The Large Fleet Collider - 2.0 Fleet Balance

[Jmes Snowscoran]

You'd think the cap would be on (evasion-tracking), not on evasion itself. So you can't stack evasion to be literally untouchable, but you can stack evasion to counter tracking.Small differences in combat formulas have such huge consequences on overall balance.

Not a great idea as it would make corvettes even more batshit than they are.

Also, something I discovered today, which there seems to be some confusion about, is shield regen in combat which is definitely a thing, even without shield capacitors. You can observe it the first time you take out a pirate station with your starting corvettes. The thing is that a ship needs to go a few ticks without being attacked to start regenerating, which means it's going to be marginal in most engagements, but when using monofleets where enemies are liable to switch targets a lot you might see a small benefit when ships are left alone for a bit.

 

[Kraegor]

Corvette mono-fleet utterly wiped the floor with the Unbidden for me yesterday. All hail evasion.

Against the Scourge they did well. I went in with mixed fleets, lost roughly 50-75% of my fleet everytime. Batteship and Cruisers mono fleets suffered the same fate. But pure 200 corvette missle/laser ships only lost 15-25% every time.

With a 2 seperate 6 shipyard each stations in the area, I was able to restock the fleets in no-time flat. Was able to push back the scourge while every other FE and Big Empire was losing massive amounts of ships.

 

[EntropyAvatar]

Not a great idea as it would make corvettes even more batshit than they are.

It's probably true that this change in isolation would make balance worse. However, I don't think rules should be designed to balance the game with the current set of parameters. Rather you should choose rules that make sense and are robust, then choose parameters that create the desired balance. Note that a bad set of rules often means that it is difficult to impossible to find the right balance.

To me, having a max 90 evasion is just a bad code smell. It says that the designers are worried about evasion, but not enough to consider a different way of making it work. Probably the worst thing about evasion is increasing marginal returns. A small evasion advantage is hardly noticeable, but each incremental boost in evasion becomes a bigger and bigger step towards invulnerability. This is a common problem with subtractive systems (similar to percent damage resistance - which subtracts damage resistance from 100%).

What about a system where hit chance is proportional to max(1, tracking/evasion)? In that model (as now), evasion below tracking is not useful, but evasion above tracking basically linearly improves your expected survival time.

 

[skizz]

Against the Scourge they did well. I went in with mixed fleets, lost roughly 50-75% of my fleet everytime. Batteship and Cruisers mono fleets suffered the same fate. But pure 200 corvette missle/laser ships only lost 15-25% every time.

With a 2 seperate 6 shipyard each stations in the area, I was able to restock the fleets in no-time flat. Was able to push back the scourge while every other FE and Big Empire was losing massive amounts of ships.

Arc Emitter / Tachyon Artillery battleship with either destroyer or cruiser as PDs + anti missile flyers is your friend here. battleship with hangar and PD works, too. But I prefer to have destroyer with double PD and long ranged gun. I usually do 4 destroyers per battleship. On my last game I messed up game settings, tho. 400k fleet power vs waves of ~100k scourge is not even close to balanced and I deleted my PD ships (11x repeatables). Even awakened empires cleaned up scourge : - /

 

[Talamare]

Did some tests and the results are here
https://imgur.com/a/uDlUh

For the Mixed Fleet used for Testing, I used
30 Guns Corvettes
20 Picket Destroyers
12 Cruisers
6 Battleships

For the Monofleets, I tried to always have 150 (152 Fleet Capacity, if 150 wasn't possible).
Resource for the Monofleets were usually around 43k to the point that I stopped bothering recording it after a little bit.
Guns for the Monofleet were always a balanced between Gauss Cannon and Gamma Laser; with a slight preference to the higher DPS Gauss Cannon when perfect equality was impossible.

Here are some observations.

Guns vs Missiles Corvette - They both performed insanely well, both were easily able to destroy the Mixed Fleet with near equal loss. Remember that the Mixed Fleet did have 40 PD.
Disruptors - Proved to be fairly poor against a Mixed Fleet, resulting in narrow losses.
Hangars - Performed the worst of everything! Essentially Complete Garbage! Needs massive buffs!
Medium Destroyers vs Large Destroyers - Destroyers focusing on Medium Weapons lost to the Mixed Fleet, but Destroyers focusing on Large Weapons actually won!
Med/Large Cruisers - Pretty similar results here as well; Bringing Large Weapons helped secure the Win.
Battleships - Battleships without their X Weapon won but spamming Large Weapons won, but it was fairly narrow. With the X Weapon, it was a MASSIVE DOMINATION!

