Mer wrote:
How is it working on lasers? Critical hits are calculated before or after shield dmg absorption?

Someone said that critical hits are applied after absorption, but damage bonuses are applied before absorption (giving Beams a completely unfair advantage over the other two weapon types in terms of beating defenses). I'm not sure if this has been verified though.

That's exactly why I'm looking for confirmation in either way.



I'm writing combat simulator so i could put genetic algorithms on optimization task. And every combat detail is quite important to be confirmed that is working in exact way so i could model it in my simulator.

(If experiment become successful then it could be cloned(with some optimization) into higher level AI... that's just my speculation. And there's a lot of whining about extremely crappy AI ship design)

Mer wrote:
Is it confirmed information(didn't found such info anywhere yet)?

How interception chance is affected by flak accuracy and missile evasion?



Also anyone know what's exact mechanic of critical hits?

for Kinetic/Missile I guess it's: If particle/missile hits(surpassed defense) then there's chance to multiply damage.

How is it working on lasers? Critical hits are calculated before or after shield dmg absorption?

In other words

offence has laser(100% accuracy) that deal avg. 30 dmg and 100% chance for critical hit x3

defense has shield that absorb 70 damage

Q: Is one shot from offence weapon going to surpass defense shield and deal damage to health?

Regarding missile defenses, I haven't rigorously tested it, but a careful parsing of the descriptions says that they have a "chance per round" to shoot down a missile. However, I know from experience that, for example, 3 flaks can shoot down 5+ incoming missiles. (I saw 5 missile trails from the ships... I'm not sure if any of the targets got double-teamed, though.) There was someone on the forums complaining awhile back that he was firing enough missiles that they had to have been shooting down three per defense just because there was no way they could have gotten as many of his missiles otherwise, too, which is what inspired me to try it. In general, though missile defenses turn out to be ... surprisingly effective, although I think shields have a similar tonnage advantage at this point. (That was not the case until v1.0.5, however.)



I also recall a report awhile back that, as someone else posted in this thread, +damage applies before shields but +crit applies after shields, so if you want to melt highly-defended ships, you shouldn't bother with +crit. (It could be entertaining on missiles with their base 20-30% crit rates, though... the question is whether 100% crit rate takes does better or worse than an equal number of +damage modules...)



Edit:

I forgot to comment on targeting... from what I understand, each of your ships selects a random enemy ship for each round of combat in a phase. The round 2 (IIRC) target also gets missiles fired at it, which hit 3 rounds later on round 4. Someone wrote a post explaining this better somewhere on the forums, but it's late, so I won't dig it up right now.



Also, Mer, if you're interested in testing things carefully, look me up on Steam (same handle) and I'd be happy to do things carefully with you. Your goals interest me.

Registered to add to this thread.

Missile spam fleets with nothing but Ion Torpedos can be churned out endlessly.

Playing pilgrim right now, and a 15 stack (even with 20% reduction in tonnage) is obliterating 20k fleets from other factions.

Single player, obviously. Serious difficulty. 3rd game, and I'm working my way up.



It's been noted in other threads, but it's worth mentioning here again. A commander with 'emergency shelter' increases the resilience of a destroyer spam fleet, simply because they rarely die. The commander might take injury on occasion, but if you pick up a couple of +dust on CP heroes, and give the suicide commander -healing cost, it's mitigated very easily, and you can just put him straight into another fleet (or back into the same one in some cases - emergency shelter can be weird).

Even a 4 heavily flakked dreads seemed unable to counter 15 low tech destroyers (which cost ~50 each).



Obviously, a smart player would simply counter it, but the AI is not smart.



I'm up to turn 180, 3rd in tech, and have researched the combat tree only to unlock the artifacts, and invasion modules. I've researched most of the other trees. I haven't even bothered keeping beam fleets as backup. I just spit out 30 or so 'Martyrs' (well, they are the pilgrims) every round on each front (currently three) and they seem capable of holding everything down. Once I have a couple of spare stacks, I send them to invade a planet. The AI will usually sue for peace at that point, so I get 20 or so rounds to terraform stuff before it starts all over again.



