Plunging fire and historical armour values.

[madham82]

9 minutes ago, Cpt.Hissy said:

Regarding plunging fire, nobody finds it strange that ALL guns actually have nearly identical pen curves? Aren't lighter shells supposed to loose energy faster, have steeper arching trajectories, and therefore have penetration in general drop faster and have that spot where horizontal pen overcomes vertical pen at closer ranges?
This was the main reasoning behind going bigger in earlier years as i remember. For big gun's flatter trajectory and greater distance of effective penetration.
  If i don't understand something in artillery ballistics, please correct me, but i think this one part i understand correctly...

Definitely truth in that. Their trajectories are more arcing since they don't have the energy to carry that flat trajectory as far. All that is needed is to take a couple representative calibers and their data, then plot what the rest should look like. Not asking Nick and team to recreate every caliber, just the pattern similar calibers should follow. 

[Bigjku]

1 hour ago, Cpt.Hissy said:

 A question regarding SHP caps. It's due to hydrodynamics based upper limit on size and rotation speed of the screws, right? That seems to be quite flexible actually, especially in smaller than ultrahuge ships.
Anyway, this limit isn't directly tied to SHP, so that'll be arbitrary cap, and that doesn't usually work well in games. Maybe have some alternative?

Regarding plunging fire, nobody finds it strange that ALL guns actually have nearly identical pen curves? Aren't lighter shells supposed to loose energy faster, have steeper arching trajectories, and therefore have penetration in general drop faster and have that spot where horizontal pen overcomes vertical pen at closer ranges?
This was the main reasoning behind going bigger in earlier years as i remember. For big gun's flatter trajectory and greater distance of effective penetration.
  If i don't understand something in artillery ballistics, please correct me, but i think this one part i understand correctly...

SHP caps seem to have a number of factors (material, raw ability to generate it based on boiler pressure and gear ratios and hydrodynamic on the screws) but my proposal for them in game is simply that it would force a trade off between size and speed and it would keep speeds down historically when they should be lower on heavy ships.  You could honestly achieve a similar result simply by limiting hull size more than the system seems to in historical eras I suspect.

But logically I am much more comfortable with the idea that the US could lay down in 1914 or so a ship with a displacement of 40 plus thousand tons if they wanted to than I am with them coming up with a 111,658 SHP ship using conventional turbines (not geared) and achieve a speed of 27.5 knots on 37,000 tons.  If I limit conventional turbines to about what the QE class had (75,000) I can do 24 knots on the same basic ship.  This makes a lot more logical sense in that period.  The QE was basically a battle cruiser sized power plant in a battleship armored hull and traded 4 knots of speed from battlecruisers for reasonable armoring.

When I go to design a British battlecruiser in 1913 it defaults to to 123,392 SHP on M-expansion steam engines which is just absurd.  If I use a power plant similar to say the Tiger of 85,000 SHP which was a huge power plant I can get 28 knots on 26,000 tons roughly which is realistic but runs a bit light weight wise.  If I use that weight on engines to get to 28,500 tons I end up at 33 knots in 1913 on 130,306 SHP.  That speed is too cheap for that era and I don’t think was even possible.

Renown laid down later than that is the best comp I can find.  But it had armor in the 3-6 inch range on the belt and only 6 main guns.  Basically it traded armor and number of barrels off the Tiger for yet more engines and speed.  But you were carrying 15 inch guns and not protected against 11 inch German shells since the armor was on the level of the early British battlecruisers that blew up with regularity.

I bring this back up because the focus of the campaign will be on the British and German race from what I can tell.  26 knot battleships and 33 knot battlecruisers with good protection and plenty of heavy guns being spit out by the system.  That just isn’t realistic.  It may “work” as you see absurd speeds in cruises and destroyers too.  But it will be strange to people heavily into naval history.  And that’s the target audience for the game.

I am fine letting someone have 30 knot ships but they need to have lots of compromises.  Even the 28 knot ships had compromised heavily to get there.

The game isn’t going to feel right in that period if they screw this up.

