Doomstack Mitigation Discussion -- hard cap vs. soft cap

[Imp0815]

Starbases can't be destroyed in a normal war, and I don't like the idea that a Total War gives you more supply for free as you conquer systems.

So let's think about some kind of other structure -- not in a Starbase -- which could give Supply.

Perhaps just a custom Kilostructure which can be destroyed, or a non-megastructure (regular) stations like a mine or research station.

Why not redefine what it means to capture a starbase then? It's not like this is set in stone, and capturing a supply hub should be somewhat important.

For example, we could increase the repair time until a starbase is rebuilt and operational again, and have a system where a supply hub needs to be in range of another supply hub to function. This would encourage smart positioning of supply hubs around your empire to prevent enemy empires from gaining extended supply range in your space, while maintaining a stable supply range for your fleets.

In addition, you could create a scorched earth mechanic where you scuttle your station instead of losing it.

Building and maintaining supply hubs could also be made a costly and lengthy affair, making it more feasible to peace out of a war when your supply range is exhausted instead of waiting for a new forward supply base.

Additionally, supply as a mechanic could be extended and perceived differently for different empires, resulting in different military playstyles. For example, a raiding empire could receive supply from raiding planets at the cost of further destroying the planet, thus reducing its value for conquest in exchange for short-term gains.

P.S. Kilostructures would be a good compromise as it somewhat decouples the supply system from the rest, making it easier to understand. But I like the added depth through the decision of "do I want my starbase to be a frontline fortress, or a supply hub to support my fleets?" More decisions make a game a better experience. I vaguely remember there are major game design outlets that define game fun in terms of decisions per minute or something like that, and games become fun when we can make decisions. A decoupled system would take that down a peg.

EDIT: Found it: Sid Meier's Interesting Decisions

 

[HFY]

Why not redefine what it means to capture a starbase then?

Because that's got a lot of potential for unintended consequences, given how starbase ownership works with respect to system ownership, and the nuances of the three war categories (regular, total, crisis).

It seems much more reliable to make Supply a separate thing which has its own clear and simple rules -- fewer chances for bugs & unintended corner-cases to occur.

That said, starbases could interact with Supply -- perhaps each Anchorage pushes Supply range +1 so a system with a focus on Anchorages would also be a great place to build up a Supply depot.

 

[Imp0815]

Because that's got a lot of potential for unintended consequences, given how starbase ownership works with respect to system ownership, and the nuances of the three war categories (regular, total, crisis).

It seems much more reliable to make Supply a separate thing which has its own clear and simple rules -- fewer chances for bugs & unintended corner-cases to occur.

That said, starbases could interact with Supply -- perhaps each Anchorage pushes Supply range +1 so a system with a focus on Anchorages would also be a great place to build up a Supply depot.

I don't see a lot of potential for unintended consequences.

Regular could lead to starbases taking much longer to be operational again.
Total, the same.
Crisis could disable them until recaptured.

On top the option to scorch the starbase. (It's even odd, with all the sci-fi stories and even real-life stories where captains would rather sink their own ships than let them be taken, that we don't have such a depiction in Stellaris yet.)

It's rather unfair, so to speak, that taking a starbase and making it operational again after a short amount of time leads to a disadvantage for the defender while providing benefits for the aggressor.

 

[HFY]

I don't see a lot of potential for unintended consequences.

It's complex code which you can't see at all. So the fact that you can't seen any problems with that complex code isn't particularly informative.

Adding more complexity to already-complex code is how you get bugs, like Orbitals running away from combat, which is apparently very difficult to fix because it's still happening.

Starbases are more complex than Habitats.


I'd just put Supply in a separate, simple structure which does not need to overlap with the system ownership code.

There is no need to stuff the two different things into a single structure, so there's no reason to make it harder to code & maintain.

 

[Ikael]

I am quite a fan of the supply radius idea, which kind of simulates the whole resupply thing without needing to actually move fleets to resupply them. The more important targets you have to defend outside the doomstack, the better!

As for turning starbases into supply depots, I would be down with it, for it would present quite a lot of interesting decisions when designing starbases (do you devote their precious slots to defense, supply, or trade collection, or...?) but only as long as you can destroy them in regular wars.

Alternatively, I also like the whole logistic base-superstructure idea, as long as 1) There is a hard cap on them (to be expanded through traditions and the likes) and 2) As long as you can't place it on a system with a starbase present (so you need to choose between defense and logistics).

And then there would be a third option: Logistic nodes as a starbase module... but with an empire-wide cap. So it doesn't matter that you might blot out the stars with starbaases, not all of them would be able to provide logistic support to your fleets, only those you decide to specialize.

