The Game Theory of Fleet Composition

[Currywurst44]

Currently we have a meta where it is best to have a fleet that is entirely composed out of battleships supported by a small fleet of corvettes.
The ideal balance many people want is to have mixed fleets that utilize all 4 ship types.

First lets look at the possibilities if there were less ship types in the game:
2 ships
On the battlefield one has to be better than the other. The optimal fleet only consists of the better ship type (factoring in cost, etc.). They can be balanced externally by having different requirements but optimally you always want to build only one of them.

3 ships
If you have three ship types the relations can be more interesting. The first possibility is that one class beats the other and that a third is stronger than both of them. In this case you can just build the third one.
The more interesting option is for their relation to be like rock paper scissors. In this case you cant just build one ship type because then your fleet can just be countered. The optimal fleet consists of all three ship types. The ratio depends on how strongly they counter each other.

4 ships
This is where you run into a problem. You cant place the four types in a circle like you did with three. To make it work the ships on opposite sites would have to be perfectly balanced so they are equally strong when fighting each other. Practically this is impossible so you always end up with something different.
If you are lucky you get rock paper scissor, or if not you get the system we currently have in Stellaris of just one best ship type.
If you want balance we have to limit ourselves to 3 ship types.

The 3 ship system would be extremely easy to implement. For example if cost, size and DPS are all equal then corvettes with 90 evasion and 1 health, destroyers with 90 tracking and 2 health and battleships with 0 tracking and 4 health would be perfectly balanced.
A optimal fleet then would be about one third Corvettes, one third Destroyers and one third Battleships.

You of course dont have to use evasion and tracking to create this kind of system. You can use any other pair of modifiers like-Penetration-Armour, Range-Speed, Missles-PointDefence etc. You just need a single pair. I just think evasion is the most natural for different ship classes.
Currently there is no balance between ship types despite having so many mechanics to achieve it. I think that the ship designer is a little too bloated. Most options are really inconsequential and only give an illusion of diversity. Thats not necessarily bad but can be very misleading.

Of course ship combat doesnt exist in a vacuum. There is the problem that you will get hard countered if you dont have all ship types available to you. This also is one of the reasons why the ship types are so similar in the current game.
To solve this you add back the ship type number 4 we removed earlier. Something like 1.5 health, 50 Evasion and 10 Tracking will lose against any other class without getting hardcountered by either one. This should be the first unlocked ship type so even if your tech is worse you can always fall back to this class to have a fighting chance.
My suggestion would be to use corvettes, destroyers and battleships for rock paper scissors and use cruisers as your basic starting ship. (Or swap Destroyers and Cruisers with this if you prefer.)

Rock paper scissors may seem like a boring and rigid game but if you factor in the external influences like research it becomes very interesting. For example:
Tier 0: Basic ship (Cruisers)
Tier 1: Corvettes
Tier 2: Destroyers
Tier 3 Battleships

Tier 0 vs Tier 1
Only basic ships vs only Corvettes

Tier 1 vs Tier 1
Only corvettes vs only corvettes

Tier 1 vs Tier 2
Only basic ships vs only Destroyers

Tier 2 vs Tier 2
Only destroyers vs only destroyers

Tier 1 vs Tier 3
Only basic ships vs only Destroyers

Tier 2 vs Tier 3
2 possible strategies depending on how strong the basic ships are:
-Only basic ships vs 2/3 destroyers 1/3 battleships
-1/3 corvettes 2/3 destroyers vs 2/3 destroyers 1/3 battleships

Tier 3 vs Tier 3
1/3 Corvettes 1/3 Destroyers 1/3 Battleships each

There are lots of interesting dynamics going on.
You dont always want to build the newest ship type.
You dont want to always build battleships. Battleships are now specialised tools only to be used against other advanced empires.
Destroyers are good general purpose ships that stay useful in every situation and during the entire game.
Except for the early game Corvettes cant effectively operate on their own and need support from other ships.

The system is very robust. As long as each ship type heavily counters another one there will always be this dynamic. Only the ratios shift around but the general strategy stays the same. This means that it is no problem to have technologies that change stats around and improve one ship class as long as they only improve the ship in its role. So dont give battleships small weapons with tracking or destroyers too much evasion and everything will stay well. If the counters are strong enough you can tack on all the other systems you want like different speeds, different disengagement chances, etc.

What do you think about implementing something like this?

 

[Olterin]

A solid analysis, however it does not consider out-of-combat ship parameters, most notably strategic mobility. Now, it's fair to say that it currently doesn't really matter much, but it could - if a fleet of cruisers, while getting outclassed by a battleship armada, can outmaneuver it and accomplish strategic goals, then the battleships suddenly require the support of smaller, more mobile means of pinning the cruisers down to get them into a fight. This dynamic currently does not really exist, but would be very interesting to have - it might require some specialist modules or weapons to be introduced that currently do not exist, or something else.

