Ship Damage - some empiral data, and observations

Distant Worlds is a vast, pausable real-time, 4X space strategy game which models a "living galaxy" with incredible options for replayability and customizability. Experience the full depth and detail of large turn-based strategy games, but with the simplicity and ease of real-time, and on the scale of a massively-multiplayer online game. Now greatly enhanced with the new Universe release, which includes all four previous releases as well as the new Universe expansion!

Moderators: elliotg, Icemania

User avatar
Osito
Posts: 878
Joined: Thu May 09, 2013 8:55 am

Ship Damage - some empiral data, and observations

Post by Osito »

Edit: My own post below is now redundant, as it is clearly not a fair description of how armour is supposed to work. See Erik's post a little below, to see a good description of the way armour works. Hopefully Erik's post might be extracted and put in a sticky somewhere.

There are a few other threads relating to the issues I want to discuss, but since I’ve taken a particular approach to the subject, I have started a new thread. Hope this doesn’t end up too confusing.

I recently realised that you can use the planet editor to set up fights between ships of your own design, and you can check the damage that's being caused. So, I’ve been doing some experiments with armour, shields and weapons to see if I can figure out how it works (I haven’t seen a definitive explanation elsewhere). My conclusions are based on the evidence I’ve seen in my own tests. I freely admit my tests could be inadequate, or my conclusions could be flawed, so I’m not trying to say that I have any definitive answers. Others could repeat what I’ve done to verify my findings (or otherwise).

Note that I’ve only tested beam, projectile, missile and torpedo weapons. I haven’t tested area weapons, or gravitic weapons.

The available armours are:
Standard – rating 10; reactive rating 2
Enhanced – rating 18; reactive rating 4
Reactive – rating 25; reactive rating 7
UltraDense – rating 40; reactive rating 10

I had thought that UltraDense armour would be 4 times as effective as standard. But it’s not. I’d also thought that the reactive rating indicated the level of damage that would be ignored before any damage is done. But it’s not.

What I observed was this:

(1) Every single unit of armour you add to a ship counts as one extra ship component. If you add 100 units of armour, it counts as 100 components. By "unit" I mean a single piece with a size of 1.

(2) When an unshielded ship is hit by a weapon, at least one component will be damaged by each attack. Yes, every single time, regardless of the weapon or the armour level. Every single weapon in the game is capable of damaging at least one ship component (where the term “ship component” includes armour) every time it hits a ship. Furthermore, to all intents and purposes, a “damaged” component seems to be synonymous with a “destroyed” component. A single pulse blaster can take down a ship armoured with UltraDense armour. Every time the pulse blaster hits a unit of UltraDense armour, it will damage/destroy the unit it hit. Given that pulse blasters hit for 4, while UltraDense armour has a reactive rating of 10, this indicates that the reactive rating does not ignore the damage of any weapon with a damage below 10. (Interestingly, the Galactopedia says "absorb"; the word "ignore" is mine, and the difference may be significant.)

(3) The more units of armour you have, the more likely it is that the armour will be hit instead of another ship component. You have 1000 units of armour and 10 other components? Then your other components only have a 1% chance of being hit. You have 1 unit of armour and 9 other components? Then your other components have a whopping 90% chance of being hit! That’s why you often find ship components being damaged before armour.

(4) Sometimes weapons will damage more than one component when they hit. The likelihood of this seems to depend on the weapon damage: higher damage weapons have a greater chance of damaging more than one component. Furthermore, higher level armour will reduce the chance that more than one component will be damaged.

These findings suggest that it is virtually useless to add just one or two units of armour. They also suggest that armour is likely to be most useful on big ships where you can place a lot of armour, or on planetary bases where you can place an almost unlimited amount of armour.

I also found that UltraDense armour is nowhere near four times as effective as Standard. In practice, I found that an unshielded ship armoured in UltraDense armour would survive about 1.5-2 times as long as an equivalent ship armoured in Standard armour (i.e. with the same number of units of armour). The armour value you see in ship design seems to be meaningless. According to the ship design screen, 400 units of Standard armour has the same strength as 100 units of UltraDense armour (4000 armour each), but you put these two ships in a fight and the UltraDense ship might be dead in half the time it takes to kill the Standard one.

