Stellaris Dev Diary #380 - Defenders in the Stars

[Ajacoba]

Well yes and no, you can spread damage across ships when a doomstack rolls in — but stations really can’t. They’re capped, limited, and stuck with hard platform limits per system. It’s always a numbers game, and stations will never outnumber fleets. So you either dump time and resources into fortifying one system just to slow a doomstack, or you build your own. And sure, once you're deep into soft cap and your economy’s stretched, maybe you start throwing alloys into citadels — but by then, you probably don’t need them. - Oh well, you could just make stations hilariously OP and create the precedent that fleet power numbers aren’t a real metric anymore. But where’s the fun in that?

That’s the problem. If orbital rings or stations were actually viable, we wouldn’t even be having this conversation. This Deep Space Citadel feature is just another patch on top of a broken foundation — a way to see if defenses can feel meaningful, without addressing the core issue: the game is built around doomstacks. The mechanics that make it shallow — uncapped fleets, simplified combat, brute-force meta — those are untouched. And yet we’re capping the few things that might actually diversify play. It's not that depth isn’t possible — it’s that this isn’t the way to get there.

Idk, even with just 30 platforms per citadel, which is a relatively low amount, you would have a total of 120 defense platforms

All shooting at least 2 torpedoes each at any fleet jumping into the system, that's gonna hurt your artillery fleets pretty badly

And that's not even maximum numbers, you can breach 50 capacity on a starbase, I assume citadels will be similarly silly, so you may be looking at like 200 defense platforms, which a bunch of fancy auras

That's a lot of cheap firepower, on top of the fleets your target has

You may be able to outclass the citadels eventually, but you will never be able to outclass the citadels reinforcing your victim's fleets

With the citadels and the starbase you can also stack a whole bunch of silly auras, so that anything that attacks gets lower stats and anything that defends gets higher stats

 

[Critical Ethics]

Bad choice of words on my part. I meant a system where fleet presence is measured/supported per system. I don’t know what the game designers wanted to come up with — I’d guess a hard cap (lol). If it were me, I’d just have a soft cap that, if overstepped, ramps up costs and decreases effectiveness.

I think we're on the same page we're just quibbling over per ship vs per fleet and per system vs per combat.

Basically a doomstack isn't a doomstack unless it's in combat. Out of combat it's just a bunch of ships that happen to be near each other. Effectiveness only matters in combat so there's no point in applying it out of combat, and costs out of combat lead to weird incentives like microing all your fleets one system apart to converge on the actual combat or swearing because two fleets crossed paths on different missions and dinged your treasury just by passing by each other.

So I'd be looking for out of combat:
Jam as many ships as you like wherever you like. Density doesn't change anything. Maybe have a note giving out if it's a silly amount that could cause problems if a fight broke out.

But when combat starts:
-> perform some kind of number of ships in combat (or in system) check
--> oh that's like twice the "correct" amount, I'm gonna multiply your in-combat logistics spends by a chunk lol.
--> and I'm gonna give the other guy some bonuses.
--> and a bunch of your excess ships don't get a bunch of bonuses* they're normally entitled to.

It sounds like some form of the first one is coming in, the second one is kinda already there but could be beefed up a bit, and we'd both like the last one (*except penalties feel worse than losing expected bonuses even if the result is identical)

 

[Toby-wan_Strife]

Part of me is unimpressed by this. From the artwork I expected it to be like the Maginot worlds from the gigastructures mod. I expected it to be a massive complex the size of an entire planet. I think time will tell whether or not this will be as useful as I think it will be. It honestly depends on how many defensive platforms you can build on it. If it's anything like a fully upgraded and fully defended citadel in the late game, it could potentially be enough to take on a 25x crisis.

Nevertheless, I'm excited to finally be able to properly defend the L-cluster.

 

[endersaim]

Unlike orbital rings and regular starbases, Deep Space Citadels can be placed far more freely within a system, provided they aren’t too close to a gravity well (or each other).

I'm old enough to remember when this was how defense platforms worked in general.

 

[stilgarpl]

I'm old enough to remember when this was how defense platforms worked in general.

And in another patch districts will be organized into a neat grid, so that we'll be able to see terrain features and build buildings at the same time. And perhaps we'll even get adjacency bonuses!