 

[Tacticus101]

I think this sort of system functions best in an RTS, where you can micromanage units and where new units are built quickly. There it offers interesting tactical combat as you try to maneuver your units to fight the ones they are good against whilst also avoiding their counters, and can rapidly adapt your forces if necessary depending on losses or different enemies. It is why a lot of RTS's stick with variations and more complex versions of the Spear>Cavalry>Archer>Spear balance.

Stellaris does not have that, there is no tactical combat and ships take a long time to build, which means that it adds no tactical options and fleet composition will frequently be defined more by your losses than how you intended to set it up. I dont agree it creates a lot of interesting dynamics; your fleets dont adapt or have different roles, you just build the optimal fleet for your tech level. For most levels each new ship just counters all the previous ones; Corvettes counter basic ships, Destroyers counter Corvettes and basic ships, why would you not just build that one newest ship type? Only when you unlock battleships is there an actual circle of counters involved that would make you balance your fleets out.

In addition, the system is only robust because it is simple. A lot of people playing Stellaris enjoy the ship design, even if a lot of options are superfluous. People like being able to design ships to do things even if that thing isnt the optimal path. In order to make this new system work you would have to remove a lot of those choices; Strike Craft would have to go because they allow battleships to beat Corvettes, Missiles would be impossible to balance, Lighting Weapons would go, most of the upgrades would be removed because they add things like tracking or evasion and the Armour-Shield-Hull divisions also seems to have gone in your system. You are adding robustness at the expense of depth. Plus, things like Titans, Spaceborn Entities, Starbases and various ship bonues are in the game and wouldnt fit this system.

So it is an interesting idea, but i feel that Stellaris fleet combat has to be balanced using strategic implications rather than tactical ones, ie using supply lines and ships having strategic roles to give each a place.

 

[InvisibleBison]

The problem with this analysis is that it attributes success in combat purely to ship size and ignores the effects of weapon composition. A destroyer with an L kinetic weapon and 2 PD weapons is not the same as a destroyer with 4 S lasers and a PD weapon, and neither of those are the same as a destroyer with an M energy and kinetic weapon and an S energy and kinetic weapon.

 

[Farbros]

I don't know, I usually make fleets with 1-2 titans, +20 battleships X+L and few with X + hangars and PD. Battleships have always 1 truster, so that titans would never be in front of them. I use corvette only as a fast reaction group, if the enemy or pirates hit somewhere where my fleet is too far away.

 

[Currywurst44]

I think this sort of system functions best in an RTS, where you can micromanage units and where new units are built quickly. There it offers interesting tactical combat as you try to maneuver your units to fight the ones they are good against whilst also avoiding their counters, and can rapidly adapt your forces if necessary depending on losses or different enemies. It is why a lot of RTS's stick with variations and more complex versions of the Spear>Cavalry>Archer>Spear balance.

Stellaris does not have that, there is no tactical combat and ships take a long time to build, which means that it adds no tactical options and fleet composition will frequently be defined more by your losses than how you intended to set it up. I dont agree it creates a lot of interesting dynamics; your fleets dont adapt or have different roles, you just build the optimal fleet for your tech level. For most levels each new ship just counters all the previous ones; Corvettes counter basic ships, Destroyers counter Corvettes and basic ships, why would you not just build that one newest ship type? Only when you unlock battleships is there an actual circle of counters involved that would make you balance your fleets out.

There are some choices from an empire point of view. If you are ahead in tech you are right, you generally want to build the newest ship type. If you want research to have an impact it has to be this way. (Except when fighting with Tier3 against Tier1 or if you already have over 1/3 battleships. So its not even as simple as that.) The really interesting decisions arise if you are not the most technological advanced empire and dont focus on research. You take a gamble because if you build something apart from basic ships you can get countered if you fall behind. Do you rush ahead and then focus your economy on producing corvettes for a short term gain? Do you slow your research and only build basic ships and destroyers because your corvettes would already get countered when you unlock them? Dont focus on researching ships at all and use the basic ships for most of the game untill you unlock all classes?

In addition, the system is only robust because it is simple. A lot of people playing Stellaris enjoy the ship design, even if a lot of options are superfluous. People like being able to design ships to do things even if that thing isnt the optimal path. In order to make this new system work you would have to remove a lot of those choices; Strike Craft would have to go because they allow battleships to beat Corvettes, Missiles would be impossible to balance, Lighting Weapons would go, most of the upgrades would be removed because they add things like tracking or evasion and the Armour-Shield-Hull divisions also seems to have gone in your system. You are adding robustness at the expense of depth. Plus, things like Titans, Spaceborn Entities, Starbases and various ship bonues are in the game and wouldnt fit this system.

