The Large Fleet Collider - 2.0 Fleet Balance

[Fa1nan]

I can contribute a bit to that Disruptor/CL vs Neutron debate:

I've begun my quest to determine the best loadouts against FEs by pitching test fleets against event-spawned FE fleets of one Titan, 30 Battlecruiser and 60 Escorts five times for each FE type. This gives me more refined results opposed to just fighting the Doom Fleet of Doom. Also, I've edited the defines to make large enough fleets possible because there apparently is no way to nullify the command limit via console.

So far I've tested mono-BB fleets starting with the F1/5-TB-2KA2P which was best in 1.8/1.9 (let's see how the notation works out - I did replace Decoders with Auxiliary Fire Controls though; I'm trying to avoid the special tech you are not guaranteed to unlock) to get a benchmark. The very first observation with those - thanks to having split the Doomfleet - is that the 1 shield plus 5 armor setup is most definitively not optimal anymore; it gets hardcountered by Materialists and Machines. I stuck with it for comparability, though.

On the weapons side, what sticks out are GB-1KA3N (any combination of Tachyon, Giga, Kinetic Artillery and Neutron Launchers, in fact) and AB-6D1H respectively AB-4CL if one should happen to find cloud lightning. The former has the highest lethality while the latter has the highest survivability.

In more detail, the siblings of the 422 rely on high damage and low firerate weapons which causes them to quite noticeably deal damage in waves. In each wave they kill a chunk of the FE fleet with downtimes in between where damage is rather lacking. As a result, the FE ships stay in combat longer and individual ships get focused fired down, preventing emergency FTL jumps, which kills up to 90% of the FE fleet before combat ends. The downside of course is that after each volley there are healthy FE ships left which inflict high casualties on the test fleet; with the armor focused defensive setup I used these range between 20% and 50% against Spiritualists, Xenophobes and Xenophiles and up to 80% against the Materialists and Machines (which then also end with the test fleet disengaging instead of the FE fleet).

Contrast this with armor and shield ignoring weapon loadouts: They deal damage more consistently, but directly to the hull; as a result FE ships have the opportunity to emergency FTL, but are also constantly debuffed from the hull damage. They kill about 70% of the FE fleet while suffering losses of around 10% against Spiritualists, Xenphobes and Xenophiles and 30% against Materialists and Machines.

I attached the concrete results as an excel file.

Now, in the test setup the FE fleets lack the buffs they have in an actual game - and there they have considerably more armor and shields while not that much more hull. As such, I expect Disruptors respectively Cloud Lightning to perform better than the 422-esque designs in an actual game. Furthermore, due to the lower amount of losses they suffer they should generate less war attrition and they're cheaper, so for now I declare armor and shield penetrating setups to be the best choice against FE if you happen to bring mono-BB-fleets. Or in other words, to get back to the Disruptor vs Neutron debate: If the enemy has enough armor and shields compared to hull, then Disruptors will outperform Neutron.

I do suspect however that BBs are not the best hull to actually wield Disruptors into combat and am going to test other hulls for now.

Edit: Tested corvette swarms and they are... eh. Corvettes are not completely useless like in 1.9, but they cannot compete with the BBs. The corvette swarms suffer the same percentage of casualties as the FE fleets and are therefore not a good choice considering war attrition. Specifically, I tested swarms of A1/2-CV-3D, A1/2-CV-1D1T and A1/2-CV-2AC1P and they performed all rather similarly. Concrete results at a later time.

 

[Drakonn]

-L and X weapons are for killing cruisers and above. Devastator torps are also here. They might be semi useful vs destroyers depending on the setup.
-M weapons are for cruisers and destroyers or can be made useful vs corvettes with picket AI. Missiles are also here and can be a substitute for torps vs battleships if you don't have torps or battleships don't have armor
-S weapons are pretty much exclusively for hitting corvettes and the only option to have a good hit chance against them if you don't use picket AI
-Autocannons and bombers(assuming bomber accuracy does benefit from ship modules) hit pretty much always, don't use them on picket ships as it's an overkill in terms of tracking. These are mostly for Corvettes with swarm AI to kill each other or for non picket cruisers and battleships vs corvettes.