I’m afraid I was unable to discern the effect of the reactive rating other than to confirm that weapons with lower damage than the rating do get through. I have a theory that reactive rating relates to the amount of damage the armour can take before another ship component is damaged. So if you damage armour with a reactive rating of 10 (UltraDense) with a weapon of rating 4 (pulse blaster), the pulse blaster will never damage a second component in one hit. But if the weapon damage is above 10, then it can damage a second component. This theory matches what I observed but I'm unsure of it. If it is true, then it makes me wonder whether the armour value itself has any effect at all.


*********

I also looked at the effect of rail guns, as there has been a lot of discussion of those recently. The in-game notes say the following about rail guns:

(1) They have a -10% targeting handicap

(2) The can partially penetrate shields

(3) They are 50% less effective against armour.

You would expect from (1) and (3) that rail guns would take longer to kill an armoured target than an equivalent strength beam weapon. I tested both a Maxos Blaster and Rail Gun, both upgraded to strength 8. I tested at close range to reduce/eliminate the effect of distance on the Blaster. I found the two weapons took about the same time to kill an identical ship with 30 units of standard armour. The Rail gun was quicker on UltraDense armour – didn't understand that.

I also tested the speed a single Rail Gun could take down shielded and unshielded ships. The unshielded ship had 30 units of standard armour, no shields, 38 components in total. The shielded ship was the same as the unshielded, but had three Corvidan shields, i.e., 41 components altogether. My tests suggested that the rail guns were getting through the shields something like 80-90% of the time. Haven't tested whether more shields would have an effect.

So, on the subject of rail guns, I would say three things:

(i) It’s possible they are more effective against armour than stated.

(ii) It’s also possible that they are too effective at piercing shields.

(iii) Even if (i) and (ii) don’t apply, the fact that armour is relatively ineffective on smaller ships makes it inevitable that Rail Guns are going to give them a hard time, until you can get a large enough ship size to coat it with a decent helping of armour. So I think my findings support the views of others that the present game design strongly favours an early game bias towards rail guns.

Osito
Osito
Bingeling
Posts: 5186
Joined: Thu Aug 12, 2010 11:42 am

RE: Ship Damage - some empiral data, and observations

Post by Bingeling »

Does a beam damage other components before all armor is gone? That does not fit my memory, but may explain some excessive component damage from all kinds of pirates in Shadows. I don't think all of them use gravitic weapons.

When a cruiser (with some armor) has 3 damaged, it is 3 armor damaged. It is not 50% (or something) probability for each point being armor damage. Of course, gravitic weapons ignore armor, so this is no longer always true.

If this was not true, I should have seen ships taking a few hits to components (inclusive armor), that lose their hyperdrive. This did not happen, the ones "stuck" were always badly damaged with lots of damaged components. If there was a weapon that could bypass armor, the "3 damaged components and stuck" should have happened before Shadows.
Cauldyth
Posts: 932
Joined: Sun Jun 27, 2010 1:24 am

RE: Ship Damage - some empiral data, and observations

Post by Cauldyth »

Iiiiiiiinteresting. Thanks for the awesome data. Hopefully Erik gets back to us soon with a report on

1) How armour is designed to work
2) Whether or not there are bugs in the current implementation of that design

So your conclusion is that armour is essentially a decoy component. You want more units of armour to increase the odds of it working against any given blast, and you want stronger armour tech to (perhaps) guarantee that the individual piece of armour that gets hit can bear the entire hit and not have it leak through to another component.
User avatar
Larsenex
Posts: 476
Joined: Fri Dec 31, 2010 1:14 pm

RE: Ship Damage - some empiral data, and observations

Post by Larsenex »

Actually Cauldyth that would not be true. Above tests show zero value in researching ultra dense armor. Instead your best bang is basic armor and gobs of it piled up on your ships. I FAIL to see the value of putting any on your ship at all.

If the size of armor scaled below 1 on the higher tiers that would be worth it. For example if Standard were a value of 1 and enhanced was a value of .50, and Ultra being a value of .0025. Than it would be worth it to add gobs of it on the ship. Otherwise for 10 slots of armor I would much rather have a size 10 shield.
Go for the Eyes Boo!