PS: I have been thinking long about the whole radius supply for a lot of its edge cases. I really think we can make this work!

 

[Imp0815]

It's complex code which you can't see at all. So the fact that you can't seen any problems with that complex code isn't particularly informative.

Adding more complexity to already-complex code is how you get bugs, like Orbitals running away from combat, which is apparently very difficult to fix because it's still happening.

Starbases are more complex than Habitats.


I'd just put Supply in a separate, simple structure which does not need to overlap with the system ownership code.

There is no need to stuff the two different things into a single structure, so there's no reason to make it harder to code & maintain.

Are you implying PDX is incapable of making changes to their own written code?

On what basis do you claim it's not possible? And why would we get a patch with such major bugs?

Could we stay in the territory of making changes to existing systems under the premise that PDX is still capable of maintaining and working on their own code?

Continue based on that:

There is no need to stuff the two different things into a single structure, so there's no reason to make it harder to code & maintain.

The "need" would be to enhance the game with new decisions. In addition, later down the line, you will encounter the problem that starbases exist while the fleet deployment is now limited per system, giving defenders the advantage. So reducing the number of fortress stations naturally by having them serve as either fully blown frontline defenses or mitigating their effectiveness by reducing their power because you swapped guns for supply depots is a perfect fit for the system and a valuable addition to the game because this decision is in the player's hands. A separate system where stations stay the same will end up in an awkward place where too many stations are on the field, and we will need a new solution for that.
And placing supply depots without any drawback could become boring, and the question arises: why even bother?

It would be too easy to work around while providing a far too quick, obvious solution to the supply challenge we want to provide here.

It would be wasted potential, missing the opportunity to add depth without unnecessary complexity.

 

[HFY]

Are you implying PDX is incapable of making changes to their own written code?

Are you forgetting the bugs in system ownership from First Contact which have not yet been fixed?

Of course they can make changes, but they are human, and humans can & do make mistakes.

Simpler code means fewer mistakes. When it's possible to do the same thing in a simple way or a complex way, don't pick the complex way, because making changes to a complex thing is more likely to create more errors and worse errors.

On what basis do you claim it's not possible? And why would we get a patch with such major bugs?

We've gotten several patches with major bugs, some quite recently. Changes are risky because humans can make mistakes.

Again, my claim is that making a change to one of the most complex parts of the code -- Starbases and system ownership -- is not a good idea when it's very easy to just put Supply in its own structure with its own simple rules.

"Needlessly risky and error-prone when a simpler solution is also easier" would be my position. "Not possible" is a bad straw-man of that.

And placing supply depots without any drawback could become boring, and the question arises: why even bother?

They could be ships rather than stations or starbase modules, so you could move them around, and so attacking them would be baked into ship behaviors.

They'd have some kind of supply radius, visible on the map like jump radius or trade collection radius, depending on how the range is calculated.

You'd want to put them close to the conflict so the radius would overlap your attack force.

Finding (some of) the enemy's supply ships could be an interesting use of intel and cloaked scouts.

Juggernaut could be changed from a poor shipyard into an excellent supply ship with a fat supply range and high combat survivability (in contrast to regular supply ships which die to Skirmish fleets quite easily).

 

[Imp0815]

"Needlessly risky and error-prone when a simpler solution is also easier" would be my position. "Not possible" is a bad straw-man of that.

We should not aim for the "simplest solution" or "easier" one but the best solution.

They could be ships rather than stations or starbase modules, so you could move them around, and so attacking them would be baked into ship behaviors.

They'd have some kind of supply radius, visible on the map like jump radius or trade collection radius, depending on how the range is calculated.

You'd want to put them close to the conflict so the radius would overlap your attack force.

Finding (some of) the enemy's supply ships could be an interesting use of intel and cloaked scouts.

Juggernaut could be changed from a poor shipyard into an excellent supply ship with a fat supply range and high combat survivability (in contrast to regular supply ships which die to Skirmish fleets quite easily).

Then how would that impact doomstacks? Wouldn't it be easier to keep your supply ships all in range of your doomstack then?

Having the decision to either build a starbase as a supply depot in a static location has far more strategic implications than just building supply ships that tag along for the ride. It sounds like adding something to solve nothing.

You could build up a convoluted system to prevent supply overlap or some arbitrary system to prevent that, but I'm so opposed to the idea that I can't think of any feasible solution to mobile supply.

I mean, you could just put them in systems, chain them from your capital to your fleets, but that would result in some uninteresting, tedious gameplay. Mobile supply bases sound interesting, but I do not see them as a solution to the thread goal here.