Wait. What are destroyers and Cruisers considered (size)?

L or M works for Cruisers then? (trying to make sure I'm getting right and haven't been wrong the whole time.)

Missiles/torpedos (G) aren't a good idea for (against) corvettes because of their evasion right?

Aside: Did I just misread somewhere that targeting is like versus like or is it heavily dependent on weapons/combat computer setup? I think I've gotten confused somewhere. I clearly saw my artillery destroyers doing their work against Cruisers. (I don't want to load my Corvettes with Torpedoes and have them all target other corvettes).

Edit: Just saw the previous page on Targeting. Which is..confusing given torpedos and such but maybe not since everyone can have armour. But if wanted missiles/torpedoes to target ships of a different class....hmm. Things to think about now.

 

[vonriel]

I'd like to help out with this, because I'm interested in what's going on here and have a few ideas of my own, but I don't know how to run the tests on my own. I muddled around a bit with console commands, but it's late and I'm not thinking straight, and overall don't want to waste time if there's a better way than what looks obvious. If there's a guide or reference somewhere that I've missed due to largely not paying attention to the forums, a link would be welcome, or at least something to search for. Thanks in advance.

Going forward, 2.0.2's beta currently has different evasion amounts for thrusters. The stats on the beta build are 20%/16%/12%/8%/6%(?) evasion for CV/DD/CC/BB/Titans when using the new tier 5 Dark Matter Thrusters. Each tier below them removes 25% of the final amount, or 5% for CVs, 4% for DDs, etc. This is compared to the current 15% evasion with tier 4 engines for everyone, so a strict nerf to evasion for all ship classes above Corvettes until you reverse engineer those FE techs, a slight buff to Destroyers once you get those tier 5 engines, and a strict buff to Corvettes. I think I got the Titan evasion number right, but it was only a glance, and as I said, I'm very tired.

 

[Immortalits]

Since that was one of my designs, let me elaborate: Arc Emitters and Cloud Lightning do have 100% accuracy, but they have 0% tracking now. I used Line combat computer because Artillery combat computers' range bonus does not work and accuracy over 100% might counter evasion (haven't tested that though). The one stat that does surely give benefit, is the +20% attack speed. I put Enigmatic Decoders into aux slots because of their tracking bonus and (again) the off-chance that increasing accuracy over 100% might counter evasion. The utility slots were all hulltank, because the opponent fleet had weapons with more bonuses against shields and armour than hull and I didn't remember that taking hull damage lowers fire rate now.

Yeah, sorry, did not noticed, that in 2.0, arc weapons lost their 100% tracking, sorry for that, but still, there's a mechanic now, that reduces your ships combat effectiveness up to 50% at ~0% hull, so if you have only hull it's not that great, also, for survival, purposes, both the titan's regeneration aura and the regenerative hull for aux slot gives more armor regen, than hull regen, and by rough numbers, you can restore at elast double the armor points/day, than hull and the max rank techs gives you the exact same ammount to them, (dragon armor and red crystal plating). So even if the opponent deals double armor and no negative hull damage, it's better to focus on armor for sustained survivability and better combat performance in the early parts of the battles, which are both important, especially in the full arc setup. You can calculate easily with a ~30 day battle, where you restore 60% armor and 30% hull with a single armor regen aux.
Also, there is a serious setback for the cloud lightning curently, which is, that it's only a tier 2.5 ranked weapon by it's statts. I've made a mod, where I've buffed it up to be comparable with lvl 2 kinetic artillery. (Compared the stats of lvl 2 arc emitter and lvl 2 giga cannon, made a rough calculation, that said, for the same power cloud lightning need a close to 3 times damage buff to make up the power difference - also made it cost the same as kinetic artillery and gave it the same range, because arc emitter and giga cannon has the same range too.)
I'm 100% sure, with the new weapon and defensive layer reworks, the arc weapons will be great in battles of large ships.
I also don't like the fact, that I'm not allowed to make 'tank' battleships... so I've made them able to use picket computers in that mod too -> see miracles, when you let battleships do, what they should be able to with better tracking.
(cloud lightning reworked stats are: 155 mineral cost, 90 power cost, 1-395 damage, 120 range - as an extra, I also gave +10% tracking for arc and lightning, because how the fk are you evading an arc, that is only possible to create, when you are actually hit - arcs are created between 2 poins of high voltage, if you miss with an arc, you're practically create an arc between a point and nothing, which is impossible - I'm working on a rework for them, where there's more tracking on them at the cost of damage, but that's hard to balance out, this 10% tracking is already a big 10% extra dps boost, so I might lower their damage too.)

