supposedly does nothing.

 

[-Marauder-]

Well we just need to account for the fact that RKVs would take 40-100 years to go through 1 jump. 3-4 hyperlane jumps and you are looking at target hit 300-500 years since you launched the RKV. That's assuming ships with FTL can't ever intercept them. FTL unless it is incredibly limited changes the playing field completely. Things that move slower than light don't matter that much.

Hyperlanes and Jump Drives exist. Which means you could literally just yeet them through those. Alternatively have a singular ship jump in and paste every planet in the system. Once these things are launched they accelerate so quickly and to such a speed they'd be unstoppable. Which is why I personally would rather not see them in the game.

So you either spend weeks to months building a large enough RKV to bust the target world in an enemy system or build it in your system and then send to another system. And either way it needs months to get up to speed. In fact until you have anti-matter torches you can't even get RKV required speeds inside a single star system without it doing incredibly large elliptic orbits for like half a year if not more.

And all this time your opponent supposedly does nothing.

You don't need a large RKV, you just need a fast RKV. Something that even humans could already do with an Ion drive and the likes. It would just take some time to pick up speed. Given the available FTL in Stellaris, this would be a complete non-issue. In theory, a corvette could serve as an RKV. Going fast enough it would do insane damage to a planet.

Now that's a waste of a perfectly good corvette. You could simply strap a Jump Drive or two to an asteroid in an afternoon and just send it off on its journey. The scary part about this is, defending against this would take much, much, much, much, much, much higher technology and capability than making it happen. There'd be virtually no warning before it arrives, because of how fast it's going.

 

[Zoolimar]

You don't need a large RKV, you just need a fast RKV. Something that even humans could already do with an Ion drive and the likes. It would just take some time to pick up speed. Given the available FTL in Stellaris, this would be a complete non-issue. In theory, a corvette could serve as an RKV. Going fast enough it would do insane damage to a planet.

a) We don't have the capability. You need at least fusion drives to get the speed needed for around 10%c and it will need years to get there. Ion drives don't have the needed exhaust speed to do it.
b) that time needed is months even for an anti-matter drive which in a world with FTL drives is a lot.
c) lasers always move faster and you can have a lot of lasers even at the start of the game to defend yourself with
d) there is still certain minimal size needed because even anti-matter drives give you top speeds of 30-40%c unless you use the most effective thrusters but then you will need years to get up to speed or super tech - at which point RKV would be irrelevant
e) RKV can be destroyed by any debris getting in its path

There'd be virtually no warning before it arrives, because of how fast it's going.

Stellaris nations have FTL sensors. You literally can see and communicate with ships at FTL speeds. RKV compared to that is nothing.

If you want to realistically hurt a planet on a budget just strafing it with a ship loaded with fusion bombs is much easier, costs less and removes all that unneeded acceleration part where you either orbit around the sun on ridiculous orbits or spend decades getting up to speed moving from system to system at STL speeds. The fact that Stellaris models planetary bombardments as some kind of anemic activity is another matter.

 

[sillyrobot]

So you either spend weeks to months building a large enough RKV to bust the target world in an enemy system or build it in your system and then send to another system. And either way it needs months to get up to speed. In fact until you have anti-matter torches you can't even get RKV required speeds inside a single star system without it doing incredibly large elliptic orbits for like half a year if not more.

And all this time your opponent supposedly does nothing.

Yeah, because size is almost immaterial. A corvette-sized rock moving at 99.999% c will crack a planet (10 kton rock will impact with 5 x 10^10 megatons of TNT equivalent). Who cares if the impact takes between a year to half a decade? That's a "short" war in a universe where a single fleet engagement can take multiple months. We have no idea what accelerations are possible -- power is unlikely to be the limiting factor. Since the rock has no crew, it can maintain an acceleration beyond what a crewed ship can achieve and thus be pretty much impossible to engage.

 

[Zenopath]

In general terms, the OP is correct. The tactical level of space battles is non existent and I'm not sure how that could be fixed.
But strategically, you do have options when faced with an enemy that has a bigger fleet than you. Most obvious is; don't engage them. One good way to deal with a superior force is to put a lot of ground troops on a planet in it's path. If the enemy is a group of fleets rather than just 1 stack, it's very likely they will leave a single fleet bombing that world and continue invading with the rest. You can get them to seperate you can defeat them in detail, fighting with locally superior forces. Or you can counter invade, forcing them to send a fleet back into their own territory to try to chase down your invading force. Or you could send small corvette fleets to dance around their fleets, try to lure them away from important targets while you build up forces. Or you could simply turtle, park your fleet at a well defended bastion with many defense platforms and wait for them to come to you.