I wanted to try and get a non military victory this time, but it seems the AI will attack you whatever you do, and by the time you build up a fleet capable of taking them apart, you might as well take them over. Sorry, that's digression, though.

You are incorrect in your assumptions.



Most tests have been conducted without complicated fleet-wide bonuses and still found approximately a doubling of defense power (for shields).

Phase 3: Metagame

The Neutrino Glass Breaker 2 is the upper limit for all glass cannon design equilibrium.



Glass Breaker 2 (NGB2)

4 Gluon Disruptor

1 Neutrino Pulsion

Cost: 294



The low end of the equilibrium is the NGB1 fit, which counters its larger cousin just as well as it counters the NGC line (which illustrates the folly of racing against weapons with defences).



NGB1

1 Gluon Disruptor

1 AGN Slugs

1 Neutrino Pulsion

Cost: 174



The spooky thing about these ships is that they don't have to be Destroyers. Corvettes and Transports work just as well (if not better) because those are not typically combat ships. You only find out what

they are in a battle, or if they orbit a planet that can scan them with a Collaborator Network. Usually, the ideal fleet is 22 of either of these little monsters, depending on the opponent. Anything else is less cost effective, but I'm sure there are gains to be made by fluctuating around these designs by a a module or two.



I encourage you to test NON-GLASS CANNON ship fits and fleet compositions that can counter the either the NGB1 or NGB2 fleets cost effectively. Remember, you have to be able to deal with:



NGB1: 41469 beam damage per round at 6468 Industry.

NGB2: 10367 beam damage and 11384 kinetic damage per round at 3828 Industry.



These aren't the only types of glass cannon fleets that work; they just happen to be the most efficient in an environment saturated with glass cannons. It's a metagame evolution.





Phase 4: Metagame Evolution

If the NGB1 and NGB2 are the equilibrium of the glass cannon metagame, how does that fit into the broader metagame of Endless Space, and are there fleet compositions containing designs for 2CP and 4 CP ships that have an advantage against the glass cannons?



In Phase 1, we saw how easy it was for a maxed-out fleet of glass cannons to vaporize a fleet of Dreadnoughts built entirely for defence. Since their defensive measures didn't get them anywhere, let's modify our original Dreadnought designs to have sufficient firepower to each kill one NGC1 as they die.



Absorption Dreadnought (ADN)

39 Ablative Wave Shield

1 Transformative Shields

1 Neutrino Capture

4 Gluon Disruptor

1 Smart Cargo

HP: 2359 (2274 + 85 armour)

Damage Reduction: 10750

Damage Reduction in Fleet: 16770 (based on 7 Neutrino Capture, see below)

Cost: 3062



Brick Dreadnought (BDN)

26 Defensive Lensing

21 Reactive Hulls

1 Graviton Shielding

4 Gluon Disruptor

1 Smart Cargo

HP: 26520 (23637 HP + 2883 armour)

Damage Reduction: 0

Cost: 1738



Neither Dreadnought is an effective trade for an NGC1 (cost 1170). Fielding Dreadnoughts against maxed-out glass cannons seems to be a terrible idea at first, but it's not that simple. Remember that the NGC1 is itself trumped by the NGB1 and NGB2, which bookend the equilibrium range in the glass cannon metagame. Therefore, a Dreadnought is viable in a glass cannon metagame if it can survive combat with fleets of NGB1/NGB2 even if it still trades unfavourably with the NGC1s.



Such a Dreadnought would have to be expected to survive against at least equal CP of NGB2s and still kill enough of them to be an efficient investment. Each NGB2 deals an average of 1885 beam damage at a cost of 294 Industry. Assuming that the Dreadnought can kill one NGB2 each round, if it wants to survive a battle while killing 4 NGB2s, it must be able to withstand:



R1 7540

R2 5655

R3 3770

R4 1885

Total 18860



Since the plan is go for four rounds, we must re-evaluate absorbers because their effectiveness scales up with time compared to armour plating. Moreover, we want the Dreadnought's absorbers to negate at least 1885 damage per round, so the total amount of HP the Dreadnought needs to equal that damage drops to:



R1 5655

R2 3770

R3 1885

Total 11320



Three rounds is also favourable for an Ablative Wave Shield Compared to Defensive Lensing, so let's try negating 3770 per round.