[Cpt.Hissy]

So i got carried away with this topic and tried to be a spreadsheet person (which i normally don't do)
Results are curious to say the least.

Did a comparison between easily accessible penetration tables for Iowa's guns and the in-game 16" Mk3 gun, which is apparently considered by players to be it's gamey counterpart.
Gamegun was plotted with normal and "superheavy" shells and pen data as-is, and also converted to Krupp3 (basically 190%) based on assumption that in-game values represent iron plate thickness.

Here are graphs for vertical armour penetration. Notice that game's curve is noticeably steeper than real data, and dives under real graph even with raw values, with Krupp3 conversion vastly underperforming at all ranges.


Next, the deck penetration graphs. At all times, game values vastly outperform the real thing, even with conversion. And again, the graph is unrealistically steep.

 

And for dessert, combined plots of real vs game gun. See that cross between side and deck pen graphs at just above 18km? That's actually common for ALL main guns in game, all the way from 10" up to 17". And real thing converges at about 35km.

I also did some more crude plots, there's some funny things showing up. Like, turns out all secondaries have seriously different pen curves from all mains, for example. And of course, recently added anime guns are different from all the others too.
Curious...
I don't normally annoy people, but...
@Nick Thomadis pardon me for annoyance, but could you please clear this for us somehow? Why do we see this? What exactly are these values that we see in game's tables, what's their relation to actual historical penetration data (if any)? How the game handles ballistics, if these are not what it actually uses? And of course, what's the use of those values for players, if they're not factual pens?

[Bigjku]

1 hour ago, Cpt.Hissy said:

So i got carried away with this topic and tried to be a spreadsheet person (which i normally don't do)
Results are curious to say the least.

Did a comparison between easily accessible penetration tables for Iowa's guns and the in-game 16" Mk3 gun, which is apparently considered by players to be it's gamey counterpart.
Gamegun was plotted with normal and "superheavy" shells and pen data as-is, and also converted to Krupp3 (basically 190%) based on assumption that in-game values represent iron plate thickness.

Here are graphs for vertical armour penetration. Notice that game's curve is noticeably steeper than real data, and dives under real graph even with raw values, with Krupp3 conversion vastly underperforming at all ranges.


Next, the deck penetration graphs. At all times, game values vastly outperform the real thing, even with conversion. And again, the graph is unrealistically steep.

 

And for dessert, combined plots of real vs game gun. See that cross between side and deck pen graphs at just above 18km? That's actually common for ALL main guns in game, all the way from 10" up to 17". And real thing converges at about 35km.

I also did some more crude plots, there's some funny things showing up. Like, turns out all secondaries have seriously different pen curves from all mains, for example. And of course, recently added anime guns are different from all the others too.
Curious...
I don't normally annoy people, but...
@Nick Thomadis pardon me for annoyance, but could you please clear this for us somehow? Why do we see this? What exactly are these values that we see in game's tables, what's their relation to actual historical penetration data (if any)? How the game handles ballistics, if these are not what it actually uses? And of course, what's the use of those values for players, if they're not factual pens?

Love this.  Pretty on point.  Deck hits would likely be very rare events.

http://www.navweaps.com/index_nathan/HOOD 1919 Armor Tests.pdf

This makes for interesting reading on the subject of deck hits.

[Draco]

@Cpt.Hissy
Well done!

Edited March 9, 2021 by Draco

[Stormnet]

I knew something was up when a krupp IV, citadel V battleship sporting deck armor 3 times thicker than that on the Bismark could be penned at 17 km by a small 12 inch on heavies and tube powder, but I never thought armor was so underperforming.

Edited March 9, 2021 by Stormnet

[Cptbarney]

56 minutes ago, Stormnet said:

I knew something was up when a krupp IV, citadel V battleship sporting deck armor 3 times thicker than that on the Bismark could be penned at 17 km by a small 12 inch on heavies and tube powder, but I never thought armor was so underperforming.

Lol, thats pretty hilarious. That would mean a supercruiser can body most battleships, thanks to faster aiming times, shell travel times and rate of fire.