If you like the nature of arc weapons, I can upload my own mod to steam so you can use it, but it's realy easy to create, just the saving process is a bit strange for those files.

 

[moyang]

Good tests Fa1nan, thank you. I have a question though: What computer did you use, line of artillery? Apparently the range bonus from the artillery computer is bugged, so line computer gives better combat bonus.

Amongst weapons you used the weapon with lowest accuracy is KA, which have 75%. If you used line computer, it becomes 95% and 2 Auxiliary Fire Controls will make it 105%. Does accuracy over 100% have any effect?

You can calculate easily with a ~30 day battle, where you restore 60% armor and 30% hull with a single armor regen aux.

The regen from regenerative hull tissue is static +1 hull and +2 armor per day, not % of ships's total hull/armor. Not very impressive outside of small corvette battles.

 

[Torakka]

Amongst weapons you used the weapon with lowest accuracy is KA, which have 75%. If you used line computer, it becomes 95% and 2 Auxiliary Fire Controls will make it 105%. Does accuracy over 100% have any effect?

Someone wrote earlier in this thread (or some other similar thread) that accuracy over 100% has no effect and apparently had also tested it. However, if I recall correctly the line CC gives only +15% accuracy, KA (75%) + line CC (15%) + 2x aux FC (5%) would be exactly 100% accuracy.

 

[moyang]

Someone wrote earlier in this thread (or some other similar thread) that accuracy over 100% has no effect and apparently had also tested it. However, if I recall correctly the line CC gives only +15% accuracy, KA (75%) + line CC (15%) + 2x aux FC (5%) would be exactly 100% accuracy.

Thanks for quick answer.

Sapient line computer gives 20% accuracy, precog gives 15% accuracy and 10% tracking. I suspect precog will do better against FE, but you don't always go that path.

 

[Fa1nan]

Good tests Fa1nan, thank you. I have a question though: What computer did you use, line of artillery? Apparently the range bonus from the artillery computer is bugged, so line computer gives better combat bonus.

Amongst weapons you used the weapon with lowest accuracy is KA, which have 75%. If you used line computer, it becomes 95% and 2 Auxiliary Fire Controls will make it 105%. Does accuracy over 100% have any effect?

The regen from regenerative hull tissue is static +1 hull and +2 armor per day, not % of ships's total hull/armor. Not very impressive outside of small corvette battles.

Ah, yeah, forgot to mention the computers. Sapient Line computers for all precisely because the artillery ones don't work; but I did not put much thought into it because this first test series is more about testing the waters. No point in min-maxing BB fleets if some other hull should happen to prove itself as clearly superior. For the Corvettes mentioned in the edit I used Swarm. Should probably add that to the notation somehow...

 

[Immortalits]

The regen from regenerative hull tissue is static +1 hull and +2 armor per day, not % of ships's total hull/armor. Not very impressive outside of small corvette battles.

It says static in the game files, but it interprets into % values in the game, at least, my ships regenerate with insane speed and can go almost infinitely in the enemy territory, waiting a bit before moving to the next larger fight. If it would realy be just 1 point of hull and 2 point of armor regenerated, then my ships won't be at max from 50% remaining hull in just a month, I don't know if the 2.0.2 changes it or I just misinterpreted these things or anything, so I'll check them, when I get home, if they're realy just flat ammount, I'll straight up mod it to have a real use, because flat ammount of hull/armor restore realy isn't good for even corvette fights...