 

[namewhichisnottakenyet]

I've scanned through the entire thread and couldn't find anything about it, so I think this hasn't been brought up yet:

With yet another megastructure in the game, I suggest you have another look at the megastructure construction UI and consider overhauling it.

With all the once-off megastructures, several habitat buildings, orbital rings, ringworlds and now DSCs, the UI is getting really cluttered.

In pretty much every game I have to think about whether or not I've already built this or that unique megastructure because they keep appearing in the build list even if I can't build them anymore, or if they're just greyed out because I selected the wrong celestial body. Like, does the mega shipyard go around a star and the strategy center around a planet or vice versa? Or did I just forget about the sentey array at the edge of the system and that's why I can't build anything else?

Adding an option to remove all megastructures for which one has reached the build limit already would already clean up the UI a lot, but maybe it makes sense to build two separate UIs, one for megastructures which can only be built once (together with a few quality of life features to identify them on the galaxy map at a glance or search for possible sites for new constructions), and one for structures which can be built multiple times.

 

[Grenartia]

Putting a hard system or combat limit on the number of ships in a fight doesn't work in Stellaris because ships move and, unlike SotS, Stellaris doesn't have a distinct battle map. "You can't build that here" is straightforward and only comes up once, but you're either preventing ships moving into a system or having a bunch of ships that are visibly "there" but not fighting, both of which would be irritating. "GUI" is not a real solution to either of those, unless we're talking removing in-system views from the game except during combat and working from there which I'd be fine with tbh but I'm probably in a minority.

I would be absolutely fine with a soft cap where your "combat" fleet cap is (base fleet size) + (a bit per extra admiral present) + a bit more depending on the size of the enemy fleet) and every ship after that doesn't get admiral bonuses and such, though it would be hard to implement that transparently with the current setup.

Yep. Doomfleets are appealing in general and in Stellaris in particular because:
1) Using a big pile of ships to obliterate a smaller pile of ships is fun.
2) Managing one big pile of ships is easier than managing several small piles of ships.
3) Using a big pile of ships to obliterate a smaller pile of ships loses you less ships than using a small or medium pile of ships to fight a small pile of ships.
4) Using a small pile of ships to fight a medium or big pile of ships or a medium pile of ships to fight a big pile of ships just doesn't work.

It'd be nice to fix everything with a single magic bullet but sometimes you need to nibble away at a problem pieces by piece. It's fine to come at each of these one by one as long as the end result is a coherent whole.

Point 1 isn't a problem except that any solution would ideally allow for point 1 to still be an option.

Fleet limits help with point 2 without getting in the way of point 1, and that they don't solve points 3 to 4 doesn't take away from that.

Good doomfleet mitigation for points 3 and 4 (and rewards for putting the extra effort into point 2) include rewarding the player for bringing a knife to a knife fight, a gun to a gun fight, and a nuke to a nuke fight (and yes we're all picturing the same starship troopers scene but let's leave that for now). I like the trade cost concept for part of this because it mitigates the economic incentive to always bring the nuke. You may more decisively win the knife vs gun fight but you had to pay out extra for bullets, and you're not actively (or effectively) prevented from bringing the gun or nuke if that's what you'd have more fun with. Numbers will be tricky though - finishing a fight faster due to bringing more ships could end up resulting in savings anyway.

It could be neat for if debris scaled inversely with the relative size of your opponent, since unloading a few dozen battleships onto a half a dozen cruisers will result in fewer intact cruisers than a fairer fight. Or is that already a thing?

I think a good way of doing that knife vs gun vs nuke fight thing, would be if ships carried limited munitions/ammo.

Obviously, energy weapons would require minimal ammo (lasers probably none except ship energy, which gets regenerated every tick), kinetic weapons a little more, scaling with slot size, missiles and physical torpedoes more than that, but with all their existing advantages, and strike craft highly limited (still in like, the 100s, but a lot less than what other weapon types have), but also carrying their own weapon supplies (this would mesh really well with implementing designable strike craft). For simplicity's sake, each weapon component's ammo would ideally be pooled with similar weapons on each ship, and that pool would basically just be HP that goes down each time the weapon is used. Probably would also open up a niche for combat replenishment ships (maybe another good one for cruisers?). That one could be implemented as a short-ranged aura, but handwaved as an abstraction layer (in practice, kind of like the electronic warfare ships from the Realistic Ships mod, though, hopefully, with a bit more customizability in the sections).