You can keep a lot of things if they stay inconsequential. A low amount of tracking or evasion wont topple the system. Armour-Shield-Hull can stay because it doesnt do much. The same with bonus damage for weapon types because it is independent from all of this. Limiting weapon types to certain classes (because they are too heavy or too risky on large ships or something like that) makes sure that you can in either case continue to use them.
Anyways, I would argue most of these things just add complexity but no depth but that is another issue. They are nice to have for roleplay.
Not everything has to fit with the system, only the basic ships. Titans, spaceborn entities and starbases are all very limited so they wont be a problem and only make the game more interesting. You have to rework some bonuses but apart from tracking you dont have to pay too much attention.
You pick one aspect like evasion and increase its impact until it becomes the defining characteristic of ships types. Clearly communicate that ships have a certain role and you can build anything you want on top of that.

So it is an interesting idea, but i feel that Stellaris fleet combat has to be balanced using strategic implications rather than tactical ones, ie using supply lines and ships having strategic roles to give each a place.

A solid analysis, however it does not consider out-of-combat ship parameters, most notably strategic mobility. Now, it's fair to say that it currently doesn't really matter much, but it could - if a fleet of cruisers, while getting outclassed by a battleship armada, can outmaneuver it and accomplish strategic goals, then the battleships suddenly require the support of smaller, more mobile means of pinning the cruisers down to get them into a fight. This dynamic currently does not really exist, but would be very interesting to have - it might require some specialist modules or weapons to be introduced that currently do not exist, or something else.

That would be a better solution but I dont believe that anything more complicated can be made to work. The devs already tried a few times but they always failed.

Edit:

The problem with this analysis is that it attributes success in combat purely to ship size and ignores the effects of weapon composition. A destroyer with an L kinetic weapon and 2 PD weapons is not the same as a destroyer with 4 S lasers and a PD weapon, and neither of those are the same as a destroyer with an M energy and kinetic weapon and an S energy and kinetic weapon.

The issue with considering armament of ships is that it can be very quickly changed so it cant really be a solution to balance different ship classes.

What you can do is allow the possibility to sacrifice some of the counter potential of a class to give it a very slight increase against its own class. So destroyers with L weapons are a bit better against other destroyers but they would loose the ability to counter corvettes. Though they still should get destroyed by battleships.

 

[Zentay]

What do you think about implementing something like this?

I think it would not make the game better. For some reason people believe that a rock-paper-scissor system would be fun and interesting but I think it never is (at least not when it's so simplistic). We need something different.

Currently you only go to war to take planets by defeating the enemy fleet, then starbases and then invading. Battleships are the best at achieving this with the least losses and therefore battleships are the best. It can be useful to have a small fast fleet that follows and takes systems so that the main fleet does not waste time with undefended systems but it's more convenience than anything. This is all the depth the game offers.

Essentially, ships have only one role in the game and with several hulls competing for this role one of them must necessarily be the best, making the others nearly obsolete. A rock-paper-scissor system totally misses the point and in a sense we already have this with the different weapon types, and the choice of shield vs armor. That system is underwhelming isn't it?

 

[MathyM]

The problem with this analysis is that it attributes success in combat purely to ship size and ignores the effects of weapon composition. A destroyer with an L kinetic weapon and 2 PD weapons is not the same as a destroyer with 4 S lasers and a PD weapon, and neither of those are the same as a destroyer with an M energy and kinetic weapon and an S energy and kinetic weapon.

According to Montu’s analyses, this is closer to the ideal composition. You can drop the PD, though.

If I recall correctly, you get one titan, 80% Artillery BBs, and 20% Carrier BS. You don’t need corvettes mixed into the BB fleet, they are doing little more than wasting fleet capacity. You can still make a dedicated corvette fleet for screening, though.

 

[Currywurst44]

I think it would not make the game better. For some reason people believe that a rock-paper-scissor system would be fun and interesting but I think it never is (at least not when it's so simplistic). We need something different.

Currently you only go to war to take planets by defeating the enemy fleet, then starbases and then invading. Battleships are the best at achieving this with the least losses and therefore battleships are the best. It can be useful to have a small fast fleet that follows and takes systems so that the main fleet does not waste time with undefended systems but it's more convenience than anything. This is all the depth the game offers.