R1 3770

R2 1885

R3 0

R4 0

Total 5655



Three rounds is STILL favourable so let's try negating 5655 per round.



R1 1885

R2 0

R3 0

R4 0

Total 1885



Glass Melter Dreadnought (GMDN1)

8 Gluon Disruptor

16 Ablative Wave Shield

1 Neutrino Capture

Cost: 1780

Comments: Assuming a fleet of five GMDN1s, the absorbers negate 5600 beam damage per round. Recall that a L4 Dreadnought has initial 2088 HP with all improvements in the shipyard system, and we now have a Dreadnought that can kill 4 NGB2s and live through it. The extra 4 Gluon Disruptors look like overkill at first, but remember that we also want this ship to be able to kill NGC1 (the maxed out glass cannons); the GMDN1 still does not trade favourably with the NGC1, but killing the NGC1 as the GMDN1 dies at least prevents a completely lopsided outcome. Besides, the additional guns are a good start toward combating bigger ships.



With 5 GMDN1s against 22 NGB2s, it's very likely that the idealized 4-on-1 scenario won't happen; there will be GMDN1s fighting 5 or more NGB2s, with the following results.



5 NGB2s Attrition and Damage

R1 3825 (9425 – 5600) <--GMDN1 dies, killing only one NGB2

R2 1940 (7540 – 5600)

R3 55 (5655 – 5600)

R4 0 (3770 – 5600)



6 NGB2s Attrition and Damage

R1 5710 (11310 – 5600) <--GMDN1 dies, killing only one NGB2

R2 3825 (9425 – 5600)

R3 1940 (7540 – 5600)

R4 55 (5665 – 5600)



These scenarios are even worse than they originally appear because if a GMDN1 dies in R1, the four or five NGB2s that survived their engagement with it switch targets to support their wingmen against the remaining GMDN1s. The battle plays out like this:



Round 1: 5 GMDN1s and 22 NGB2s

3 GMDN1s each fighting 4 NGB2s (-3 NGB2; each GMDN1 takes 1940 damage)

1 GMDN1 fighting 3 NGB2s (-1 NGB2)

1 GMDN1 fighting 5 NGB2s (-1 NGB2; -1 GMDN1)



Round 2a: 4 GMDN1s and 17 NGB2s

3 damaged GMDN1s each fighting 4 NGB2s (-3 NGB2s; -3 GMDN1s)

1 GMDN1 fighting 5 NGB2s (-1 NGB2; -1 GMDN1)



OR



Round 2b: GMDN1s and 17 NGB2s

2 damaged GMDN1s each fighting 4 NGB2s (-2 NGB2s; -2 GMDN1s)

1 damaged GMDN1s each fighting 5 NGB2s (-1 NGB2; -1 GMDN1)

1 GMDN1 fighting 4 NGB2s (-1 NGB2; 1 GMDN1 takes 1940 damage)



From here, the best case scenario is that Round 3 occurs and another NGB2 dies, making the total casualty report 5 GMDN1s and 10 NGB2s, which is an economic blowout against the Dreadnoughts. Therefore, the GMDM1s need to be able to survive at least 5 NGB2, so we'll boost their absorbers to negate roughly 4 NGB2s per round (7540 beam damage).



GMDN2

8 Gluon Disruptor

22 Ablative Wave Shield

1 Neutrino Capture

Cost: 2140

Comments: This build absorbs 7700 beam damage per round, so it kills 4 NGB2s without getting damaged, and it kills 5 NGB2s while taking 1725 damage as per the following table (363 HP remaining). Instant death only occurs against 6 NGB2s. Let's look at the 6-on-1 scenario, focusing on only one 6-on-1 occurring each round.