 

[yski2moreletters]

I don't know how to run the tests on my own.

The way I do it, is I set up a galaxy with two identical empires for the tests. I colonize right next to the other with one of them, and have the fleets meet in a nearby system.

Console commands wise, ai to disable ai. play 0 and play 1 to switch between empires, and instant_build to instantly build your fleets. research_technologies for tech, or research_technologies 1 if you want all the special leviathan tech etc as well. finish_research to finish current research if you want repeatables.

If you want to test admirals, debugtooltips to see IDs when you hover over them. add_trait_leader 123 leader_trait_traitname, where 123 is the leader ID and leader_trait_traitname is the name of the trait you want to add, look here for list: http://stellaris.paradoxwikis.com/Console_commands#List_of_Traits_.28Individuals.29
Similarly, remove_trait_leader 123 leader_trait_traitname to remove traits.

..I think that's about it. Save before running a test, and name saves appropriately. Then running the test becomes a matter of loading the save, opening console and and re-disabling AI and enabling instant_build to avoid any resource shortages or AI suddenly adding leaders or something messing up the tests. The up arrow key gives you previous console commands, so you can do it really quickly.

 

[yavorh55]

Can we please get a mod which makes the ship hulls do what they are ACTUALLY meant to do as roles?
Vettes kill cruisers and BBs in swarm and kill each other in PD.
Destroyers kill vettes in picket/line and kill BBs, then cruisers, then Arty destroyers inn artilery.
Cruisers kill destroyers in picket, then cruisers in line.
BBs kill cruisers and each other.

Those are the actual roles of ships! There aren't different hulls simply to pummel at each other, there are different hulls to counter each other- like in EVE!

 

[moyang]

It says static in the game files, but it interprets into % values in the game, at least, my ships regenerate with insane speed and can go almost infinitely in the enemy territory, waiting a bit before moving to the next larger fight. If it would realy be just 1 point of hull and 2 point of armor regenerated, then my ships won't be at max from 50% remaining hull in just a month, I don't know if the 2.0.2 changes it or I just misinterpreted these things or anything, so I'll check them, when I get home, if they're realy just flat ammount, I'll straight up mod it to have a real use, because flat ammount of hull/armor restore realy isn't good for even corvette fights...

Just tested it, it works just like ingame tooltip says. To make things worse, the regen don't work in combat.

screenshots: https://imgur.com/a/UBYPd

 

[Immortalits]

Can we please get a mod which makes the ship hulls do what they are ACTUALLY meant to do as roles?
Vettes kill cruisers and BBs in swarm and kill each other in PD.
Destroyers kill vettes in picket/line and kill BBs, then cruisers, then Arty destroyers inn artilery.
Cruisers kill destroyers in picket, then cruisers in line.
BBs kill cruisers and each other.

Those are the actual roles of ships! There aren't different hulls simply to pummel at each other, there are different hulls to counter each other- like in EVE!

Yeah, but it creates a game, where corvette spam rules, like now and before along the whole damn hystory of the game... corvette spam FTW, everyone plays other ship sizes because it's cool, not because it's effective.
Corvettes has the best cost effectiveness because of their high evasion and absolutely the best dps in the entire game. They provide 1 G slot / naval point, which no other ships are capable of which is great vs everything.
Also, you just wrote down the problem itself: corvettes kill everything exept specialized destoyers, which sucks against everything else -> why bother building anything else other than corvettes then?
Larger weapons has a serious disadvantage, which is tracking, after they finish their 'size', they're highly ineffective vs smaller sizes, especially XL weapons.
Heck, giga cannon has only 75% hit chance, if you don't boost it with +25% you won't even hit 1/4 of the shots, which is huge with an extremely slow firing speed weapon.
Then comes the next problem, tracking, with 0 tracking and cruisers having 20-30% evasion, they already miss quite a lot, then they supposed to go down to destroyers and corvettes with 50-60% and 90-110% evasion... you're joking right?
Also, XL weapons are roughly worth 3 L weapons in dps, but they cost around 2 L weapons, but are solely good vs their size, after that, they're the most inefficient weapons in the game.
so you're telling, that we should build mixxed fleets, where:
BB - BB
Cruiser - Cruiser + Destroyer
Destroyer - Cruiser + Corvette
Corvette - Corvette, Cruiser, BB (and I think they're also good at killing Destroyers too, even specialized ones)