I'm not saying that supply depot building and deciding where to put them can't result in something similar, I just find it more engaging to make strategic decisions. For example, deciding if I want offensive supply depots that can be lost and turned against me if my offensive fails, or putting them back in my empire to have a 3-4 system buffer before it gets hairy, forcing my enemy to build up a frontline supply that I can utilize if their offensive fails. This is just one example of what static supply depots and the tradeoffs for stations could result in. From my perspective, engaging and fun gameplay writes itself if you really care about strategy, warfare logistics, and the implications and tradeoffs that result from them.

I fully agree that such a system should elevate the juggernaut (in general, they should be elevated!). Maybe even allow it to be self-sufficient without the need for supply, to have a really heavy-hitting, dangerous flagship that the enemy needs to monitor closely. If it breaks through into your empire, it could force its way to your important worlds and supply depots. This sounds far more exciting and fun compared to just a shipyard without the academy bonus. :S

 

[HFY]

Then how would that impact doomstacks? Wouldn't it be easier to keep your supply ships all in range of your doomstack then?

The point of a supply mechanic is to give you something you must defend to keep your doomstack at top performance.

If you keep your supply in range of your doomstack, that's intended, but it means you must split your fleets to defend your supply, or you risk the enemy destroying your supply.

Splitting your fleets means NOT putting all your fleet into a single doomstack.

That is how Supply impacts doomstacks.

We should not aim for the "simplest solution" or "easier" one but the best solution.

Sure, and putting Supply in a Starbase is worse than the other two ideas which have been posted (kilostructures / ships / both kilostructures and ships).

Mobile supply bases sound interesting, but I do not see them as a solution to the thread goal here.

They are ships you must defend to keep your doomstack at top performance.

That means splitting your fleets up, which is the goal.

I'm not saying that supply depot building and deciding where to put them can't result in something similar, I just find it more engaging to make strategic decisions. For example, deciding if I want offensive supply depots that can be lost and turned against me if my offensive fails, or putting them back in my empire to have a 3-4 system buffer before it gets hairy, forcing my enemy to build up a frontline supply that I can utilize if their offensive fails. This is just one example of what static supply depots and the tradeoffs for stations could result in. From my perspective, engaging and fun gameplay writes itself if you really care about strategy, warfare logistics, and the implications and tradeoffs that result from them.

The problem here is that there is no such thing as a fully defensive war in Stellaris.

The best way to get an enemy to back down is to capture their starbases and invade their colonies -- exactly like you'd do in an aggressive war, except you don't get to keep them afterwards.

This is a fundamental issue in Stellaris, that most wars look a lot like Total War, except you don't get as much from winning (nor get it as fast).

But IMHO this means we can't really make a strategic decision about playing defense. That decision must be tactical -- where I have my supply on day 0 of the war -- because to win a defensive war, I often must invade the enemy after smashing their doomstack(s).

 

[Ikael]

Sure, and putting Supply in a Starbase is worse than the other two ideas which have been posted (kilostructures / ships / both kilostructures and ships).

Why would that be? I mean, aside from technical issues (I did not know that the starbase code was that hard to modify :S). As long as you can obliterate starbases by default in regular wars, it should be fine I guess?

They are ships you must defend to keep your doomstack at top performance.

That means splitting your fleets up, which is the goal.

I am not sold on mobile supply ships at all. You would just need to merge them into the doom stack in order to go back to having a single target to attack during a war (your cluster of ships + supply ships). Having squishy, static supply depots to defend makes for a much, much more interesting target and incentive for fleet-splitting.

The obvious exception to this would be the Juggernaut, which makes total sense for it to act like a mobile supply point (not to mention that is a late-game tech, and limited to one per empire, or so I remember).

 

[HFY]

Why would that be? I mean, aside from technical issues (I did not know that the starbase code was that hard to modify :S). As long as you can obliterate starbases by default in regular wars, it should be fine I guess?

- Technical issues.
- Worse for integrating with relates mechanics -- e.g. if there's a fleet that can destroy Supply specifically, or a fleet that can capture Supply without taking ownership of a system, then ... how are either of those supposed work?
- Worse fit for other related mechanics like Juggernaut being a supply source.
- Inflexible.

I am not sold on mobile supply ships at all. You would just need to merge them into the doom stack in order to go back to having a single target to attack during a war (your cluster of ships + supply ships). Having squishy, static supply depots to defend makes for a much, much more interesting target and incentive for fleet-splitting.

Then make the basic model only work in your space, or in allied space for joint wars. We already have mechanics around ships behaving differently outside / inside your space, so this isn't added complexity.

Juggernaut is an exception. Perhaps Titans are also exceptions.