Round 1: 5 GMDN2s and 22 NGB2s

2 GMDN2s each fighting 4 NGB2s (-2 NGB2)

2 GMDN2 fighting 3 NGB2s (-2 NGB2)

1 GMDN2 fighting 5 NGB2s (-1 NGB2; -1 GMDN2)



Round 2: 4 GMDN12s and 17 NGB2s

2 GMDN2s each fighting 4 NGB2s (-2 NGB2)

1 GMDN2 fighting 3 NGB2s (-1 NGB2)

1 GMDN2 fighting 6 NGB2s (-1 NGB2; -1 GMDN2)



Round 3: 3 GMDN2s and 13 NGB2s

1 GMDN2 fighting 6 NGB2s (-1 NGB2; -1 GMDN2)

1 GMDN2 fighting 4 NGB2s (-1 NGB2)

1 GMDN2 fighting 3 NGB2s (-1 NGB2)



Round 4: 2 GMDN2 and 10 NGB2s

1 GMDN2 fighting 6 NGB2s (-1 NGB2; -1 GMDN2)

1 GMDN2 fighting 4 NGB2s (-1 NGB2)



Round 5: 1 GMDN2s and 9 NGB2s

1 GMDN2 fighting 11 NGB2s (-1 NGB2; -1 GMDN2)



This time, the 5 GMDN2s trade with 13 NGB2s. This still looks grim, but consider this: those GMDN2s are down 2CP in the battle. Those 2 CP could be a battleship (or maybe even just two little hulls carrying more Neutrino Captures), which would immediately swing the result in favour of the Dreadnoughts by boosting their absorbers and killing extra NGB2s per round.



Dreadnought fleets can defeat glass cannons.

EXPERIMENT 2: LOW TECH GLASS CANNONS



Considerations an Assumptions

Same as Experiment 1, with the following modifications.

-no tech requiring more than 1800 to unlock

-ships are level 2

-ship get +25% initial HP

-fleet size is 11CP

-no discount on shipbuilding



Unfortunately, there will be some cost distortion because I'm still using the same test game. The good news is that it's all really low tech, and costs should remain relative. Cruiser hulls will be very cheap, though. At this point, we have no access to fleet boosts of any kind, so I'll tech down the swarm. It's worth noting up front that the base HP of a L2 1CP ship at this tech level is 394, and that of a L2 2CP ship (only the Cruiser is available) is 788.



Baby Glass Breaker (BGB1)

9 Synchrotron Laser

1 High Energy Couplings

Cost: 172

Comments: This ship has 315 HP due to the High Energy Couplings, and averages 321 damage, enough to kill itself in one volley.



It takes exactly 12 Active Mirroring to negate the average damage from the BGB1 fit, a defence compliment that costs 180 Industry and weighs 120 tons. CP1 ships at this tech level only get 135-138 tonnage (depending on how Quantum-Damped Circuits interacts with Advanced Containers), and the offensive loadout for instant death is 86 tons on a Destroyer, which only has about 52 tons to spare after that. With High Energy Couplings, every additional Synchrotron Laser trumps 1.4 additional Active Mirroring, so even before the defending ship runs out of tonnage, adding defences in the arms race is just screwing you out of Industry.



A more economical approach would be to add Reflective Armour, which adds 58 HP (once you factor in its actual armour bonus) per module in the first round, 46.2 in total if you use High Energy Couplings. This module actually outpaces the available weapons tech, even with High Energy Couplings. A defensive counter to the BGB1 would be a BGB1 fit with 6 Reflective Armour, and to kill it, the BGB1 needs to mount more guns...17 Synchrotron Lasers in total. Even on a Destroyer, though, that's impossible at this tech level; the maximum armament is 15 Synchrotron Lasers and High Energy Couplings, which averages 536 damage. The following ship kills such a fit while surviving the first round.



BGB2

9 Synchrotron Laser

5 Reflective Armour

1 High Energy Couplings

1 Advanced Containers

Cost: 262

Comments: This fit has 12 tonnage to spare, but it doesn't need it for anything, so it will remain empty. If we're going to be in the second round, any absorbers we add to this ship count twice as much in terms of damage

we don't need to repair on our ships after the battle. We won't add them because they won't save the ship from a 2-on-1 situation in battle.



It's worth noting that we have access to Cruisers at this tech level (remember, starting HP of 788), so let's see if they're effective at stopping equal CPs of BGB2s, a pair of which deals 643 damage on average. The Cruiser only fires at one BGB2 per round, so it only needs to be able to kill one, and the rest of the tonnage is devoted to survival. At this tech level, the Cruiser has ~248 tonnage capacity.