[Cptbarney]

Also i've noticed the armour scheme seems to follow that or a pre-dread for the most part, so it shows we don't really have armour sets based on citadel types yet and only using one type atm.

Which is fine for an alpha but i want more armour types and internal armour so we can test these out before steam release.

[Nick Thomadis]

11 hours ago, Cpt.Hissy said:

@Nick Thomadis pardon me for annoyance, but could you please clear this for us somehow? Why do we see this? What exactly are these values that we see in game's tables, what's their relation to actual historical penetration data (if any)? How the game handles ballistics, if these are not what it actually uses? And of course, what's the use of those values for players, if they're not factual pens?

As I already said, the applied values are different from the estimations, because they do not include the actual angle of hit, which significantly reduces the penetration of the shell. There are also some hard coded angled armor values according to hull, which are also not included. We could improve later all this, but it is not a priority. Our priority is to provide the next hotfix update which will fix/improve all the significant and critical problems that the community has found.

If you made a statistical evaluation of how many ship hits are transformed into actual full penetration deck hits, you would get values close to reality. The so called "Immunity zones" do not work as "deflector shields" making ships impervious to any shell attack, but make less probable to inflict critical hits to a ship that is well protected on this zone.

And this is how the game works now at these ranges, making deck armor essential in any case, because without it, any hit would cause ammo detonations, engine damage and explosions that would disintegrate a ship. Would the game be more enjoyable if ships were immune, taking unrealistic zero damage by 16 inch shells? Maybe for you, but not for the majority, and this would not be realistic either.

Arguably, we could remove all the statistical estimations of the game, (I do not know many games that offer so much data) to stop this frequent disorientation that they cause.

 

[Cpt.Hissy]

@Nick Thomadis Thanks for reply.. though it doesn't really explain anything.
Mainly what's the use of these values for the player, if they don't represent anything real. If there are some additional calculations based on angles, probably easiest and most useful way to do that will be to include those calculations in the tables, assuming that "side pen" is against perfectly vertical armor at right horizontal  angle, and "deck pen" is for perfectly horizontal armor. That should give the numbers generally similar to those in IRL tables. Well, at least actually useful and not this levels of nonsense we have now.
+ as an option, values could be converted for the type of armor currently installed (and to one of the target in battle)

I understand, and hope most other people here do too, that the game is still in heavy development, everything may change and the team needs to prioritise things one at a time due to limited resources. But there are issues - like this one - that are much more disturbing for players than some of the ones actually being worked on right now, that's why so much fuss around them.

We still need to have (proper) pen tables, as they are the basis for designing our protection and for picking our battle range. One of the most important aspects of the whole thing, much more important that yen another imaginary post-dreadnought superbattleship in a realistic game about dreadnoughts. (and i believe any other game focused on realistic-ish artillery combat gives these tables as well. Even friggin war thunder)
But some other data could actually go away if it causes confusion.

Still hope your team will succeed in making this the game we all wished it to be

@Cptbarney Yep looks like actual collision box for ships is roughly their outer skin (that's not even predread level). But this is a thing that can be brought down to satisfactory lifelike behaviour by just RNG coefficients, and considering the potential of dozens of ships in single battle, and that those ships will be not carefully handcrafted models but unpredictable random mishmash, it'll probably be the best way.
Highly detailed internal armor layouts will require a lot more work than simple hull model, and the team seems to struggle even with those (so many rescaled clones), and it's almost guaranteed to not fit properly with some of the possible configurations of other parts. On top of that modeling and calculating every hit at every complicated multilayered armored box when there's a lot of them may be problematic for performance. Unity is bad at such things.
Instead there may be predefined chances for a hit on the side or the deck to count as penetration, or as upper belt hit, or as hit on deck slope, etc, with corresponding effects. A penetrating hit near a turret may have additional factor for hitting it's barbette and still doing nothing much. And so on, so on. All based on installed armour scheme etc.
That may look not very realistic with a single ship, but should give believable overall results for entire fleet over time. And it'll be easier to implement and run faster, which is not of least importance.