That's a valid observation of a potential problem, but the problem has a solution.

 

[Imp0815]

- Technical issues.
- Worse for integrating with relates mechanics -- e.g. if there's a fleet that can destroy Supply specifically, or a fleet that can capture Supply without taking ownership of a system, then ... how are either of those supposed work?
- Worse fit for other related mechanics like Juggernaut being a supply source.
- Inflexible.

I'm still skeptical if you are the authority on deciding if something is technically difficult or not.

I'm not sure if there is a misunderstanding, but let me try to explain how I would do it: Supply is provided by supply modules on a station. Depending on the amount, you get increased range and capacity. A fleet needs supply to function without debuffs. Each fleet has a supply use per time unit that needs to be fulfilled. If too many fleets are within the range of a supply depot, taking up too much supply, all fleets in the radius will start to suffer.

So mechanically, players need to decide where to place their fleets and supply depots to create a military presence in a radius around a supply depot/station. If the supply depot stops providing supply, either by being captured or otherwise, the fleets are on a timer to find systems where they can be supplied.

This would create a natural system where stacking all fleets into one system would lead to too much supply usage in one location, and should result in your empire not being properly defensible. Nor would it be possible to conquer large amounts of territory in a short amount of time, as you would leave your maxed-out supply radius and thus stack up penalties. It's not rocket science—it's basically the system we have for HoI IV and its supply systems.

If we really have technical difficulty making captured stations work during a war, then just let them be disabled, and the war ends as soon as you try to push the offensive beyond your supply range. This would lead to a better system for any empire on the offensive or defensive, as wars would be slowed down and conquest or loss would be staged based on how well you placed your supply depots.

I would not create any mobile supply depots because it creates a non-solution that just adds annoyances to the mix, ultimately resulting in the same minute-to-minute war gameplay—stacking all forces into one stack and rolling up your enemies.

The same goes for juggernauts. I would not make them a mobile supply base, but rather make them not dependent on supply, turning them into an unpredictable threat that needs to be taken seriously, and giving them a role far enough down the line to be useful as an endgame element.

I'm not fully aware what you mean by inflexibility when we are playing a grand strategy game. Strategies are meant to be somewhat inflexible, as their entire purpose is to build a long-term progression of activities leading to a favorable end result. And this is all based on player decisions. If we mitigate strategies by provide too much flexibility in a strategic game player decisions stop to matter and we go from strategies to tactics to reactions. Then we are not playing a strategic game anymore.

And to make it clear again, yes, this system might not be perfect, and it should not be added in a vacuum, so everything else attached to fleets, wars, and stations must be revised. And finally, this whole discussion is about preventing doomstacking via hard/soft caps. So any compromise that can be predicted to lead to doomstacks in the end is just wasted energy, as we are really trying to discuss solutions that prevent that from happening—ideally, forever.

 

[HFY]

I'm still skeptical if you are the authority on deciding if something is technically difficult or not.

Dude I provided some evidence that starbase code was complex -- we have bugs from a not-that-recent DLC which have not been fixed yet, and the behaviors are complex. You don't need to trust my """authority""" about this if you are familiar with the game's mechanics.

There's really no need to go so personal or so hostile.

I'm not fully aware what you mean by inflexibility when we are playing a grand strategy game.

I mean that in mechanical terms, capturing a Starbase means something. When I declare Total War on you and take your stations, does that automatically give me Supply right in the heart of your empire? With no effort on my part beyond doomstacking my way directly through your stations?

Strategies are meant to be somewhat inflexible, as their entire purpose is to build a long-term progression of activities leading to a favorable end result. And this is all based on player decisions. If we mitigate strategies by provide too much flexibility in a strategic game player decisions stop to matter and we go from strategies to tactics to reactions. Then we are not playing a strategic game anymore.

Yeah I think that rolling a doomstack through your stations and getting free Supply for being so smart as to DO THE MOST OBVIOUS THING POSSIBLE might not be a great example of the nuances of a grand strategy.

Capturing Supply should be a separate, distinct strategy from just bulldozing through in the most direct & obvious way possible.

Lumping the two things together creates an inflexible situation with respect to mechanics.


Better design is to make Supply separate from Starbases, so capturing one or the other can be a strategic and/or tactical choice.

 

[mrt1212]

Why not redefine what it means to capture a starbase then? It's not like this is set in stone, and capturing a supply hub should be somewhat important.

For example, we could increase the repair time until a starbase is rebuilt and operational again, and have a system where a supply hub needs to be in range of another supply hub to function. This would encourage smart positioning of supply hubs around your empire to prevent enemy empires from gaining extended supply range in your space, while maintaining a stable supply range for your fleets.