Edit
My suggestion in which Corvettes are reduced to only 2 Defense Items instead of 3.
- The Corvettes had slightly increased losses, but overall still had a dominating Win.

I would say they probably need their Base 60% Evasion reduced to 50%.

 

[GloatingSwine]

Not sure if it still the case in 2.0, but the Picket Bow had an extra defense slot which is... meh. Destroyers currently don't have access to missiles and point defense seems really inadequate right now.

No, the whole layout of defence slots changes in 2.0.

The big problems right now are:


L slots are too good. Especially the exclusive L weapons with tier 6 versions. The increase in damage scaling with slot size was not offset by anything, and it's very easy to boost their accuracy and/or tracking. That means there is no reason to use anything but the maximum number of L slot weapons your fleet can sustain*. And that means Battleships because they're the only ones that can have all L slot weapons. (That's their real bonus now. That and loads of hitpoints.)

The least accurate L/XL slot weapon on a Battleship is going to hit the most evasive possible Destroyer 48% of the time.

Evasion is not good enough. Only corvettes have enough evasion to slightly mitigate those L slot weapons. For every other ship class evasion is a useless dead stat. Also, tracking is far too easy to get. Tracking should be restricted to only that granted by the weapon. That's the only way to actually make non-L weapons necessary. Also there should be no 90% hardcap. L weapons should be unable to hit Corvettes specced for Evasion, you should need S weapons to stop them, as it stands Corvettes simply do not need most of their Evasion options by the end of the game because 110% evasion does nobody any good, and nobody needs any S weapons because you can trivially add enough chance to hit and tracking to hit anyway.

Point Defence is a failed experiment. It needs to go away and missile weapons need to be balanced against the normal set of defences. Otherwise they will simply vacillate back and forth between irrelevant and overpowered as the balance of point defence wanders between easily stopping all the missiles and not stopping any of them.


* By which I of course mean use infinite torpedo corvettes because Torpedoes make L slot weapons look like they're not even trying. 100% accuracy and 55 tracking because you don't need the Swarm computer ever means 65% to hit evasion capped corvettes and guaranteed hits on everything else, ignore shields and +100% armour damage. that's what I mean by making them interact with normal defences, they're balanced as if point defence is going to stop some of them but that balance point is far too fine to ever reach in real life, even if PD wasn't utter garbage, it will always be all or nothing.

 

[Jmes Snowscoran]

It's probably true that this change in isolation would make balance worse. However, I don't think rules should be designed to balance the game with the current set of parameters. Rather you should choose rules that make sense and are robust, then choose parameters that create the desired balance. Note that a bad set of rules often means that it is difficult to impossible to find the right balance.

To me, having a max 90 evasion is just a bad code smell. It says that the designers are worried about evasion, but not enough to consider a different way of making it work. Probably the worst thing about evasion is increasing marginal returns. A small evasion advantage is hardly noticeable, but each incremental boost in evasion becomes a bigger and bigger step towards invulnerability. This is a common problem with subtractive systems (similar to percent damage resistance - which subtracts damage resistance from 100%).

What about a system where hit chance is proportional to max(1, tracking/evasion)? In that model (as now), evasion below tracking is not useful, but evasion above tracking basically linearly improves your expected survival time.

This is some pretty solid reasoning. I think your proposed system has some glaring flaws, though. First off I think your max should be min. Second it doesn't take into account weapon accuracy when calculating hit chance. Third it does nothing for the problem where ships with mediocre evasion (cruisers and to a certain extent destroyers) get little or no benefit against most attacks.

Let's say we define a ships evasion chance instead as min(0.9, evasion*max(0, 1-tracking)). In this model weapon tracking takes away a portion of the defending ship's evasion rather than a fixed number. Thus a weapon firing with 50% tracking would give a corvette with 140% evasion a 70% chance to evade, and a cruiser with 20% evasion would have a 10% chance to evade. If the same corvette is shot at by a weapon with 0% tracking, its effective chance to evade becomes 90% (where it's capped). If the weapon has 100% tracking, all evasion is negated regardless of how high it is.

 

[Talamare]

No, the whole layout of defence slots changes in 2.0.

The big problems right now are:


L slots are too good. Especially the exclusive L weapons with tier 6 versions. The increase in damage scaling with slot size was not offset by anything, and it's very easy to boost their accuracy and/or tracking. That means there is no reason to use anything but the maximum number of L slot weapons your fleet can sustain*. And that means Battleships because they're the only ones that can have all L slot weapons. (That's their real bonus now. That and loads of hitpoints.)