It's not completely hopeless if you are attacked by a stronger force, but sometimes, the best solution is just to surrender and accept a humilation loss of 100 influence rather then let your fleet be chopped up to ribbons. Then you build up your forces and prepare an invasion of your own in 10 years.

 

[Baldamundo]

1. I never said I want only a hoi4 MtG implementation but to use Paradox's best renowned knowledgeable concepts, which include HOI4 ( the HOI 3 of paradox 2.0 like the OOB) has it central tenet for it's shared design philosophy with Stelaris.
2. This will indeed change it's DNA like Megacorp have done (what should have been Apocalypse instead of a band-aid). This may be impossible to implement fully with the current state of developement but should definitly be considered for a hypotetic sequel for a 4X (should be a 5X like MOO3 was destined to be) GSG. Like the tiles system, the solar system should not be rendered unless Paradox want a 4x RTS game akin to SoaSE.

Increasingly think this is actually the answer - one of the fundamental problems with Stellaris is that it has a bunch of incoherent RTS elements clumsily implemented halfway. They basically just need to bite the bullet and decide if it's trying to be an Sins of a Solar Empire style RTS or a more abstract strategic naval simulation like Hearts of Iron. The fleet system's changed a lot since release, but each iteration's still had the same basic confusion

 

[MichaelJanuary]

RKVs in a stellaris type universe may not be as effective as you might imagine.

If one civ can fit engines to asteroids or other bodies and start accelerating them, so can the other. With FTL sensors, they will just end up playing space billiards, throwing rocks around the star system. At the least, the defending side could have prepared by keeping a number of large asteroids (trojans) in handy orbits that they could just move into place as a blocking/deflecting force.

Remember, the scenario here are two Civs with equivalent techs, so they can both play the same game, offensively and defensively.

 

[Maethendias]

not only that, but there is currently an incredible "exploit" or problem with things like "retreats" from battle, which do not give correct war attrition, which means, you could utterly destroy a whole armada of an enemy, and GAIN war attrition, cause you lost 1 corvette, while their 60 corvettes mostly just.,, "retreated", which obviously makes no sense, and utterly removes 2 weaponstyles out of 3 from the game completly

 

[Pancakelord]

honestly the doomstacking problem seems unfixable without adding utterly unrealistic penalties to the game. i can only think of a few ways that could work and they all involve debuffing fleets based on how many ships there are in a system and that is just plain unrealistic.

What happens when lots of units stack up in a military conflict? be it land, sea or air? you use explosives - Artillery, nuclear weapons or airburst devices (i'm grossly oversimplifying here). This is also true of most videogames, particularly strategy games and MMOs (e.g. in "10 man" trials, raids whatever) If too many units or men stack up on one spot theres usually some sort of mechanic that can hit them all concurrently whilst doing a lot of damage.

This is what Stellaris needs: Area of Effect (AOE) weapons. We had the code to add stuff like that pre 2.0 but

1. it was ripped out
2. it was never used to it's full potential

Today people decry (rightfully) the state of PD vs Missiles. Either you have too many missiles and overwhelm the enemy (e.g. prior to G slots being added) or the enemy has too much PD and virtually no missiles get through. Not enough for them to be competitive anyway. Likewise we have Ion cannons, massive expensive and ultimately useless weapons that are supposed to make starbases hard to crack, instead their slow rate of fire means they get easily engulfed by swarms of ships. These both suffer from the same issue, they are rather specialised weapons but they don't do enough damage to be relevant vs the hordes of ships they go up against.

The solution? Make missiles and Ion cannon deal AOE damage, If an Ion cannon fires at a ship, and hits, the explosion would then harm all other ships in a sphere/circle (of some radius, e,g, 2x the length of a corvette) around the target. same for missiles (but to a lesser degree as they're ship-mobile).

• They could be raw DPH weapons (Damage per hit - I.e. a "nuke") which would serve as an effective way to weed out hordes of corvettes as each shot would damage or outright kill multiples of them all packed together (another issue the game's had since its inception).
• Or the weapons could deal debuffs to the properties of ships - e.g.
  • Evasion. you could have a "gravitic ion beam" (total scifi word salad) that reduces the evasion of all ships it hits in the explosion radius by 80% for N days, making them easy pickings for other defence cannons or allied fleets.
  • Speed. ships hit by "shackle missiles" could be slowed to a stop, keeping them outside of engagement range, or keeping your ships at range.
  • Sensor jammer / "electronic warfare" warheads, they do minimal damage but upon detonation, warheads reduce the tracking and/or accuracy of enemy ships in the blast radius.