In this setup, you should just spam corvettes, because the enemy fleet's 'big guns' will become nonegsitent, giving you a big edge in effetive fleet power.

 

[Immortalits]

Just tested it, it works just like ingame tooltip says. To make things worse, the regen don't work in combat.

screenshots: https://imgur.com/a/UBYPd

Thanks for testing, I'll right away mod it for myself, because it's plain stupid for regen not be % based... I don't have a quick idea to solve the combat regen issue unfortunately :/

 

[Saori]

Thanks for testing, I'll right away mod it for myself, because it's plain stupid for regen not be % based... I don't have a quick idea to solve the combat regen issue unfortunately :/

Well. it does make it an asymetrical choice - I actually like those.

Its good for smaller hulls and worse for bigger ones. Instead of one auxilarity you slap onto all your hulls one has to think now about what kind of Aux is useful when.
The two other regen buffs in the game btw are %-based. So you have variety in your game. (Admiral trait and Living Metal are both % regen effects)

This asymmetry would really shine if the target behaviour would get updated somehow or be able to be player adjusted - to decrease the dominance of monohull fleets.

 

[Immortalits]

The 1 point hull and 2 point armor regen is plain useless, it does not affect combat at all, since there are so much bonuses to hull and armor numbers, that are % based (repeatable techs, racial passives - devouring swarm, etc) after a few repeatable tech and the flat bonuses to hull for certain ship sizes, that small regen will disappear, making even the 5% hit chance boost ultimately better than the regenerative hull aux. Battles rarely enough time to make those few point noticable even in large numbers, while giving 10% shield from shield charger is just great, even as shield can be ignored in a lot of weapon setups.
However, considering the game's curent state, the 1% boost should be a bit OP, if we look at trait bonuses or the devouring swarm innate 0,5% regen abilty, so I think I'll make it 0.1% regen instead of 1 for hull and 0.2% for armor, so it will actually be usefull, but not overpowered and it will take 100+ days to come out ahead of a flat 10% shield boost.

 

[Saori]

The 1 point hull and 2 point armor regen is plain useless, it does not affect combat at all, since there are so much bonuses to hull and armor numbers, that are % based (repeatable techs, racial passives - devouring swarm, etc) after a few repeatable tech and the flat bonuses to hull for certain ship sizes, that small regen will disappear, making even the 5% hit chance boost ultimately better than the regenerative hull aux. Battles rarely enough time to make those few point noticable even in large numbers, while giving 10% shield from shield charger is just great, even as shield can be ignored in a lot of weapon setups.
However, considering the game's curent state, the 1% boost should be a bit OP, if we look at trait bonuses or the devouring swarm innate 0,5% regen abilty, so I think I'll make it 0.1% regen instead of 1 for hull and 0.2% for armor, so it will actually be usefull, but not overpowered and it will take 100+ days to come out ahead of a flat 10% shield boost.

so that leaves us with an auxilarity thats mostly usefull for the early and midgame? Good for corvetts and the hit and run doctrine? that lets a fleet operate deep behind enemy lines without a need to to fly back home to repair themself?

In the endgame that becomes useless - but okay? Thats fair enough.

People in balancing discussion tend to only look at endgame fleets - thats a bit wrong imo.

I personally dislike % boosts - they are static over all hull types and tech levels and thus really boring. I rather have a bunch of auxilarities which are only usefull on certain setups or during specific phases of the game then a bunch of fairly static choices.