Essentially, ships have only one role in the game and with several hulls competing for this role one of them must necessarily be the best, making the others nearly obsolete. A rock-paper-scissor system totally misses the point and in a sense we already have this with the different weapon types, and the choice of shield vs armor. That system is underwhelming isn't it?

The thing that would change is defeating the enemy fleet. A research based rock paper scissors system would create a lot of interesting interactions in that area at least.
The reason I am advocating for this kind of system is that you can make it consequential. I showed that ship to ship combat can be non-transitive so that there is no best hull when it comes to fleet combat. The equipment of ships can just be changed very easily but the ships classes you build can not.

I agree that other areas of warfare can still be greatly improved.
It is hard to decide if the amount of work for something like this is worth it but seeing how often people discussed this aspect in the past it surely has some importance.
Something that solves multiple issues at once instead of just the fleet composition would be great but can be very hard or costly. I think if you have a solid base system it becomes a lot easier to continue to work from there.

 

[baraterek]

I'm beginner, so I'll write something

What about making a system, where ships needs each other? Like Battleship and Colossus should have "weak spots" so that there is % chance that small ships will make critical damage to them (blow), then such ships need escort (Cruiser). From other side, Big Ships have strong fields, so Corvette operates badly if close to it - it's better to use Destroyers to destroy Big Ships (weak spot exploit) etc.

I hope it's understandable. Instead of just numbers, lets add RNG based on incidents. Would be nice to see, that one exploding Battleship could damage nearby ships, or see that one "hero" Corvette destroys Colossus (minimal chance of course, but it's so Star Wars fun level) or while torpedoing starbase counting, that strikecraft and Corvette won't break defence and reach heavy guns ...

 

[Losttruppen]

I don't like the idea of a rock paper scissors change to stellaris because you could not go to war with an enemy if you did not have intel on their ship builds. This would either lead to you and the AI declaring war as you do now and then finding out you already lost before your first engagement, or you see the enemy fleet and have to go back and refit all your ships which then leads to you just stomping on the AI with a perfect counter fleet as they aren't going to go back refit their ships.

The "Ideal" fleet composition in this new system would be to build an even distribution of rock, paper, and scissor ships and we would be right back where we started with the decision coming down to numbers and tech levels. It might solve the variety issue but it won't change the gameplay much other than a higher repair bill from losses.

 

[Northernwwater]

Great post, but I would like to see new ways to add variety outside of new weapon mixes.

For example, mutually exclusive build choices that for example increased survivability at the expense of damage output. So, for example, if you use an X slot you have less armor, hull or shield affordability. All we really have now is a small amount of power trade off. To small in my opinion.

Traditions trees that emphasized survivability verses damage output.

 

[fourteenfour]

For me the most efficient means to make it desirable to have each hull type represented is to make it inefficient to mount weapon types on specific hulls.

Examples. Please understand the numbers provided in modules/weapons are very rough and without detail as in exact counts of types when using mixed weapons in one section

Corvettes should be the choice for harassment and such, never mounting more than beam weapons or trading them for a single slow loading missile or torpedo weapon. I would actually separate missile and torpedo launchers into different component groups leaving corvettes only the torpedo option.

Destroyers we could focus on point defense so that their module slots three point defense weapons at a time and no other weapons. They would have that option for both sections they could equip. Perhaps their front section could mount four. They would never have more than medium mounts.

Cruisers we could focus on hangars and missiles. So each section would accommodate more either than any other hull. As in as a CV a cruiser could be 2+3+1 with other configurations being similar for missile mounts of 2+4+2 or with torpedoes as 1+2+1. They would also be able to sport large mounts for beam weapons as 1+2+1. However their mounting of point defense weapons would be 2 + 2 + 2 meaning any DD would be a cheaper solution.

Battleships would favor large mount weapons and even the "X" mount but I would swap the X mount to be a singular weapon mount occupying the center slot to represent that sacrifice of three large slots to mount it. They would have similar point defense issues in that at most they would have 2 + 2 + 2 and each would replace the space required for whatever other weapon type could fit. A battlesip as a carrier would be 2 + 2 + 1

Then we would change combat rules such that all weapons start not ready to fire and only start their cool down timer once a ship has successfully acquired its first target.


One other change and more controversial would be to change how star bases are handled.

When taking control of a base all modules are stripped. Simply put you don't know how to operate what the other guy built as they may not have the same number of limbs, digits, vision , and more. The second is in repair. Destroyers and down could be repaired at any starbase that has crew service facilities. Cruisers would need a Starhold sized facility to repair and Battleships would need Star Fortress or shipyard facilities to repair and would take a build slot while repairing. Titans require both Star Fortress and multiple shipyards.