Intel 8700K Oc'd to 4.8ghz
32 gigs ram
GTX 1070 w/ 6gigs ram.
Using a cache drive from intel with a 60gig flash & 1 terrabyt hd accelerated.
User avatar
Osito
Posts: 878
Joined: Thu May 09, 2013 8:55 am

RE: Ship Damage - some empiral data, and observations

Post by Osito »

Bingeling, it's definitely true that beam weapons can damage other components before all the armor is gone. How soon that happens depends on how many units of armour are on the ship. If a ship has 3 damaged components, it's certainly possible (likely, actually) that some of the damaged components are not armour. I saw this over and over again in the testing. Your point about the hyperdrive is correct, in practice. Ships only have one hyperdrive component. So if a ship has 100 components, only one of those is the hyperdrive. Once the ship components start to get attacked, it's initially only a 1% chance that the hyperdrive will get hit. As other components get destroyed, the probability goes up (e.g., when there are only 50 components left, the hyperdrive has a 2% chance of being the next one to go). That's my theory, anyway, and it is supported by the data I was seeing in the tests. Of course, I'm not sure there isn't a weighting in favour of hitting armour. I didn't see one, but frankly I didn't do enough testing to detect one.

Cauldyth, I think you have summarised (my understanding of) the position correctly.

Larsenex, that's not quite right. There is a theoretical value in researching armours after Standard, as they reduce the chance that more than one component is damaged with each shot. In practice, think UltraDense armour is probably somewhere between 1.5-2.5 times as effective as Standard armour. Also, I believe the effectiveness was better against higher level weapons. But I take your point about whether it's worth putting armour on a ship. I think, on a small to medium ship, the amount of armour you can spare is so small that it would hardly last any time against the enemy's weapons. I'm not sure about the larger ships, though. If you can spare several hundred units of armour, it might make a difference, but I haven't really tested it (and I'm not sure whether I could spare the space).

Osito
Osito
invaderzim
Posts: 211
Joined: Mon Jun 03, 2013 10:48 am

RE: Ship Damage - some empiral data, and observations

Post by invaderzim »

These findings suggest that it is virtually useless to add just one or two units of armour. They also suggest that armour is likely to be most useful on big ships where you can place a lot of armour, or on planetary bases where you can place an almost unlimited amount of armour.

This sounds about right. When I'm killing a spaceport, I'll notice there is a period of time after the shields go down before the firepower starts to go down. After the firepower starts to go down, it drops at a steady pace, suggesting that most of the armor has been removed.

Also this suggests that ion weapons can be more effective than normal against spaceports since they should only ignore all components except weapons and engines.
User avatar
Erik Rutins
Posts: 39324
Joined: Tue Mar 28, 2000 4:00 pm
Location: Vermont, USA
Contact:

RE: Ship Damage - some empiral data, and observations

Post by Erik Rutins »

Hi everyone,

Here's how armor and ship damage are _intended_ to work. From the results above, if we can validate them, we have some investigating to do to make sure it's working as intended.

Armor is struck by weapons fire when shields are fully depleted, or when the weapon is a rail gun and it bypasses shields. When the weapon is a gravity beam or gravity area weapon (these gravity weapons completely bypass both shields and armor)

Phaser weapons are also better at penetrating armor: the reactive rating is effectively halved when a phaser strikes it, thus allowing easier destruction of the armor.

Armor protects other components from damage. The only instance where other components should be damaged before armor is in the case of a hit from a gravitic weapon, which bypasses armor. Gravitic weapons damage components randomly once they hit and they can damage armor as well as normal components. No other weapon types completely bypass armor. By design, all other weapon types must first destroy armor before any damage occurs on other components.

When weapons fire strikes Armor there may be damage to the armor when the remaining weapon strength exceeds the Reactive Rating of the Armor component, i.e. the weapon is powerful enough to penetrate the reactive rating of the armor.

If the remaining weapon strength is less than or equal to the Armor reactive rating, then there is a small chance that the armor will still be breached. The actual chance depends on the ratio of remaining weapons strength to the reactive rating of the armor, but is never more than 20%.

If the reactive rating of the armor is penetrated then there is a random chance that the armor component will be destroyed (i.e. damaged component). The actual chance depends on the ratio of the standard Rating of the armor component to the remaining weapons damage amount, but is never less than 10%.