 

[Raithnor]

They

Yeah, but it creates a game, where corvette spam rules, like now and before along the whole damn hystory of the game... corvette spam FTW, everyone plays other ship sizes because it's cool, not because it's effective.
Corvettes has the best cost effectiveness because of their high evasion and absolutely the best dps in the entire game. They provide 1 G slot / naval point, which no other ships are capable of which is great vs everything.
Also, you just wrote down the problem itself: corvettes kill everything except specialized destroyers, which sucks against everything else -> why bother building anything else other than corvettes then?
Larger weapons has a serious disadvantage, which is tracking, after they finish their 'size', they're highly ineffective vs smaller sizes, especially XL weapons.
Heck, giga cannon has only 75% hit chance, if you don't boost it with +25% you won't even hit 1/4 of the shots, which is huge with an extremely slow firing speed weapon.
Then comes the next problem, tracking, with 0 tracking and cruisers having 20-30% evasion, they already miss quite a lot, then they supposed to go down to destroyers and corvettes with 50-60% and 90-110% evasion... you're joking right?
Also, XL weapons are roughly worth 3 L weapons in dps, but they cost around 2 L weapons, but are solely good vs their size, after that, they're the most inefficient weapons in the game.
so you're telling, that we should build mixed fleets, where:
BB - BB
Cruiser - Cruiser + Destroyer
Destroyer - Cruiser + Corvette
Corvette - Corvette, Cruiser, BB (and I think they're also good at killing Destroyers too, even specialized ones)

In this setup, you should just spam corvettes, because the enemy fleet's 'big guns' will become nonexistent, giving you a big edge in effective fleet power.

BB's should also specialize in taking out Star Fortresses and Defense Platforms.

I've been using Squadrons (12) of Corvettes as Pirate Hunters and Commerce Raiders. If you keep hitting a warring empire's back line of outposts their main battle fleets will keep trying to retake the outposts before waging war on your territory. You can Rope-a-dope them this way until war exhaustion kicks in.

I'm thinking to discourage Corvette Spam, they need to be so fragile that the cost of replacing them is more than a destroyer in the same amount of time. Another idea would be to buff accuracy or overall toughness, or even discount the build cost so you could get 2 Destroyers for the same minerals as 3 Corvettes.

There needs to be a narrative to the technology of Starship design that mirrors the blue water equivalent. Destroyers were invented to destroy Corvettes, Cruisers were invented to Destroy Destroyers, and so on. Once bigger ships came on the scene Destroyers were used to screen for other ships.

Honestly, I'd like to see the Hearts of Iron Team and Stellaris team collaborate to correct issues in both games. Since Hard Military Sci-Fi is usually just WWII Navies in Space.

 

[Fa1nan]

The way I do it, is I set up a galaxy with two identical empires for the tests. I colonize right next to the other with one of them, and have the fleets meet in a nearby system.

Console commands wise, ai to disable ai. play 0 and play 1 to switch between empires, and instant_build to instantly build your fleets. research_technologies for tech, or research_technologies 1 if you want all the special leviathan tech etc as well. finish_research to finish current research if you want repeatables.

If you want to test admirals, debugtooltips to see IDs when you hover over them. add_trait_leader 123 leader_trait_traitname, where 123 is the leader ID and leader_trait_traitname is the name of the trait you want to add, look here for list: http://stellaris.paradoxwikis.com/Console_commands#List_of_Traits_.28Individuals.29
Similarly, remove_trait_leader 123 leader_trait_traitname to remove traits.

..I think that's about it. Save before running a test, and name saves appropriately. Then running the test becomes a matter of loading the save, opening console and and re-disabling AI and enabling instant_build to avoid any resource shortages or AI suddenly adding leaders or something messing up the tests. The up arrow key gives you previous console commands, so you can do it really quickly.

"attackallfleets" saves you the trouble of setting up another empire.

 

[Luchofeio]

New to the thread, can someone just please tell me it is ok to use carriers vs the AI??! That´s all I want to know =D (even if you are lying)