[Cptbarney]

3 minutes ago, Cpt.Hissy said:

 

@Cptbarney Yep looks like actual collision box for ships is roughly their outer skin (that's not even predread level). But this is a thing that can be brought down to satisfactory lifelike behaviour by just RNG coefficients, and considering the potential of dozens of ships in single battle, and that those ships will be not carefully handcrafted models but unpredictable random mishmash, it'll probably be the best way.
Highly detailed internal armor layouts will require a lot more work than simple hull model, and the team seems to struggle even with those (so many rescaled clones), and it's almost guaranteed to not fit properly with some of the possible configurations of other parts. On top of that modeling and calculating every hit at every complicated multilayered armored box when there's a lot of them may be problematic for performance. Unity is bad at such things.
Instead there may be predefined chances for a hit on the side or the deck to count as penetration, or as upper belt hit, or as hit on deck slope, etc, with corresponding effects. A penetrating hit near a turret may have additional factor for hitting it's barbette and still doing nothing much. And so on, so on. All based on installed armour scheme etc.
That may look not very realistic with a single ship, but should give believable overall results for entire fleet over time. And it'll be easier to implement and run faster, which is not of least importance.

Ye collisions boxes and be re-sized doe in unity so no need for code, Too be fair im expecting a simple internal scheme using primitives (cubes basically), and if its worth it expand from there, if they can do a warthunder or better style system then more power to everyone i guess. Also yeah, i hope we get less clones for hulls as well.

Really want to make some pretty unique ships (i mean we can sort of, but we are still locked by nation and by year group for custom battles and when you select unlock all, it just causes a mess really).

I think using primitives and then maybe going to simple low-poly models with primitives should be fine for the most part, if it doesn't cause too many performance issues.

[madham82]

1 hour ago, Nick Thomadis said:

As I already said, the applied values are different from the estimations, because they do not include the actual angle of hit, which significantly reduces the penetration of the shell. There are also some hard coded angled armor values according to hull, which are also not included. We could improve later all this, but it is not a priority. Our priority is to provide the next hotfix update which will fix/improve all the significant and critical problems that the community has found.

If you made a statistical evaluation of how many ship hits are transformed into actual full penetration deck hits, you would get values close to reality. The so called "Immunity zones" do not work as "deflector shields" making ships impervious to any shell attack, but make less probable to inflict critical hits to a ship that is well protected on this zone.

And this is how the game works now at these ranges, making deck armor essential in any case, because without it, any hit would cause ammo detonations, engine damage and explosions that would disintegrate a ship. Would the game be more enjoyable if ships were immune, taking unrealistic zero damage by 16 inch shells? Maybe for you, but not for the majority, and this would not be realistic either.

Arguably, we could remove all the statistical estimations of the game, (I do not know many games that offer so much data) to stop this frequent disorientation that they cause.

 

So it boils down to the tables aren't useful (since they do no take angle of fall into consideration), and the in game pen popup has too many quirks to be useful (and doesn't help us at all in the design stage). We need a better solution when designing ships. It would seem the game has some kind of table to determine the angle of each shell at specific ranges. Could that be incorporated into the table in the designer to give a better indicator of pen performance? I can understand the reason to put off a fix in this area for now, it just needs to be fixed because obviously the tables in game are confusing everyone. 

I don't believe anyone here believes immunity zones are anything "magical". But they were a key design philosophy (especially in the later years) just like using armor thickness to match against your own gun caliber. They were the rule of thumb to "proof" designs, but if you don't have good data to utilize you can't proof anything. I think most people think of Hood when it comes to this subject. You don't want to have your nice expensive ship Hood'ed because you had no real idea at what ranges your deck armor is ineffective. 

 

[Draco]

Right @Nick Thomadis there is no immediate need to fix this before release, I concur with as much.

As with expecting a historical deck armour value like 6-8" to make you completely immune at theoretical ranges, no one ever expected that. We all agree it would be unrealistic if it worked 100% every single time.
However, currently there is no point whatsoever in equipping any deck armour at all, and it sucks.
I've had 15" guns bounce on my 0" extended deck only to then pen my 14" mid deck at the exact same range seconds later.
I've bumped up deck values to the order of 16-17" (over two yamatos worth of deck armour) only to still get penned at 15km or below by medium sized guns.