Glass Melter Cruiser (GMC1)

18 Synchrotron Laser

10 Reactive Hulls

1 Predictive Plating

1 High Energy Couplings

1 Advanced Containers

Cost: 413 (582)

Comments: The cost I present for this Cruiser, 413, is artificially lowered by my test game's monoply on Titanium 70 and Hexaferrum. If we were to add the discounted 169 from the ship's hull, Reactive Hulls, and Advanced Containers, this Cruiser is still a good counter to glass cannon Destroyers in the early game. Has an effective HP of 1150 in the first round and can kill a BGB2 with 6 Reflective Armour (remember, it must kill a BGB each round). Not only that, but it survives round 2 after killing the second BGB2. It will only die if targeted by three BGB2s, and even then it will kill two of them, nearly trading at only a 58 Industry disadvantage, which is a small price to pay for average losses of zero GM1s. Anyone who starts building these should gain the upper hand over a glass cannon Destroyer fleet at the 1800 tech level.



Cruiser fleets can defeat glass cannons.

Reserved for Updates

Reserved for Updates

Reserved for Updates

Nice update, however your figures still seem to disagree with the results of my ingame testing.



I fought 22 NGC1s with 11 BS with 5200 shield and neutrino capture, the BS took 0 dmage before killing the destroyers.

5200*11*1.88 = 107536 shield

Your data suggests 178364 damage from the NGCs but this doesn't appear to be the case - in practice I think either shields are higher than stated in beta or weapons are lower. I believe the most likely explanation is that defences have had a 2* multiplier added in the beta.



When I fought 9 BS against 22 NGC1s 4 were killed - this would be the 4 that were attacked by 3 destroyers each while the other 5 were attacked by 2 each.

178364/22 = 8107 damage per NGC1

107536/11 = 9776 shield per BS

add multiplier = 19552 shield

2 NGC1 = 16214 = no damage done

3 NGC1 = 24321 = 4769 damage = killed ship





also not sure why you test against dreads as I would assume 11 BS (with -20% armour tonnage) or 22 shield destroyers ( for 22 stacking +8 def) would give more overall defence than dreads as an anti glass cannon build.

FluffyBunny wrote:
Nice update, however your figures still seem to disagree with the results of my ingame testing.

From a scientific perspective, this is fantastic news. Theoretical models are much more instructive when they are found wanting.

FluffyBunny wrote:
also not sure why you test against dreads as I would assume 11 BS (with -20% armour tonnage) or 22 shield destroyers ( for 22 stacking +8 def) would give more overall defence than dreads as an anti glass cannon build.

I tested against the modified dreadnought fits from Experiment 1 for continuity reasons; although I agree that battleships are a much better counter to glass cannons than dreadnoughts, I wanted to prove that even the ships with a poor reputation in combat can hold their own against the glass cannon fad.

Well if my 2*multplier idea is correct you could go 5 absorbtion dread and 1 absorb BS

Dread shield 10750*1.48*2 = 31820

vs 4 NGC1 = 8107*4 = 32428 = 608 damage taken

BS = 5200*1.48*2 = 15392

vs 2 NGC1 = 16214 = 822 damage taken



Providing they kill one destroyer per volley each they would survive but take some damage in the process.



5 absorb dread and two neutrino 1 CP ships would allow the dreads to take 0 damage, not sure of the shields on the 1cp ships so they might die.

That said, it is uncertain if that 2x defense effect is a bug right now.



I don't have time to fully respond to this post, but it looks interesting. I should be able to get to it in about 10-12 hours.



Kreios, when you have time I would be interested in what you think of this combat system modification.

/#/endless-space/forum/28-game-design/thread/11229-discussion-ship-designing-and-an-aggravating-equilibrium

FluffyBunny wrote:
Well if my 2*multplier idea is correct you could go 5 absorbtion dread and 1 absorb BS

Dread shield 10750*1.48*2 = 31820

vs 4 NGC1 = 8107*4 = 32428 = 608 damage taken

BS = 5200*1.48*2 = 15392

vs 2 NGC1 = 16214 = 822 damage taken



Providing they kill one destroyer per volley each they would survive but take some damage in the process.