 

[Critical Ethics]

I think a good way of doing that knife vs gun vs nuke fight thing, would be if ships carried limited munitions/ammo.

Obviously, energy weapons would require minimal ammo (lasers probably none except ship energy, which gets regenerated every tick), kinetic weapons a little more, scaling with slot size, missiles and physical torpedoes more than that, but with all their existing advantages, and strike craft highly limited (still in like, the 100s, but a lot less than what other weapon types have), but also carrying their own weapon supplies (this would mesh really well with implementing designable strike craft). For simplicity's sake, each weapon component's ammo would ideally be pooled with similar weapons on each ship, and that pool would basically just be HP that goes down each time the weapon is used. Probably would also open up a niche for combat replenishment ships (maybe another good one for cruisers?). That one could be implemented as a short-ranged aura, but handwaved as an abstraction layer (in practice, kind of like the electronic warfare ships from the Realistic Ships mod, though, hopefully, with a bit more customizability in the sections).

Fights eating logistics is really just a more abstract version of this. Firing guns eats up bullets and laser coils and stuff, these all take misc resources to replace (represented by increased logistics) and there's additional transport overhead if you're in unfriendly territory (represented by more logistics).

A big advantages of using an empire level logistics instead of a per-ship bar is that at the scale Stellaris operates under we can't really pay attention to what an individual ship is up to. Individual ships running out of bullets is just going to get lost in noise. Seeing your trade stockpile plummet is much more visible and something you can actually react to.

A fleet or combat level bar could work from a scale point of view but smaller fleets would obviously have a smaller bar, so you'd have even more incentive to doomstack due to the possibility of a small fleet failing to kill a bigger fleet because they flat ran out of bullets!

A bunch of special weapons that eat an ammo bar that depletes during combat and can only refill out of combat and eats trade and/or SR reserves as it does could be cool though...

 

[Imp0815]

I think we're on the same page we're just quibbling over per ship vs per fleet and per system vs per combat.

Basically a doomstack isn't a doomstack unless it's in combat. Out of combat it's just a bunch of ships that happen to be near each other. Effectiveness only matters in combat so there's no point in applying it out of combat, and costs out of combat lead to weird incentives like microing all your fleets one system apart to converge on the actual combat or swearing because two fleets crossed paths on different missions and dinged your treasury just by passing by each other.

So I'd be looking for out of combat:
Jam as many ships as you like wherever you like. Density doesn't change anything. Maybe have a note giving out if it's a silly amount that could cause problems if a fight broke out.

But when combat starts:
-> perform some kind of number of ships in combat (or in system) check
--> oh that's like twice the "correct" amount, I'm gonna multiply your in-combat logistics spends by a chunk lol.
--> and I'm gonna give the other guy some bonuses.
--> and a bunch of your excess ships don't get a bunch of bonuses* they're normally entitled to.

It sounds like some form of the first one is coming in, the second one is kinda already there but could be beefed up a bit, and we'd both like the last one (*except penalties feel worse than losing expected bonuses even if the result is identical)

I see your approach is to keep things simple when not at war, but really attention-hogging when at war.

I wouldn’t recommend splitting a supply system into states, but rather applying it universally — to make it easier for players, automation, and the AI.

You get a limit of ships/fleets per system, provided by something else or as a ruleset, so positions can be automated more easily and the system is more transparent and understandable.

But I think this is all derailing the topic, and I fear getting my posts removed again for going off-topic.

So to swing back around: Deep Space Citadels are a cool concept, and I like the idea of a valid defensive playstyle — but until we address the reason for needing a third defensive structure just to slap a band-aid on that playstyle and desperately try to make it competitive with doomstacks, it’s just frustrating once you understand the core issue of the game.