The armor damage is in two phases:

1. Penetrating the Reactive rating of the armor (max 20% chance for weapons with insufficient power)
2. Actually damaging the armor component (always at least 10% chance)

Some equations:

For Phasers: Damage = (Remaining Phaser strength at time of impact - (Armor Reactive rating / 2))
For other weapons: Damage = (Remaining weapon strength at time of impact - Armor Reactive rating)
BUT when Armor Reactive rating > Remaining weapon strength at time of impact THEN:
chance of penetrating Armor Reactive rating = (Armor Reactive rating / Remaining weapon strength at time of impact) * random factor (with Maximum chance of 20% when Remaining weapon strength at time of impact = Armor Reactive rating)
AND in the above case remaining weapon damage that passes to phase 2 below is ALWAYS 1 (i.e. minimal damage)
Actual chance to destroy armor = Max(0.1, (Damage / Armor standard Rating))

Using the above equations against Standard Armor (level 1):

Level 1 Maxos Blaster from range 100:

strikes target with remaining strength of 4, Armor Reactive rating (2) reduces damage to 2
chance to destroy armor component is 20% (Armor Standard Rating = 10, damage of 2 divided by 10)

Level 1 Rail Gun from range 100:

strikes target with remaining strength of 6 (no distance losses), Armor Reactive rating (2) reduces damage to 4
rail guns (and missiles) have less chance of destroying armor (50% reduction), thus actual damage is 2
thus chance to destroy armor component is 20% (Armor Standard Rating = 10, damage of 2 divided by 10)

Level 3 Impact Assault Blaster from range 100:

strikes target with remaining strength of 9 (3 distance loss), Armor Reactive rating (2) reduces damage to 7
chance to destroy armor component is 70% (Armor Standard Rating = 10, damage of 7 divided by 10)

Level 3 Shockwave Torpedo from range 100:

strikes target with remaining strength of 20 (4 distance loss), Armor Reactive rating (2) reduces damage to 18
chance to destroy armor component is 100% (Armor Standard Rating = 10, damage of 18 divided by 10)

Keep in mind Rail guns are also less accurate and much shorter ranged.

While these values are unchanged since Legends, in Shadows we have improved the random number generator to be more random. Thus this may have also increased the chance of reactive armor penetration and/or armor destruction.

Note that there are also some special cases for Armor: Shandar have stronger armor (including reactive rating) at their spaceports (racial bonus). Also, rail guns and missiles inflict less damage on armor (i.e. less chance to destroy once penetrate reactive rating).

Against rail guns, armor, speed/maneuver (as they are short range) and ECM are the intended counters. Ships that do not have enough armor, are not faster/more maneuverable than the rail gun ship and don't have an ECM advantage should not do well.

Against gravitic weapons, the intended counters were ECM, speed/maneuver (as while they are not short range, missiles and torpedos significantly outrange them), larger ships and repair bots.

What about when firing at unarmored ships, or ships which have had their armor destroyed?

For normal components, the remaining damage is reduced by the SIZE of each destroyed component, until the damage is exhausted (i.e. single shot can damage multiple components). Damage that hits normal components (NOT armor components) is also reduced by the damage reduction percentage first, then the remaining damage amount destroys components.

Regards,

- Erik
Erik Rutins
CEO, Matrix Games LLC


Image

For official support, please use our Help Desk: http://www.matrixgames.com/helpdesk/

Freedom is not Free.
User avatar
DeadlyShoe
Posts: 217
Joined: Sat Jun 01, 2013 10:15 pm

RE: Ship Damage - some empiral data, and observations

Post by DeadlyShoe »

I feel that's how it worked in Legends - Shakturi were nearly impossible to hull damage with Shatterforce, missiles, railguns, or fighters.

but my non-empirical weather eye testing believes that Osito is correct that 1 weapon hit = 1 component damaged right now.
Against gravitic weapons, the intended counters were ECM, speed/maneuver (as while they are not short range, missiles and torpedos significantly outrange them), larger ships and repair bots.
If we're talking about general weapon balance, I feel I need to bring up the AI problem with gravitics: an escort will force a capital ship to retreat with a single hit, because component damage = retreat... :|

Outside this retreat problem gravitics feel fairly weak

I do try to rush repair bots but on 200k tech price I only barely got them before the Eruktah showed up. Pretty late game. In practical terms it's alot easier to bribe Ikkuro for their repair bot. xD
User avatar
Erik Rutins
Posts: 39324
Joined: Tue Mar 28, 2000 4:00 pm
Location: Vermont, USA
Contact:

RE: Ship Damage - some empiral data, and observations

Post by Erik Rutins »

One thing to add to my above post, in the case of the Shockwave Torpedo example, I forgot to explain how the additional damage works. The remaining torpedo damage after you reach 100% of the Standard Armor Rating will carry over and destroy the next armor component, or if no undamaged armor remains then will destroy a normal component.
Erik Rutins
CEO, Matrix Games LLC


Image

For official support, please use our Help Desk: http://www.matrixgames.com/helpdesk/

Freedom is not Free.
User avatar
Erik Rutins
Posts: 39324
Joined: Tue Mar 28, 2000 4:00 pm
Location: Vermont, USA
Contact:

RE: Ship Damage - some empiral data, and observations

Post by Erik Rutins »

Osito's testing concerns me. A save file here would help a lot so that we can reproduce these results and examine what's going on in the game.
(2) When an unshielded ship is hit by a weapon, at least one component will be damaged by each attack. Yes, every single time, regardless of the weapon or the armour level. Every single weapon in the game is capable of damaging at least one ship component (where the term “ship component” includes armour) every time it hits a ship. Furthermore, to all intents and purposes, a “damaged” component seems to be synonymous with a “destroyed” component. A single pulse blaster can take down a ship armoured with UltraDense armour. Every time the pulse blaster hits a unit of UltraDense armour, it will damage/destroy the unit it hit. Given that pulse blasters hit for 4, while UltraDense armour has a reactive rating of 10, this indicates that the reactive rating does not ignore the damage of any weapon with a damage below 10. (Interestingly, the Galactopedia says "absorb"; the word "ignore" is mine, and the difference may be significant.)

This should not be the case, by design.
(3) The more units of armour you have, the more likely it is that the armour will be hit instead of another ship component. You have 1000 units of armour and 10 other components? Then your other components only have a 1% chance of being hit. You have 1 unit of armour and 9 other components? Then your other components have a whopping 90% chance of being hit! That’s why you often find ship components being damaged before armour.

Same with this - armor should be hit before other components. The only weapons where it should work like this are gravitic weapons.
Erik Rutins
CEO, Matrix Games LLC


Image

For official support, please use our Help Desk: http://www.matrixgames.com/helpdesk/

Freedom is not Free.
Cauldyth
Posts: 932
Joined: Sun Jun 27, 2010 1:24 am

RE: Ship Damage - some empiral data, and observations

Post by Cauldyth »

Thanks for the detailed explanation Erik. Is it possible that all weapons accidentally inherited the gravitic weapons' ability to randomly destroy components? Osito's analysis, as well as others posted in the other thread, suggest that may be the case. All weapons seems to be randomly destroying components, sometimes armour, but sometimes other components when there is still armour present.
invaderzim
Posts: 211
Joined: Mon Jun 03, 2013 10:48 am

RE: Ship Damage - some empiral data, and observations

Post by invaderzim »

If we're talking about general weapon balance, I feel I need to bring up the AI problem with gravitics: an escort will force a capital ship to retreat with a single hit, because component damage = retreat... :|

Yeah I've experienced this auto-retreat on many occasions. I really wish there was a way to disable all automatic behaviors of a specific fleet (auto-retreat, auto-refueling, auto-regrouping, etc.) and take full control over ships during important battles.
Cauldyth
Posts: 932
Joined: Sun Jun 27, 2010 1:24 am

RE: Ship Damage - some empiral data, and observations

Post by Cauldyth »

Whoops, guess Erik was responding at the same time I was typing my post. [:)]
Cauldyth
Posts: 932
Joined: Sun Jun 27, 2010 1:24 am

RE: Ship Damage - some empiral data, and observations

Post by Cauldyth »

Okay, I managed to put together an example that seems to show the problem. At least I think it does, I'm not entirely sure how to examine the ships more closely to see what's going on. I've put the savegame here:

https://dl.dropboxusercontent.com/u/897 ... -05-03.dwg

Steps to reproduce

1) Once the game starts, pause immediately.
2) Find the cruiser Arcadia. I've stripped it down to no shields, and just a single Maxos blaster.
3) Order it to attack the pirate destroyer Fearsome Eclipse, which is right in front of it. The cruiser will keep trying to run away because it has no shields, so just keep repeatedly clicking attack to get it to close in.