5 absorb dread and two neutrino 1 CP ships would allow the dreads to take 0 damage, not sure of the shields on the 1cp ships so they might die.

Another possibility is that the damage multipliers from Neutrino Pulsion apply after the absorbers reduce incoming damage. Compare the following two cases for 1 Gluon Disruptor assisted by 22 Neutrino Pulsion vs. 1 Ablative Wave Shield (includes 85% accuracy):



Case 1: Damage Multiplier --> Absorber

0.85 x 210 x 2.64 - 250 = 221 damage



Case 2: Absorber --> Damage Multiplier

[(0.85x210)-250] x 2.64 = 0 damage



I have been assuming that the game rules use Case 1, but the results you're seeing are consistent with Case 2. This also makes my current interpretation of critical hits less certain; since I assume that they function like Case 2, it doesn't make sense that there would be a linguistic distinction between two functionally identical mechanics.



I wonder if the damage and defence stats summarized in the in-game battle reports can shed some light on the issue.

The specific numbers in this guide are out-of-date, although I believe the methodology is still useful. I will update this guide fairly soon, but for the moment, please use it only as a means to help you evaluate efficient fleet composition, rather than look for some magic ship fitting that will solve all of your woes.





Introduction

The primary purpose of this guide is to illustrate how analyzing fleet composition from an evolutionary perspective can help you adapt to a variety of conditions in your games. This guide is a living document examining how fleet composition and ship fittings evolve. Fleet composition is fluid, so what you find in this guide today will not necessarily be what you find in it tomorrow.



I would not have gotten to this point in my analysis without the people who commented on earlier versions of this article, so many thanks to everyone involved. Keep the feedback coming!



Experiments

The following are two experiments, one a max tech and the other at low tech, each which shows the evolution and equilibrium of glass cannon fleet composition at their respective levels, and how glass cannons are not the all-powerful solution many people, myself included, thought they were.



EXPERIMENT 1: MAXIMUM TECH GLASS CANNONS



Conditions and Assumptions

-tech level is maxed out

-ships are level 4 (highest level upon construction)

-ships all benefit from +74% initial HP (all relevant buildings constructed)

-ship HP was optimized during fitting where appropriate, then checked upon construction

-ships are built at a 20% discount due to Self-Organizing Cities

-22 Command Points per fleet

-all beams are assumed to have 85% accuracy at Long Range

-all kinetic weapons are assumed to by 70% accurate at long range (this is my own arbitrary figure until either I find the time to test this myself or get reliable test results from someone else)

-all weapons do average damage

-final damage calculations per ship and per fleet reflect the assumed accuracy and average damage

-each round, damage will be distributed evenly onto the opposing ships

-no bonuses except from ship fittings (no cards, heroes, race, etc.)

-all decimals in final calculations rounded to the nearest integer

-all costs presented in this post are obtained from the actual ship design screen of a 200-turn game, which my Amoeba have researched the entire tech tree and have a monopoly on all strategic resources; the most pronounced effect is that Defensive Lensing costs 46.4 and the trio of Strange Warheads/Disruptor Beams/Non-Baryonic Warheads all cost 23.2



Regarding Critical Hit Calculation

I oversimplified the first few versions of these experiments, assuming that multiplying bonuses to critical hit % and damage was valid in all cases. It is not. It is only valid when the target has undefended HP. For this version, I have simplified my fleet structure for the glass cannons. I dropped the Extreme Fusion modules entirely, using only Neutrino Pulsion for raw damage boost. As a result, all of the names and abbreviations of the ship fittings have changed, as has a considerable amount of calculation.



Weapon Evolution

The removal of the critical hit interactions prompted the re-introduction of kinetic weapons as a selective pressure for defence development. Under previous critical hit assumptions, the beam weapons forced defence development of a much greater pace due to interactions between number of projectiles fired (and deflected) and damage multipliers. That disparity is still there at a lesser ratio than before (2:1 instead of 5:1), so I left it as part of the analysis.



Also, note that I will be adding higher tech weapons early in the selection process because the goal is to drive the opposing fits toward equilibrium quickly. Fine tuning is more likely near the equilibrium position.