The reason people doomstack is because there are literally no hard counters to doomstacking. Its a strategy with only upsides, no downsides. The AI does it because it knows no better.
Trying to add other modifiers onto the fleets does not solve the root problem: After 4 years, Stellaris still does not have any viable counterplay to doomstacking.

• Its like zerging in other games, you overwhelm your enemy with a tidalwave of fodder, with the only counter being bombs, nukes or even magical spells and explosions (depending on the setting).

If you suddenly add AOE weapons into the game you'll quickly see

• people go "oh shit" and start using smaller fleets for all but the most difficult battles, as the losses would be too great with so many ships packed on top of each other (either in a single big fleet or many small fleets all in a conga-line),
  • If you dont want an "AOE alpha-strike" you could
    • make ion-cannons deal some fancy scifi type of disruptor energy that deals the damage over several ingame combat days,
    • or have missiles with corrosive/nanite/hive-spore warheads that will slowly eat into enemy armor and superstructures, dealing ticking damage over time ("DOT") to support the overall damage of other defences or fighting ships.
• it would also make starbases a useful fortification, rather than a mid-game road-bump,
  • they still might not "out-dps" a fleet or last too long on their own, but a starbase with some Ion cannons firing volleys of AOE shells could shred any corvette swarms thrown at it, leaving the capital ships exposed and easy to kill by your own fleets.
  • This might even force players to build a fleet to counter enemy starbases specifically (gasp, strategy? In my Stellaris?!)

 

[CrazyJ]

Spoiler: your quote

What happens when lots of units stack up in a military conflict? be it land, sea or air? you use explosives - Artillery, nuclear weapons or airburst devices (i'm grossly oversimplifying here). This is also true of most videogames, particularly strategy games and MMOs (e.g. in "10 man" trials, raids whatever) If too many units or men stack up on one spot theres usually some sort of mechanic that can hit them all concurrently whilst doing a lot of damage.

This is what Stellaris needs: Area of Effect (AOE) weapons. We had the code to add stuff like that pre 2.0 but

1. it was ripped out
2. it was never used to it's full potential

Today people decry (rightfully) the state of PD vs Missiles. Either you have too many missiles and overwhelm the enemy (e.g. prior to G slots being added) or the enemy has too much PD and virtually no missiles get through. Not enough for them to be competitive anyway. Likewise we have Ion cannons, massive expensive and ultimately useless weapons that are supposed to make starbases hard to crack, instead their slow rate of fire means they get easily engulfed by swarms of ships. These both suffer from the same issue, they are rather specialised weapons but they don't do enough damage to be relevant vs the hordes of ships they go up against.

The solution? Make missiles and Ion cannon deal AOE damage, If an Ion cannon fires at a ship, and hits, the explosion would then harm all other ships in a sphere/circle (of some radius, e,g, 2x the length of a corvette) around the target. same for missiles (but to a lesser degree as they're ship-mobile).

• They could be raw DPH weapons (Damage per hit - I.e. a "nuke") which would serve as an effective way to weed out hordes of corvettes as each shot would damage or outright kill multiples of them all packed together (another issue the game's had since its inception).
• Or the weapons could deal debuffs to the properties of ships - e.g.
  • Evasion. you could have a "gravitic ion beam" (total scifi word salad) that reduces the evasion of all ships it hits in the explosion radius by 80% for N days, making them easy pickings for other defence cannons or allied fleets.
  • Speed. ships hit by "shackle missiles" could be slowed to a stop, keeping them outside of engagement range, or keeping your ships at range.
  • Sensor jammer / "electronic warfare" warheads, they do minimal damage but upon detonation, warheads reduce the tracking and/or accuracy of enemy ships in the blast radius.

The reason people doomstack is because there are literally no hard counters to doomstacking. Its a strategy with only upsides, no downsides. The AI does it because it knows no better.
Trying to add other modifiers onto the fleets does not solve the root problem: After 4 years, Stellaris still does not have any viable counterplay to doomstacking.

• Its like zerging in other games, you overwhelm your enemy with a tidalwave of fodder, with the only counter being bombs, nukes or even magical spells and explosions (depending on the setting).