Then we toss in supply. We can base supply on the level of improvement a starbase has or perhaps components, example would be anchorages might provide the range and ship class supply needs. Example, a Starport which is the first level of improvement with at minimum of crew quarters can hold any number of destroyers or smaller to a range of five jumps. Anchorage modules extend this by one. Starholds support cruisers and can support battleships provided they have at least one anchorage. Outside of support means you fleet costs are treated as undocked.


TL;DR The key is to limit what each hull type can mount efficiently so that if you want that weapon type you are best served by a specific hull type.

 

[Tacticus101]

There are some choices from an empire point of view. If you are ahead in tech you are right, you generally want to build the newest ship type. If you want research to have an impact it has to be this way. (Except when fighting with Tier3 against Tier1 or if you already have over 1/3 battleships. So its not even as simple as that.) The really interesting decisions arise if you are not the most technological advanced empire and dont focus on research. You take a gamble because if you build something apart from basic ships you can get countered if you fall behind. Do you rush ahead and then focus your economy on producing corvettes for a short term gain? Do you slow your research and only build basic ships and destroyers because your corvettes would already get countered when you unlock them? Dont focus on researching ships at all and use the basic ships for most of the game untill you unlock all classes?

Those are false choices though.
If you dont build anything other than basic ships then you lose anyway, because basic ships are weak against every kind of ship AND you are behind on tech so your ships are all worse anyway.
How would you need to focus your economy any more than your currently do when you unlock a new ship type?
Why would you willingly be behind on tech given the bonuses it gives across the board?
Unless you are completely reworking the tech system, how do you specifically reduce your research on ships?

The end result of these changes is that the 'optimal' fleet composition changes. You end up with rather than all Battleships (or whatever the currently meta is) being optimal, you want fleets of exactly 1/3 Corvettes/Destroyers/Battleships. That isnt a decision, you are just doing a different same thing every game.

You can keep a lot of things if they stay inconsequential. A low amount of tracking or evasion wont topple the system. Armour-Shield-Hull can stay because it doesnt do much. The same with bonus damage for weapon types because it is independent from all of this. Limiting weapon types to certain classes (because they are too heavy or too risky on large ships or something like that) makes sure that you can in either case continue to use them.
Anyways, I would argue most of these things just add complexity but no depth but that is another issue. They are nice to have for roleplay.
Not everything has to fit with the system, only the basic ships. Titans, spaceborn entities and starbases are all very limited so they wont be a problem and only make the game more interesting. You have to rework some bonuses but apart from tracking you dont have to pay too much attention.
You pick one aspect like evasion and increase its impact until it becomes the defining characteristic of ships types. Clearly communicate that ships have a certain role and you can build anything you want on top of that.

Numbers are hard to balance though.
Add a few too many bonuses to evasion from Psionic/Engines/etc and suddenly destroyers might have enough to be viable vs Battleships.
Leave weapons the way they are and you get particle launcher destroyers beating battleships.
How do you keep starbases viable at defending against all types of ships?

You have a force a lot of things for balance, and i think that makes things uninteresting if you are going to rigidly stick to a rock-paper-scissors system. Most (good) RTSs do break their own rules to keep things interesting. Cavalry that can kill spear units or Archers that counter cavalry and similar units. They recognise that too simple a combat is boring, and those are in games where the micromanagement makes battles way more interesting anyway.

That would be a better solution but I dont believe that anything more complicated can be made to work. The devs already tried a few times but they always failed.

The devs havent really tried to add an actual strategic role system. What they did try to do was force a more tactical one similar to what you suggested, with destroyers being the main source of point defence and tracking.

What they need is a logistics/fuel system similar to how HOI4 balances ships. Battleships are big and powerful but cost a lot when out of port, Corvettes are cheap but have limited range.

 

[Currywurst44]

I don't like the idea of a rock paper scissors change to stellaris because you could not go to war with an enemy if you did not have intel on their ship builds. This would either lead to you and the AI declaring war as you do now and then finding out you already lost before your first engagement, or you see the enemy fleet and have to go back and refit all your ships which then leads to you just stomping on the AI with a perfect counter fleet as they aren't going to go back refit their ships.

In my post I was talking about specific roles for the different ship classes like Destroyer, Cruiser,... You cant refit those to each other. Thats why it is so important to include a ship type that doesnt get hardcountered by other ones. It will just be a bit more expensive.

Those are false choices though.
If you dont build anything other than basic ships then you lose anyway, because basic ships are weak against every kind of ship AND you are behind on tech so your ships are all worse anyway.

The basic ships in my examples are not significantly worse just enough to reward research. Its not a false choice. Sometimes you have to chose between losing and less losing, especially if you are an AI.