The least accurate L/XL slot weapon on a Battleship is going to hit the most evasive possible Destroyer 48% of the time.

Evasion is not good enough. Only corvettes have enough evasion to slightly mitigate those L slot weapons. For every other ship class evasion is a useless dead stat. Also, tracking is far too easy to get. Tracking should be restricted to only that granted by the weapon. That's the only way to actually make non-L weapons necessary. Also there should be no 90% hardcap. L weapons should be unable to hit Corvettes specced for Evasion, you should need S weapons to stop them, as it stands Corvettes simply do not need most of their Evasion options by the end of the game because 110% evasion does nobody any good, and nobody needs any S weapons because you can trivially add enough chance to hit and tracking to hit anyway.

Point Defence is a failed experiment. It needs to go away and missile weapons need to be balanced against the normal set of defences. Otherwise they will simply vacillate back and forth between irrelevant and overpowered as the balance of point defence wanders between easily stopping all the missiles and not stopping any of them.


* By which I of course mean use infinite torpedo corvettes because Torpedoes make L slot weapons look like they're not even trying. 100% accuracy and 55 tracking because you don't need the Swarm computer ever means 65% to hit evasion capped corvettes and guaranteed hits on everything else, ignore shields and +100% armour damage. that's what I mean by making them interact with normal defences, they're balanced as if point defence is going to stop some of them but that balance point is far too fine to ever reach in real life, even if PD wasn't utter garbage, it will always be all or nothing.

Basically we are talking about vastly increasing the extremes on both Evasion and Tracking?
Small Weapons should be pushing 70-90% Tracking
Medium Weapons probably 40-60% Tracking
Large Weapons maybe 10-20% Tracking
X-Large Weapons probably 0-10% Tracking

Then changing the Tracking System to Evasion minus Tracking, with 90% Hard Cap. (Instead of Evasion 90% Hard Cap, minus Tracking)
Edit - Including Jmes Snow suggestion, Evasion multiplied by Tracking, with a 90% Hard Cap.

Finally Buffing Evasion on Corvettes to 70%?
Destroyers to 50%
Cruisers to 30%

 

[EntropyAvatar]

First off I think your max should be min. Second it doesn't take into account weapon accuracy when calculating hit chance. Third it does nothing for the problem where ships with mediocre evasion (cruisers and to a certain extent destroyers) get little or no benefit against most attacks.

Max should be min: You are right. My brain doesn't work!

Missing weapon acc: I meant to-hit is proportional to the bit I wrote. You can always multiply by accuracy. Though I find the concept of separate accuracy and tracking a bit weird. I mean, if the target is literally the size of a moon and 10 km away, if there any weapon that could miss? I guess it's a way to make damage more unpredictable.

Mediocre evasion: Well, the beautiful thing about the multiplicative function is you can set parameters to a very wide range. So you can set the appropriate sized weapon to be at say 80% of the typical evasion of the target. Then you make modifiers also multiplicative (e.g. engines boost evasion by 10%). So you pick it so that a "correct" sized weapon for a ship class will have maybe 80% to hit on the typical evasion. Smaller weapons are a more reliable hit, but are less damage-efficient.

Say a large weapon is 3x the damage of a medium, and so on. So the "right" sized weapon will be 3x0.8=2.4 damage, so still 20% better than using a "too small" weapon. You try to make it so a Cruiser might be agile enough that large weapons aren't a great choice against it (say evades half the shots), but it will only evade 10% of a medium.

Of the top of my head, let's say a CR is 50 evasion. Then a large might be 40 tracking (80% of base accuracy). Maybe a really buffed CR is 80 evasion (60% bonus!), so the large would be only 50% of base acc. But a medium would be about 70 tracking. Or if it swings the other way (tracking bonuses exceed evasion bonuses), maybe it becomes efficient to use XL weapons on cruisers.

Let's say we define a ships evasion chance instead as min(0.9, evasion*min(0, 1-tracking)). In this model weapon tracking takes away a portion of the defending ship's evasion rather than a fixed number. Thus a weapon firing with 50% tracking would give a corvette with 140% evasion a 70% chance to evade, and a cruiser with 20% evasion would have a 10% chance to evade. If the same corvette is shot at by a weapon with 0% tracking, its effective chance to evade becomes 90% (where it's capped). If the weapon has 100% tracking, all evasion is negated regardless of how high it is.