If you suddenly add AOE weapons into the game you'll quickly see

• people go "oh shit" and start using smaller fleets for all but the most difficult battles, as the losses would be too great with so many ships packed on top of each other (either in a single big fleet or many small fleets all in a conga-line),
  • If you dont want an "AOE alpha-strike" you could
    • make ion-cannons deal some fancy scifi type of disruptor energy that deals the damage over several ingame combat days,
    • or have missiles with corrosive/nanite/hive-spore warheads that will slowly eat into enemy armor and superstructures, dealing ticking damage over time ("DOT") to support the overall damage of other defences or fighting ships.
• it would also make starbases a useful fortification, rather than a mid-game road-bump,
  • they still might not "out-dps" a fleet or last too long on their own, but a starbase with some Ion cannons firing volleys of AOE shells could shred any corvette swarms thrown at it, leaving the capital ships exposed and easy to kill by your own fleets.
  • This might even force players to build a fleet to counter enemy starbases specifically (gasp, strategy? In my Stellaris?!)

i love all of these but i question whether the devs could implement them without breaking something important. every time a core mechanic changes the devs break something, and they don't usually fix it these days, hell the ai has been crap since 2.0 and has steadily gotten worse. crisis have been worthless since the ftl debacle, micro has been hell since 2.2 and performance has been crap since 2.2. . sectors have never worked properly and likely never will, not technically a change but still worth pointing out. large mechanical changes don't end well with stellaris.

My point is that if the devs implement this sort of change we don't what would it break and how could they fix it, or if they even bother trying. my suggestions to fix the game are almost always rooted in changing numbers and not changing mechanics because there is way less of a chance of screwups that will never be resolved. seriously I'm pretty sure there is still to-do code in the planet files left over from 2.2, it's that bad. if they screw up a number anyone can fix it, a mechanic like aoe if coded incorrectly could cause all kinds of havoc. finding a way to add something within the current system is the best bet

if they just add a modifier or change a number, and it turns out the change isn't enough or is too much it can easily be tuned by a modder til it actually works, but with new mechanics they are usually extremely difficult or impossible to mod out if they just don't work. why do you think there aren't any mods that bring tiles back or ftl types back?

the games broken enough as it is, i am somewhat opposed to any idea that could result in an even more broken game. for now fleet combat works, but if the devs start screwing with it there is no guarantee it will stay that way.

 

[Oscot]

The reason people doomstack is because there are literally no hard counters to doomstacking. Its a strategy with only upsides, no downsides.

Wellllllll... there is a downside: when you only have one fleet, you can only cut one swath of destruction through enemy territory at once. Two fleets means twice the speed of planet capture, twice the speed of arriving at your claims objectives.

The problem is that this isn't enough of a downside because (perennial Stellaris complaint) no-one bothered to try and synchronise mechanics to each other. Slooooowwww conquest progress on the galactic map is actually no problem to the aggressor in Stellaris because the way war exhaustion mechanics are calculated, it's always necessary to basically bore your enemy into submission through time more than anything else. When the war's going to last for ~10 years anyway, there's no disadvantage to crawling through their territory at a snail's pace and popping off worlds in serial, especially when sieges / ground combat are so fast compared to fleet movement.

Contrast EUIV, where if you don't start your fort sieges on day 1, you're gonna do badly. Time is a resource to spend wisely in EUIV wars. In Stellaris, conversely, you're looking for any way to waste it.

 

[methegrate]

Wellllllll... there is a downside: when you only have one fleet, you can only cut one swath of destruction through enemy territory at once. Two fleets means twice the speed of planet capture, twice the speed of arriving at your claims objectives.

The problem is that this isn't enough of a downside because (perennial Stellaris complaint) no-one bothered to try and synchronise mechanics to each other. Slooooowwww conquest progress on the galactic map is actually no problem to the aggressor in Stellaris because the way war exhaustion mechanics are calculated, it's always necessary to basically bore your enemy into submission through time more than anything else. When the war's going to last for ~10 years anyway, there's no disadvantage to crawling through their territory at a snail's pace and popping off worlds in serial, especially when sieges / ground combat are so fast compared to fleet movement.

Contrast EUIV, where if you don't start your fort sieges on day 1, you're gonna do badly. Time is a resource to spend wisely in EUIV wars. In Stellaris, conversely, you're looking for any way to waste it.

Not to mention that when it comes to war exhaustion and claim score, the value of destroying their fleet vastly outweighs the value of taking planets.

You can take every planet in the empire, but they won't surrender until you blow up the damn fleet. So there's no reason to split the fleet for defensive purposes. The best move is to keep it intact and keep it together, knowing that sooner or later they'll have to confront you.