The "Ideal" fleet composition in this new system would be to build an even distribution of rock, paper, and scissor ships and we would be right back where we started with the decision coming down to numbers and tech levels. It might solve the variety issue but it won't change the gameplay much other than a higher repair bill from losses.

The end result of these changes is that the 'optimal' fleet composition changes. You end up with rather than all Battleships (or whatever the currently meta is) being optimal, you want fleets of exactly 1/3 Corvettes/Destroyers/Battleships. That isnt a decision, you are just doing a different same thing every game.

Thats one of the main points of wanting mixed fleets. With mono fleets there is no flexibility. The only possible change is to switch to another ship class altogether if there was suddenly some extreme bonus for a different one.
The beauty of a mixed fleet is that even if you research a small improvement for one of your classes the optimal ratios immediately shift a little the whole game long. All of the sudden everything becomes flexible. So for one empire it could be 1/4, 1/4, 2/4 and for another one 2/5, 1/5, 2/5. It wont be a big loss to differ from this "optimum" but if you only research the improvements for one class you know what you should do.

How would you need to focus your economy any more than your currently do when you unlock a new ship type?
Why would you willingly be behind on tech given the bonuses it gives across the board?
Unless you are completely reworking the tech system, how do you specifically reduce your research on ships?

I would switch to militarized economy and turn a few scientists into metalurgists and stop expanding. Sometimes you have different priorities, for example if you want to capture an opponents homeworld or a chokepoint. It will be worth the sacrifice in the short or medium term. Just dont research the techs to unlock/improve the ship class. But you are right, you cant really to that to any extend.

Numbers are hard to balance though.
Add a few too many bonuses to evasion from Psionic/Engines/etc and suddenly destroyers might have enough to be viable vs Battleships.

Only having to pay attention to evasion and tracking would certainly help with that. If there are very specific lategame setups for a portion of empires it might not even be that bad and rather interesting.

Leave weapons the way they are and you get particle launcher destroyers beating battleships.

No, both can use shields and battleships will just have more DPSHealth than destroyers. They would be able to use the particle launchers just as well as neither of them has any significant evasion.

How do you keep starbases viable at defending against all types of ships?

Personally I would make them similar to destroyers in stats. So no evasion and high tracking with medium DPSHealth. Then gradually over the course of the game they get worse at defending against fleets as you research battleships.

You have a force a lot of things for balance, and i think that makes things uninteresting if you are going to rigidly stick to a rock-paper-scissors system. Most (good) RTSs do break their own rules to keep things interesting. Cavalry that can kill spear units or Archers that counter cavalry and similar units. They recognise that too simple a combat is boring, and those are in games where the micromanagement makes battles way more interesting anyway.

The thing is that it wont be worse than it is now. To break away from rock paper scissors you first have to have rock paper scissors and the base mechanic behind that should be rock solid. You can always add other ship classes later that violates it but is limited by certain resources or something else.
Many RTS have multiple overlapping ones of these circles of counters. I dont think you can theoretically do that with just 4 units. You probably need at least 5 or 6. For example with AoE2 you have Cavalry-Spear-Bow but you also have Sword-Bow-Skirmisher.

 

[TrotBot]

Currently we have a meta where it is best to have a fleet that is entirely composed out of battleships supported by a small fleet of corvettes.
The ideal balance many people want is to have mixed fleets that utilize all 4 ship types.

First lets look at the possibilities if there were less ship types in the game:
2 ships
On the battlefield one has to be better than the other. The optimal fleet only consists of the better ship type (factoring in cost, etc.). They can be balanced externally by having different requirements but optimally you always want to build only one of them.
View attachment 783265
3 ships
If you have three ship types the relations can be more interesting. The first possibility is that one class beats the other and that a third is stronger than both of them. In this case you can just build the third one.
The more interesting option is for their relation to be like rock paper scissors. In this case you cant just build one ship type because then your fleet can just be countered. The optimal fleet consists of all three ship types. The ratio depends on how strongly they counter each other.
View attachment 783266 View attachment 783267
4 ships
This is where you run into a problem. You cant place the four types in a circle like you did with three. To make it work the ships on opposite sites would have to be perfectly balanced so they are equally strong when fighting each other. Practically this is impossible so you always end up with something different.
If you are lucky you get rock paper scissor, or if not you get the system we currently have in Stellaris of just one best ship type.
If you want balance we have to limit ourselves to 3 ship types.
View attachment 783268

The 3 ship system would be extremely easy to implement. For example if cost, size and DPS are all equal then corvettes with 90 evasion and 1 health, destroyers with 90 tracking and 2 health and battleships with 0 tracking and 4 health would be perfectly balanced.
A optimal fleet then would be about one third Corvettes, one third Destroyers and one third Battleships.