 

[Imp0815]

Idk, even with just 30 platforms per citadel, which is a relatively low amount, you would have a total of 120 defense platforms

All shooting at least 2 torpedoes each at any fleet jumping into the system, that's gonna hurt your artillery fleets pretty badly

And that's not even maximum numbers, you can breach 50 capacity on a starbase, I assume citadels will be similarly silly, so you may be looking at like 200 defense platforms, which a bunch of fancy auras

That's a lot of cheap firepower, on top of the fleets your target has

You may be able to outclass the citadels eventually, but you will never be able to outclass the citadels reinforcing your victim's fleets

With the citadels and the starbase you can also stack a whole bunch of silly auras, so that anything that attacks gets lower stats and anything that defends gets higher stats

Ok, I see your point — so we go with the last option: "make them silly."

We’ll see how this turns out and how it balances against doomstacks in the game. This might work, but it still doesn’t make me happy, as I see it as a band-aid and yet another feature that misses the main issue.

 

[Critical Ethics]

I see your approach is to keep things simple when not at war, but really attention-hogging when at war.

I wouldn’t recommend splitting a supply system into states, but rather applying it universally — to make it easier for players, automation, and the AI.

Having it on all the time would be far more attention-hogging during war. Multiple fleets traveling to targets through allied or captured enemy territory would need to make sure they spent the entire time in completely different systems to prevent dings to your treasury. Imagine trying to factor that into automatic pathfinding. Imagine trying to keep track of that yourself. Imagine trying to teach that to the AI, both opponents and allied fleets.

Combat is already a different system state. Having different rules apply already happens.

You get a limit of ships/fleets per system, provided by something else or as a ruleset, so positions can be automated more easily and the system is more transparent and understandable.

Again, are we talking about a hard limit of ships per system or a soft cap? What happens if you have ships in a system, it's "full", and you want to send a different fleet through that system to the other side? Don't say "Magic GUI".

But I think this is all derailing the topic, and I fear getting my posts removed again for going off-topic.

So to swing back around: Deep Space Citadels are a cool concept, and I like the idea of a valid defensive playstyle — but until we address the reason for needing a third defensive structure just to slap a band-aid on that playstyle and desperately try to make it competitive with doomstacks, it’s just frustrating once you understand the core issue of the game.

Static defences aren't supposed to be competitive with doomstacks. They're supposed to provide a numbers boost at lynchpin systems for fleet-to-fleet fights. Static defences shouldn't be completely untouchable by pure fleet combat, that would be an awful, terrible way to set up the game.

 

[Nabil ad Dajjal]

Arguing against a supply soft cap because it would make microing fleets around to re-form doomstacks in combat challenging misses the point.

The whole purpose of that kind of soft cap is to disincentivize putting all of your fleets / ships into one massive engagement. If it's obnoxious to keep all of your fleets together, that means that it's working; the soft cap is meant to signal "you're playing wrong, have those fleets fighting along separate fronts instead."

 

[Calvax]

Arguing against a supply soft cap because it would make microing fleets around to re-form doomstacks in combat challenging misses the point.

The whole purpose of that kind of soft cap is to disincentivize putting all of your fleets / ships into one massive engagement. If it's obnoxious to keep all of your fleets together, that means that it's working; the soft cap is meant to signal "you're playing wrong, have those fleets fighting along separate fronts instead."

A soft cap on its own won't do that. If the idea is that there's a limit per system, per empire, and above that limit there's diminishing returns then using a larger fleet still wins. If it's a hard cap war becomes tedious since you have to micro each battle front to drip feed in new fleets as space becomes available.

 

[grommile]

you're playing wrong, have those fleets fighting along separate fronts instead."

There may only be one front available, because the default galaxygen settings are tuned to provide chokepoints and you and your enemy may both be surrounded by empires that hate both of you and won't give military access.

 

[Nabil ad Dajjal]

A soft cap on its own won't do that. If the idea is that there's a limit per system, per empire, and above that limit there's diminishing returns then using a larger fleet still wins. If it's a hard cap war becomes tedious since you have to micro each battle front to drip feed in new fleets as space becomes available.

I disagree.

CK3, for example, has a supply soft cap and it actually does a decent job at punishing doomstacks (it's not perfect, but still miles better than Stellaris in that sense). If you try to keep your armies in a single huge column then you'll end up starving them before long which degrades performance in combat: it's usually better to keep multiple smaller columns with distinct tasks. And this leads to more reasonable behavior overall, for example the ability to defeat enemy forces in detail.