You'll see the first volley or two that hits will immediately start knocking out components, even though it has 20 Ultradense Armour on it. I've done the test a few times, and in one it even lost its hyperdrive on the first hit.

Strat_84
Posts: 84
Joined: Thu Dec 08, 2011 8:35 am

RE: Ship Damage - some empiral data, and observations

Post by Strat_84 »

Thanks for the explanations. One more detail please: how are the damage control components involved in the calculations ?

Here is a save about the armor problem: http://dl.free.fr/hLxP3Q5oi
The screen should be centered on the Fleet Victory, with a dozen of ships damaged after a fight against pirates (no gravity weapon involved as far as I know). All have internal components damaged with still intact armor. Especially check the Capital ship leading this fleet, it took only a few railgun hits and it resulted in 1 hab module being destroyed, with all the armor intact.

The files are massive for my connection and Cauldyth proposed something equivalent to the Big Jim test, so I don't think it's really necessary I upload it. [;)]
User avatar
Osito
Posts: 878
Joined: Thu May 09, 2013 8:55 am

RE: Ship Damage - some empiral data, and observations

Post by Osito »

Thanks for the information, Erik, that's really helpful. I've uploaded a save file "Armour Test from Osito.dwg".

Rather than bloat the forum with another wall of text explaining what I did in detail, I've sent you an email on the matter. Hope it helps.

Osito
Osito
User avatar
Erik Rutins
Posts: 39324
Joined: Tue Mar 28, 2000 4:00 pm
Location: Vermont, USA
Contact:

RE: Ship Damage - some empiral data, and observations

Post by Erik Rutins »

ORIGINAL: Strat_84
Thanks for the explanations. One more detail please: how are the damage control components involved in the calculations ?

From my post above:

"For normal components, the remaining damage is reduced by the SIZE of each destroyed component, until the damage is exhausted (i.e. single shot can damage multiple components). Damage that hits normal components (NOT armor components) is also reduced by the damage reduction percentage first, then the remaining damage amount destroys components."

Thank you for the save files everyone!

Regards,

- Erik
Erik Rutins
CEO, Matrix Games LLC


Image

For official support, please use our Help Desk: http://www.matrixgames.com/helpdesk/

Freedom is not Free.
Strat_84
Posts: 84
Joined: Thu Dec 08, 2011 8:35 am

RE: Ship Damage - some empiral data, and observations

Post by Strat_84 »

My bad, I expected so much that component to reduce the damage on armor as well that I missed that part. [:'(]
User avatar
Erik Rutins
Posts: 39324
Joined: Tue Mar 28, 2000 4:00 pm
Location: Vermont, USA
Contact:

RE: Ship Damage - some empiral data, and observations

Post by Erik Rutins »

Yes, damage control is for when you take damage beyond your armor. It can significantly increase survivability by limiting damage to just one component for each hit through armor, but it does not affect damage to armor. Repair bots help repair armor and other damage, damage control makes your normal ship components more resilient.
Erik Rutins
CEO, Matrix Games LLC


Image

For official support, please use our Help Desk: http://www.matrixgames.com/helpdesk/

Freedom is not Free.
Spacecadet
Posts: 1784
Joined: Sun Apr 18, 2010 5:52 pm

RE: Ship Damage - some empiral data, and observations

Post by Spacecadet »

Interesting.

I always thought that it worked with somewhat of a "Hit Point" behavior.

For example:

Level 3 Impact Assault Blaster from range 100:

strikes target with remaining strength of 9 (3 distance loss), Armor Reactive rating (2) reduces damage to 7
chance to destroy armor component is 70% (Armor Standard Rating = 10, damage of 7 divided by 10)

In this case I thought the Armor would be reduced to 3 points after this this impact (9 - 2 = 7, then 10 - 7 = 3), and that the next hit would basically take out the remaining Armor value and have 4 damage left to apply to the next component.




CPU: Intel 2700K
RAM: 16 GB
GPU: GTX 970
OS: Windows 7 (64 bit)
Res: 1920 x 1200


Post Reply

Return to “Distant Worlds 1 Series”