Phase 1: Hyperbole

Extreme examples, while unrealistic in practice, provide valuable framing of an issue. The following is my attempt to craft the most polarized contrast between the offensive and defensive fleets in Endless Space.



Attackers

Love 'em or hate 'em, 'glass cannon' Destroyers have a monopoly on fleet composition. I have presented two fits below, which are disturbingly synergistic. If you put 8 Neutrino Pulsions and 14 Extreme Fusions in the same fleet, the fleet averages 580.16% of its base damage. However, this only occurs against undefended HP, and the focus will be on the NGC line for the remainder of this guide, since defences have triple effectiveness against the 8N/14E fleet composition.



Neutrino Glass Cannon (NGC1)

17 Gluon Disruptor

1 Smart Cargo

1 Neutrino Pulsion

Base Damage: 3570

HP: 522

Cost: 1170



Fusion Glass Cannon (FGC1)

17 Gluon Disruptor

1 Phased Plasmic

1 Smart Cargo

1 Extreme Fusion

Base Damage: 3637.5

HP: 522

Cost: 1174



Defenders

For the targets, I propose the toughest ships per CP in the game. Here are two Dreadnoughts, each fit for a different defence strategy under the current mechanics. The first uses the beam defences, and the second uses HP bloat. Each Dreadnought has used ALL of its tonnage for its respective defences.



Absorption Dreadnought (ADN)

43 Ablative Wave Shield

1 Neutrino Capture

1 Densified Plating

1 Smart Cargo

HP: 2359 (2274 + 85 armour)

Damage Reduction: 10750

Damage Reduction in Fleet: 16770 (based on 7 Neutrino Capture, see below)

Cost: 3045



Brick Dreadnought (BDN)

30 Defensive Lensing

21 Reactive Hulls

1 Smart Cargo

HP: 26520 (23637 HP + 2883 armour)

Damage Reduction: 0

Cost: 1597



Simulation

A quick glance at the above fits and their stats reveals that the math is completely unnecessary to determine the outcome of a battle between 22 NGC1s and a fleet of either Dreadnought fit. The Dreadnought fleet is vaporized by the first volley. For the sake of completeness, let's prove it. To give the Dreadnoughts that much more of a boost, the ADNs are assumed to benefit from 2 additional Neutrino Captures (placed on a pair of 1CP ships of negligible cost), and the BDNs get an extra half Dreadnought to fill to 22CP.



Attack Fleet: Glass Cannon Destroyers

Composition: 22 NGC

Damage: 178364 ([8x3570] + [14x3637.5]) x 2.64 x .85



In a single volley, this fleet can kill:

-9.32 Absorption Dreadnoughts (~37 CP)

-6.73 Brick Dreadnoughts (~27 CP)



Notice that even without the fleet bonuses to damage and critical hits, the Glass Cannon fleet still destroys ~3.5 Absorption Dreadnoughts, and destroys ~2.5 Brick Dreadnoughts in its opening volley.





Phase 2: Evolution of the Arms Race

That NGC fleet looks pretty amazing, but it still dies to one volley from 22 of these in a fleet:



Glass Breaker (NGB1)

1 Gluon Disruptor

1 AGN Slugs

1 Neutrino Pulsion

Cost: 176

Comments: The odd thing about this ship is that the base hull is irrelevant because it does not utilize all of its tonnage. This trend continues throughout its evolution.



The first task of the Glass Cannon fleet evolution is to achieve cost parity with their most convenient counter. Replacing some Gluon Disruptors with absorbers and deflectors should do the trick. Missiles will never be an effective counter because their ships die to the Glass Cannons before the missiles launch. Since the Neutrino Pulsion fits are the prevalent variants in the fleet composition I propose, I will only track their evolution from now on.



NGC2

12 Gluon Disruptor

2 Impenetrable Hulls

2 Ablative Wave Shield

1 Smart Cargo

1 Neutrino Pulsion

Cost: 1110

Comments: The absorbers negate 100% of damage from NGB1s. The reason for this strong emphasis on defence is that the NGC2 is much more expensive than the NGB1. This simulation is being overly generous to the NGC2 in assuming that opposing ships pair off against each other. The NGC2 is more than five times the cost of the NGB1, so it only 4 NGB1s need to focus fire (with another 4, creating pairs) for this fight to break even economically. Those are bad odds for the NGC2 fit. In fact, the NGC line is going to immediately evolve again to require a 3-on-1 scenario for the first round destruction of an NGC2 against NGB1s.