Putting similar supply soft caps and attrition mechanics into Stellaris to punish doomstacks would be a huge benefit.

 

[grommile]

Putting similar supply soft caps and attrition mechanics into Stellaris to punish doomstacks would be a huge benefit.

Only if the bots can be taught to recognize when the correct solution is "blow past the cap" and when it isn't.

(And if it's never the correct solution, then be intellectually honest and make it a hard cap.)

 

[Calvax]

I disagree.

CK3, for example, has a supply soft cap and it actually does a decent job at punishing doomstacks (it's not perfect, but still miles better than Stellaris in that sense). If you try to keep your armies in a single huge column then you'll end up starving them before long which degrades performance in combat: it's usually better to keep multiple smaller columns with distinct tasks. And this leads to more reasonable behavior overall, for example the ability to defeat enemy forces in detail.

Putting similar supply soft caps and attrition mechanics into Stellaris to punish doomstacks would be a huge benefit.

I'm not totally against ideas of logistics and a soft cap but also not sold you can just port systems from other games wholecloth. In CK3 the map is the world map, most provinces connect to all of their neighbours with the exception of rare impassable terrain or the sea which allows travel to everywhere. There's no such thing as closed borders and no structures that can pin an enemy in a province and force approach by specific angles.

All of that is completely different in stellaris where we have hyperlane geography. In Ck3 you can easily split your armies, spread them into neighbouring provinces, and let them resupply (only to doomstack them again when the enemy approaches). That's not the case in stellaris where there might only be one entrance to the enemy you're fighting so you can't split up and attack from all angles.

 

[Critical Ethics]

I disagree.

CK3, for example, has a supply soft cap and it actually does a decent job at punishing doomstacks (it's not perfect, but still miles better than Stellaris in that sense). If you try to keep your armies in a single huge column then you'll end up starving them before long which degrades performance in combat: it's usually better to keep multiple smaller columns with distinct tasks. And this leads to more reasonable behavior overall, for example the ability to defeat enemy forces in detail.

Putting similar supply soft caps and attrition mechanics into Stellaris to punish doomstacks would be a huge benefit.

If I have three fleets and I can micro them around the map carefully staggering them to avoid hitting any penalties, then dump them all into the one fight to doomstack anyway, then there is no actual mechanical penalty being applied before the fight, just an attention and frustration penalty. Balancing based on how much annoyance your players can tolerate is not a good approach.

So we have two options:
1) Change everything about how stellaris handles map generation, fleet movement, and combat at both an engine level and game concept level.
2) Only apply the soft cap penalties during battles, when doomstacking actually matters.

"Apply a system cap simple as" only seems simple if you ignore the context of the entire rest of the game, as most "do X simple as" claims tend to. "Apply a system cap and then completely rewrite the game from the ground up to accommodate this huge, sweeping change" has less of a ring to it though.

 

[Imp0815]

1) Change everything about how stellaris handles map generation, fleet movement, and combat at both an engine level and game concept level.

To reduce my rambling to a minimum — this is basically the only complete solution to the problem.

• Adding system caps (soft or not) with penalties for overstacking would require players to spread fleets across more locations.
• Stacking anyway should be a costly solution, de-emphasized by exposing other parts of the war front.
• Dispersion of force — like a realistic, believable star empire would do — should be the go-to.
• A complete, solid deployment and automation system for fleets would be a must-have to avoid excessive micromanagement.

Following this military rework, a redesign of the gameplay loop and economy is necessary, as you will eventually reach a point where additional fleets are no longer needed.
So additional game goals — beyond growth and endless militarization — must be introduced.
Different power mechanics like politics, utopian society systems, espionage, technology, and other avenues must be designed.

I’m just repeating myself — and what’s written in my signature here on the forum — but an adaptation of the Hearts of Iron IV land army and combat system for Stellaris would be my ideal solution.

Deep Space Citadels — and defenses in general — should have an extensive role in providing border protection, and the game should be balanced around that.

 

[grommile]

At this point you're just saying "write stellaris 2".

(Because otherwise the game keeps dragging through endless interstitial states.)