In addition, you could create a scorched earth mechanic where you scuttle your station instead of losing it.

Building and maintaining supply hubs could also be made a costly and lengthy affair, making it more feasible to peace out of a war when your supply range is exhausted instead of waiting for a new forward supply base.

Additionally, supply as a mechanic could be extended and perceived differently for different empires, resulting in different military playstyles. For example, a raiding empire could receive supply from raiding planets at the cost of further destroying the planet, thus reducing its value for conquest in exchange for short-term gains.

P.S. Kilostructures would be a good compromise as it somewhat decouples the supply system from the rest, making it easier to understand. But I like the added depth through the decision of "do I want my starbase to be a frontline fortress, or a supply hub to support my fleets?" More decisions make a game a better experience. I vaguely remember there are major game design outlets that define game fun in terms of decisions per minute or something like that, and games become fun when we can make decisions. A decoupled system would take that down a peg.

EDIT: Found it: Sid Meier's Interesting Decisions

The best part of that talk to me is the section on assessing audience (and audience cohorts) and what they're looking for in a game and what their feedback is about (and I have watched this talk a dozen time before, so yes, I have a favorite part - Mr. Designer "Why aren't you making the game I wish you would make" and a fundamental disrespect of what the game devs want to make and present). This is half of what I'm talking about these days with suggestions for improvement where players are generally speaking their POV and engagement entirely and not considering they might be 20% of the whole player pie. Not insignificant but not the whole picture, so try to consider that. Of course I'm not always good about it when proposing my own ideas and do think there's a little bit too much coddling going on, but I can spot an idea that is almost archetypical and who it screws when someone else proposes.

 

[PK_AZ]

For me, discussion about taking starbases is another proof that supply radius is not enough - there has to be literal supply, that is produced in one system, stored in another, and used for fleet actions.

 

[Shaeliss]

Ship morale seems like a very useful addition.

Militarists could get higher morale overall; Pacifists could get higher morale when defending; Psionic empires could inflict higher morale damage; reduced supply could reduce morale before reducing other stats; etc.

Some weapons could damage morale more than others, and some ships should be immune to morale (e.g. Machine Intelligence Gestalts).

Bio-Ships could treat hull weapons as direct morale attacks, too -- weapons that destroy your flesh directly might be scary.

Maybe Gestalts should have a "Cohesion" instead of "Morale" with some events and/or weapon hits causing temporary desync?

And cohesion would be easy to bring down but also easy to bring back up? Short terms ups and down, more reliable over time?

 

[mrt1212]

For me, discussion about taking starbases is another proof that supply radius is not enough - there has to be literal supply, that is produced in one system, stored in another, and used for fleet actions.

Supply radius just sucks for all the potential situations and allowances in the game, where something like Crisis Aspirant before your own wormhole access and widespread hyper relays becomes a 'only a handful of the AI buffoons can do anything about this by proximity' and it's stopping the end of the game.

 

[TSBasilisk]

Supply radius just sucks for all the potential situations and allowances in the game, where something like Crisis Aspirant before wormhole and widespread hyper relays becomes a 'only a handful of the AI buffoons can do anything about this by proximity' and it's stopping the end of the game.

Capture a system and build a supply depot, or whatever, there to give you access deeper into enemy territory. Also, allow friendly nations to share supplies when working together against a common enemy - or let a neutral party provide them for a charge or on the sly.

 

[mrt1212]

Capture a system and build a supply depot, or whatever, there to give you access deeper into enemy territory. Also, allow friendly nations to share supplies when working together against a common enemy - or let a neutral party provide them for a charge or on the sly.

What if I am precluded from capturing a far flung outpost because of supply range in the first place? I'm just tossing this out here based on how I don't have a lot of contiguous territory even as a conqueror who partially wins through conquering. I get a supply radius where its like 'ships in the radius don't exhaust supply' but the scenarios where it's like, you might need a very good reason to be out of it for a long time, Im worried and I don't particularly have beef with far flung things on the other side of the galaxy you need to help with (or be the one to lead the charge against). Paying other empires for supply would be hella cool though, but then we're at a place where the impetus and primary function of supply radius is basically...a yoke on specific empires who can't make friends to rely on for supply. Which is fine, I guess.

 

[SaberVS7]

+1 to "Supply Depots", but an input of mine

I think an Empire's owned systems should automatically count as "Depots" so long as they're held, regardless of being Starbased or not - But captured outposts/bases do not, and instead building a Forward Supply Base becomes a role of Construction Ships, like how Science Ships have the role of investigating battle debris for techpoints.

Furthermore, automatically deconstructed when the war ends.