NGC3

8 Gluon Disruptor

3 Impenetrable Hulls

4 Ablative Wave Shield

1 Smart Cargo

1 Neutrino Pulsion

Cost: 1050

Comments: Quite tough, but still very dangerous, the NGC3 has powerful defences against the two damage types that threaten it.



To destroy an NGC3 in a single volley, one needs to average or either:

522 + 1000 beam

[#projectiles-120] x projectile damage > 522 kinetic



NGB2

4 Gluon Disruptor

1 Neutrino Pulsion

Cost: 294

Comments: When your opponent diversifies their defences, focus on one of them to gain economic advantage. In this case, the beam defences are stacking up at a faster rate, so let's increase pressure on them while not wasting tonnage on guns that are lagging falling behind in the arms race.



NGC4

8 Gluon Disruptor

7 Ablative Wave Shield

1 Smart Cargo

1 Neutrino Pulsion

Cost: 1050

Comments: The NGC4 stops 100% beam damage from the NGB2, and it retains its damage loadout. However, it's still almost four times as expensive as the NGB2, so here we'll see another surge toward more absorbers to for a 3-to-1 requirement on first round casualties.



NGC5

1 Gluon Disruptor

13 Ablative Wave Shield

1 Smart Cargo

1 Neutrino Pulsion

Cost: 990

Comments: This is not a viable build. Notwithstanding the fact that it's still three times the cost of an NGB2, it can't even kill one! The NGC line needs to be re-envisioned as a counter to the NGB line.



NGC5

1 Gluon Disruptor

1 AGN Slugs

1 Hard Kinetics

1 Neutrino Pulsion

Cost: 176

Comments: This is identical to the original NGB1 build, so we've gone full circle.

Ketobor wrote:
That said, it is uncertain if that 2x defense effect is a bug right now.



I don't have time to fully respond to this post, but it looks interesting. I should be able to get to it in about 10-12 hours.



Kreios, when you have time I would be interested in what you think of this combat system modification.

/#/endless-space/forum/28-game-design/thread/11229-discussion-ship-designing-and-an-aggravating-equilibrium

I like where you're going with most of those ideas. However, my investigation of the current combat mechanics has prompted me to put my own proposals on hold until I have a clear picture of how the current system plays out in full.

The game gives max and min damage for beam weapons. It does not provide the average damage. Assuming that mean=median, as I did in another analysis, may be folly for such statistical analysis.



The fact is beam weapons may do min damage every 9 out of 10 shots and you would never know it because the feedback is just not there. You can only see if your ships have taken damage but only qualitative observations can be made using a fleet-wide HP bar.



It is probably better to do an analysis using kinetic weapons that have min=max damage. This way you can quantify %accuracy for the attackers to remain effective.



Actually, now that I think about it, it is probably best to include a % of max-damage-per-weapon for the attackers to remain effective.

supersoaker9 wrote:
The game gives max and min damage for beam weapons. It does not provide the average damage. Assuming that mean=median, as I did in another analysis, may be folly for such statistical analysis.



The fact is beam weapons may do min damage every 9 out of 10 shots and you would never know it because the feedback is just not there. You can only see if your ships have taken damage but only qualitative observations can be made using a fleet-wide HP bar.



It is probably better to do an analysis using kinetic weapons that have min=max damage. This way you can quantify %accuracy for the attackers to remain effective.



Actually, now that I think about it, it is probably best to include a % of max-damage-per-weapon for the attackers to remain effective.

I agree that using averages is an oversimplification, but the impact is minimal due to the specific examples involved. The minimum damage of a Gluon Disruptor is 95% of the average damage, and the disparity is unlikely to affect the analysis. I could always use combinatorics as a much better model for predicting how many shots are likely to hit, but the work involved relative to the slight increase in accuracy is huge.



One of the things I'm really interested in right now is getting the relative accuracies for beams and (especially) kinetics nailed down at all ranges, which would make all of the calculations much less tenous.