You of course dont have to use evasion and tracking to create this kind of system. You can use any other pair of modifiers like-Penetration-Armour, Range-Speed, Missles-PointDefence etc. You just need a single pair. I just think evasion is the most natural for different ship classes.
Currently there is no balance between ship types despite having so many mechanics to achieve it. I think that the ship designer is a little too bloated. Most options are really inconsequential and only give an illusion of diversity. Thats not necessarily bad but can be very misleading.

Of course ship combat doesnt exist in a vacuum. There is the problem that you will get hard countered if you dont have all ship types available to you. This also is one of the reasons why the ship types are so similar in the current game.
To solve this you add back the ship type number 4 we removed earlier. Something like 1.5 health, 50 Evasion and 10 Tracking will lose against any other class without getting hardcountered by either one. This should be the first unlocked ship type so even if your tech is worse you can always fall back to this class to have a fighting chance.
My suggestion would be to use corvettes, destroyers and battleships for rock paper scissors and use cruisers as your basic starting ship. (Or swap Destroyers and Cruisers with this if you prefer.)
View attachment 783269
Rock paper scissors may seem like a boring and rigid game but if you factor in the external influences like research it becomes very interesting. For example:
Tier 0: Basic ship (Cruisers)
Tier 1: Corvettes
Tier 2: Destroyers
Tier 3 Battleships

Tier 0 vs Tier 1
Only basic ships vs only Corvettes

Tier 1 vs Tier 1
Only corvettes vs only corvettes

Tier 1 vs Tier 2
Only basic ships vs only Destroyers

Tier 2 vs Tier 2
Only destroyers vs only destroyers

Tier 1 vs Tier 3
Only basic ships vs only Destroyers

Tier 2 vs Tier 3
2 possible strategies depending on how strong the basic ships are:
-Only basic ships vs 2/3 destroyers 1/3 battleships
-1/3 corvettes 2/3 destroyers vs 2/3 destroyers 1/3 battleships

Tier 3 vs Tier 3
1/3 Corvettes 1/3 Destroyers 1/3 Battleships each

There are lots of interesting dynamics going on.
You dont always want to build the newest ship type.
You dont want to always build battleships. Battleships are now specialised tools only to be used against other advanced empires.
Destroyers are good general purpose ships that stay useful in every situation and during the entire game.
Except for the early game Corvettes cant effectively operate on their own and need support from other ships.

The system is very robust. As long as each ship type heavily counters another one there will always be this dynamic. Only the ratios shift around but the general strategy stays the same. This means that it is no problem to have technologies that change stats around and improve one ship class as long as they only improve the ship in its role. So dont give battleships small weapons with tracking or destroyers too much evasion and everything will stay well. If the counters are strong enough you can tack on all the other systems you want like different speeds, different disengagement chances, etc.

What do you think about implementing something like this?

i don't think we should have to build every ship type. single ship type fleets should be still possible, but they should require different designs of the same ship type. smaller ship types can be buffed enough for battleships to not be viable if you don't have some carrier battleships to screen for the capital weapon battleships. this would open up the option of running a carrier cruiser fleet alongside them as well, instead. or you could have them escorted by corvettes or destroyers as your screen. I find this much more interesting than just locking people to being forced to just build all four at all times. let different styles all be viable.

 

[Pancakelord]

Whilst I do think a full rework is needed, I dont think its too likely. We havent had any edits of existing ship sections in a long time.

At best I can see them adding a few new loadout roles to hunt certain classes. E.g. A pair of forward-locked L-slots on a cruiser bow that adds an extra 30% Anti-battleship damage. Or some sort of "Anti-[Shipclass]" Aux-mod increasing that ship's dmg against them correspondingly. It wouldn't restrict builds - it'd make them more flexible if anything - whilst still forcing you to think a little more about sending 20 battleships against someone when the whole galaxy knows you have a battleship fetish.

But (and I said this when the 3.0 DDs were being aired) we need a dossier on known enemy/other fleet designs if combat is to ever actually be anything more complex than a moshpit of +range stats.

SOTS had this years ago, it'd give you a little database of enemy designs you'd seen so you could design your stuff accordingly. Intel levels could probably restrict this in Stellaris, it just needs a decent UI - even re-using the ship designer with a "Country (known designs)" filter would work IMO.