Having %100 tracking completely negate evasion is one way to allow uncapped evasion. I think going that route, you'd have to be very careful about tracking bonuses though. If you have too many tracking bonuses sprinkled around, evasion becomes a dead stat.[/quote][/QUOTE]

 

[Kraegor]

Arc Emitter / Tachyon Artillery battleship with either destroyer or cruiser as PDs + anti missile flyers is your friend here. battleship with hangar and PD works, too. But I prefer to have destroyer with double PD and long ranged gun. I usually do 4 destroyers per battleship. On my last game I messed up game settings, tho. 400k fleet power vs waves of ~100k scourge is not even close to balanced and I deleted my PD ships (11x repeatables). Even awakened empires cleaned up scourge : - /

Nope, I had Tacyon with PD and destroyers and Corvettes. Still lost 50-75% every single battle. My battleships went down first, and hard. I had 120k fleet power vs their 110k fleet power and I had the 50% extra damage perk. They still took out my destroyers and battleships like they were made of butter. Only the high evasion corvettes came out on top every time.

 

[GloatingSwine]

Aha! I have it!

A part of the puzzle is missing.

Let's go from first principles. Including what I've already said about removing the easy sources of accuracy and tracking, those should be defined wholly by the weapon. (Higher tiers of weapon could get more to have a progression through the game).

We'll start with a Battleship monofleet. Battleships should be functionally unable to fight Corvettes, Corvettes should be avoiding all damage from L weapons and most from M weapons Battleships should have no S slots available and only few M slots.

Battleship fleets, therefore, should need another ship to protect them from Corvettes. Which should be Destroyers, the point all along has been that Destroyers should be good at killing Corvettes. So Destroyers should predominantly be packing S slot weapons to fight Corvettes, but have others available to fight cruisers. (They probably need a 6S loadout available).

Now, even if we make Destroyers very good at killing Corvettes, 6S weapons, where S weapons have enough tracking to hit more often than they miss, there will never be enough of them to stop an all-Corvette fleet from just monstering the battleships, because there will probably be three to four corvettes for every destroyer at least. So the Destroyers need to be able to stop the corvettes not just kill them a bit better.

So we need the missing piece of the puzzle.

We need Suppression.

Every time a ship successfully evades (as in the to hit roll was between the base accuracy and the accuracy modified by evasion-tracking), its own accuracy should be reduced by the tracking of the weapon it just evaded for the next 20 days.

Ships should have a preference not to shoot at suppressed targets.

That would mean that a torpedo corvette monofleet trying to strike at a Battleship fleet protected by a reasonable amount of Destroyers simply finds its accuracy smashed into the bin, unable to meaningfully attack the Destroyers directly with Torpedoes and having its attack runs mostly thwarted against the battleships, they would be slowly picked to pieces.

But, of course, the same behaviour is replicated between Destroyers and Cruisers. Cruisers' M weapons ruin the day of Destroyers, damaging them somewhat reliably and reducing their ability to do their own job (because a missed shot causes no suppression, only one that would have hit if not for evasion, so now the Destroyers are less effective at suppressing the Corvettes which can go to town on the Battleships)


That would mean every fleet now needs at least two ship classes working together, any monofleet is going to meet something it can't deal with in even a random zoo fleet.

 

[Talamare]

We'll start with a Battleship monofleet. Battleships should be functionally unable to fight Corvettes, Corvettes should be avoiding all damage from L weapons and most from M weapons Battleships should have no S slots available and only few M slots.

Battleships are already basically lacking in Small Slots, They don't even want them either.
I would like to see Broadside Bow being changed to 4 M, instead of L/M/2S tho

As far as the current Cruiser vs Battleship disparity
Technically Cruisers could also have 4 Large Guns for the same Fleet Space as 1 Battleship.
Tho after the 4 Large Guns we are talking about either an X gun or a 4 Mediums... and everyone knows an X is better than 4M
In that regard... Even Destroyers could have 4 Large Guns (and again 4 Mediums)

Missile Corvettes can't be replicated tho
Cruisers are the only other ship that can hold Guided, they max out at 3G and 6S, which is worse than 4G and 4S.

 

[Fa1nan]

I did a few sloppy tests against crises. Setup was 2x Insane where I triggered the respective event to spawn the crisis and then tried to pick off single fleets. Only one fight for each design because doing it like that is very tedious.