 

[UniqueNameHere]

I actually think a lot of problems could be solved by eliminating the ability for battleships to mount X weapons - leave those entirely to Titans. Then, battleships could be the ONLY ships that could mount hangers. Your setups would then be limited to:
Corvettes: S or G (but NOT torpedo)
Destroyer: S, G (any), PD
Cruiser: S, M, G (any)
Battleships: L, G (any), H
Titans: L, X, H

 

[TrotBot]

I actually think a lot of problems could be solved by eliminating the ability for battleships to mount X weapons - leave those entirely to Titans. Then, battleships could be the ONLY ships that could mount hangers. Your setups would then be limited to:
Corvettes: S or G (but NOT torpedo)
Destroyer: S, G (any), PD
Cruiser: S, M, G (any)
Battleships: L, G (any), H
Titans: L, X, H

noooo, don't take away my carrier cruisers pleeeease

 

[Tacticus101]

The basic ships in my examples are not significantly worse just enough to reward research. Its not a false choice. Sometimes you have to chose between losing and less losing, especially if you are an AI.

My point is that there isnt a real choice. There is the optimal choice and there is the option you got for because you are behind in tech. That isnt an actual decision because there is always one optimal option rather than several equally good ones. At best it is a "do i build up now or horde my resources for when i unlock the next ship type", which already exists in the game.

Thats one of the main points of wanting mixed fleets. With mono fleets there is no flexibility. The only possible change is to switch to another ship class altogether if there was suddenly some extreme bonus for a different one.
The beauty of a mixed fleet is that even if you research a small improvement for one of your classes the optimal ratios immediately shift a little the whole game long. All of the sudden everything becomes flexible. So for one empire it could be 1/4, 1/4, 2/4 and for another one 2/5, 1/5, 2/5. It wont be a big loss to differ from this "optimum" but if you only research the improvements for one class you know what you should do.

In a perfect theater of mind though, maybe? Though i doubt you would achieve sufficient balance to make it work like that.

In a real game, most players are not going to constantly adjust the ratio of their fleet every tech or two. If nothing else it is generally economically better to just build more ships rather than disband your current ones anyway. I mean, currently in the game there are lots of points where the optimal decision isnt just battleships or corvettes, but people still generally stick to whatever the meta is. It is much simpler to just use the same fleet template rather than constantly change it unless you are RPing. The ships you build are generally determined by by economic reality rather than balance.

Also, unless new weapons or upgrades are giving like +100% levels of effectiveness you are not going to get that large a variation in "optimal" ratio. A +10% increase in firepower from say a weapon upgrade will mean you need 18 rather than 20 corvettes as the 'optimal' number for your fleet at the most extreme, that isnt a meaningful flexibility especially given the inherent degree of randomness in the combat system.

I would switch to militarized economy and turn a few scientists into metalurgists and stop expanding. Sometimes you have different priorities, for example if you want to capture an opponents homeworld or a chokepoint. It will be worth the sacrifice in the short or medium term. Just dont research the techs to unlock/improve the ship class. But you are right, you cant really to that to any extend.

All of which are things you currently exist in the game to the same extent.

Only having to pay attention to evasion and tracking would certainly help with that. If there are very specific lategame setups for a portion of empires it might not even be that bad and rather interesting.

That is the main metric they balance combat at the moment though, and it hasnt worked. Plus if say battleships have no tracking, it wouldnt take a lot of extra evasion to a destroyer to make them have more effective health than Battleships. Even getting close would make them viable against mixed fleets and anything short of a pure battleship fleet.

No, both can use shields and battleships will just have more DPSHealth than destroyers. They would be able to use the particle launchers just as well as neither of them has any significant evasion.

The point is that if someone unlocks one of the better weapons against someone behind on tech or with a slant towards armour techs they could build a destroyer that does sufficiently more damage to armour/hull that their destroyer could beat a battleship because of the extra damage it does. Not to mention you would have to completely remove the variety of high accuracy weapons like Arc Lightning.

Personally I would make them similar to destroyers in stats. So no evasion and high tracking with medium DPSHealth. Then gradually over the course of the game they get worse at defending against fleets as you research battleships.

So you dont have them viable at defending against all types of ships, making battleships better?

The thing is that it wont be worse than it is now. To break away from rock paper scissors you first have to have rock paper scissors and the base mechanic behind that should be rock solid. You can always add other ship classes later that violates it but is limited by certain resources or something else.
Many RTS have multiple overlapping ones of these circles of counters. I dont think you can theoretically do that with just 4 units. You probably need at least 5 or 6. For example with AoE2 you have Cavalry-Spear-Bow but you also have Sword-Bow-Skirmisher.

It wont be better than it is now though, and it will restrict options. Rock-Paper-Scissors systems add a lot to games with tactical combat because you can maneuver and there is a high replacement rate, i dont think they add much interesting to grand strategy games where the interesting thing is the strategic use and design of fleets.