Contingency: First of all, this is by far the hardest crisis. Unbidden and Prethoryn show up with about 300k per fleet (160k for Prethoryn Vanguard), whereas AI gets 426k roaming fleet and 1M at their machine worlds. Smashed 100 AB-4D2V (the anti-FE design) with max shields (because AI crisis fully relies on energy weapons; also the maximum number of shields seems to be 5 without downgrading other stuff) into a roaming fleet; 4 survived while almost killing that fleet; 2 AI BBs and 5 Escorts survived. 84 GB-1K3N could not come close with all 84 ships lost while 6 BBs and 17 Escorts remained. The clear winner here are torpedo corvettes (CV-1D1T) with 386 out of 654 surviving completely annihilating the roaming AI fleet. Conclusion: Use Torpedoswarm with max shields against Contingency.

Unbidden: 100 AB-4D2V killed an Unbidden fleet at a cost of 20 ships, CV-1D1T lost 312. I also tried a full kinetic setup GB-4K where 53 out of 84 remained. Conclusion: Against Unbidden use Disruptor/Cloud Lightning BBs with max shields (because Unbidden Weapons are bad against shields).

Prethoryn: This one is a bit more interesting because their weapons penetrate shields and have bonuses against armor, so there's the question of armor vs crystal plating and of course PD. I first tested weapon loadouts with max armor: 100 AB-4D2V won with 31 ships lost (31% of mineral cost). Fully armored CV-1D1T have similar mineral cost at 623 ships, of which 262 were lost (42% of mineral cost). An anti-armor design TB-4N failed horribly because it couldn't hit the 65% evasion Scourge destroyer.
Therefore, I took the AB-4D2V design and tested ratios of crystal plating versus armor:

Crystal : armor | remaining ships out of 100
      0 : 6     | 70
      1 : 5     | 70
      2 : 4     | 81
      3 : 3     | 78
      4 : 2     | 77
      5 : 1     | 86
      6 : 0     | 80

Despite the randomness, this does indicate that crystal plating is better than armor and knowing targeting, 5:1 is likely best. So I took this defensive setup and tested different ratios of AB-4D2V and carriers AB-4D2H with one PD, one Flak and two of the Disruptors in small slots (basically, the two hangar bay middle section). Results:

Percentage of carriers | remaining ships out of 100
                     0 | 86 (from before)
                    10 | 78
                    20 | 82
                    30 | 75
                    40 | 81
                    50 | 73
                    60 | 74

While this kinda confirms that 5:1 is a good defensive setup, these numbers point towards not swapping in any carriers. This will be especially true if you do not have Cloud Lightning and have to resort to the middle section with one hangar bay and 4 medium slots. Since you'll have corvettes lying around anyway, I conclude: Use Disruptors/Cloud Lightning with 5:1 crystal plating to armor ratio and put PD on any corvette swarms you may have lying around.

 

[Acidpunk]

Aha! I have it!

Battleship fleets, therefore, should need another ship to protect them from Corvettes. Which should be Destroyers, the point all along has been that Destroyers should be good at killing Corvettes. So Destroyers should predominantly be packing S slot weapons to fight Corvettes, but have others available to fight cruisers. (They probably need a 6S loadout available).

Now, even if we make Destroyers very good at killing Corvettes, 6S weapons, where S weapons have enough tracking to hit more often than they miss, there will never be enough of them to stop an all-Corvette fleet from just monstering the battleships, because there will probably be three to four corvettes for every destroyer at least. So the Destroyers need to be able to stop the corvettes not just kill them a bit better.

So we need the missing piece of the puzzle.

What we need is to give destroyers two missile slots on the rear ship module, this way they can obliterate corvettes.

 

[Talamare]

Serious Question tho
What should Cruisers be good at?

Should they really just be Baby's First Battleship?
Should they have a defining Role?

 

[Defiler99]

I think Cruisers should have specialized role modules, like ECM, mini auras, etc.
If they're just going to be scaled-down battleships that are slower than destroyers, they might as well just get removed.

 

[BaronNoir]

I have a serious question-my Destroyers took monstruous casualties in fleet battles. Is there a behavior/build that limit casualties ? (I tend to use them as the AA ship)

 

[Bouchart]

You are going to take casualties in all but the most lopsided battles. Based on everything I've read here people are gravitating towards either torpedo boat corvettes or battleships with long range weapons like kinetic artillery.

 

[tangled axile]

What's better against Citadels, CV swarm or BBs?

 

[Bouchart]

Probably battleships. Starbases can't evade so long range weapons work well against them. Starbases also have a mix of weapons so the long range of battleships would help avoid a few of them like autocannons.