4 hours ago, Nick Thomadis said:

If you made a statistical evaluation of how many ship hits are transformed into actual full penetration deck hits, you would get values close to reality.

I vehemently disagree with this statement, and the thread is by now full of evidence to the contrary. We have done multiple analytic studies displaying that the opposite is true.
I think the true crux of the matter might be that we simply disagree on what constitutes a "full deck pen", which in regular naval discourse only constitutes penning all defensive layers to enter the citadel and cause catastrophic damage to boilers, magazines ect. If by saying "full deck pen" you are also inferring hits that pen the upper "sacrificial" plates but fail to enter the citadel, then yes such hits should "pen" inside the immunity zone but do significantly less damage than "full" pens, which, as we have done our best to display with as many in game examples as possible, isn't currently the case.

4 hours ago, Nick Thomadis said:

Would the game be more enjoyable if ships were immune, taking unrealistic zero damage by 16 inch shells?

No. As stated earlier one should expect such hits to do about a third or fourth of a full deck pen with no chance at hurting engines or magazines, so if a 16"er does 200dmg with a full pen post-modifiers, then about 50-80dmg is what we're after, and the magazines should have no chance to detonate.
Again this is not what is happening in game.

 

4 hours ago, Nick Thomadis said:

making deck armor essential in any case, because without it, any hit would cause ammo detonations, engine damage and explosions that would disintegrate a ship.

Just this morning I got a flash fire from a 15" gun penetrating 14.5" (again, almost two yamatos worth) of deck at 20km distance. It completely disintegrated the ship well inside it's theoretical immunity zone, which is exactly what you just said can't happen. I really think I am just about running out of different ways of telling you that what you're saying is the opposite of what is happening...

Apologies, I don't mean to insult anyone, I just want this game to be as good as it can be.

 

Edited March 9, 2021 by Draco

[Skeksis]

2 hours ago, madham82 said:

I think most people think of Hood when it comes to this subject. You don't want to have your nice expensive ship Hood'ed

Checkout this video…

The Loss of HMS Hood - But why did it blow up?? - YouTube

Looks plausible to me, to be “Hood’ed” maybe not such a thing (deck armor-wise). 

Edited March 9, 2021 by Skeksis

[Draco]

I just tested a battleship with no deck armour. I take back what I said, there is definetly a difference between 0" and 14" of deck armour, it reduces mean damage by about 3-400%.

I still think deck armour overall needs to be more powerful at long-very long range, but the claim that it does nothing at all I take back.

Apologies again.

[o Barão]

8 minutes ago, Draco said:

I just tested a battleship with no deck armour. I take back what I said, there is definetly a difference between 0" and 14" of deck armour, it reduces mean damage by about 3-400%.

I still think deck armour overall needs to be more powerful at long-very long range, but the claim that it does nothing at all I take back.

Apologies again.

Did you notice any critical damage to internal components when getting pen with a 14" deck armour?

[Draco]

Yes, flash fires, lots and lots of flash fires!

[Draco]

as long as you start at max range, and as long as both you and the enemy battleship can actually spot each other and find a firing solution, you will get penned at some point whilst closing the range, regardless of deck thickness, and I can recount at least three test runs so far where I have had flash fires at 20-25K distance with 14" or above of deck armour.

[o Barão]

9 minutes ago, Draco said:

as long as you start at max range, and as long as both you and the enemy battleship can actually spot each other and find a firing solution, you will get penned at some point whilst closing the range, regardless of deck thickness, and I can recount at least three test runs so far where I have had flash fires at 20-25K distance with 14" or above of deck armour.

Well that imo shouldn't happen but i am not a naval expert like Drachnifel so i can be wrong. I expected to see deck pens related to the multi level deck armour layout but with the magazines and machinery well protected. Damage yes but internal components safe.

Maybe the shell penetrated the barbette or turret top?

Edited March 9, 2021 by o Barão

[Nick Thomadis]

10 minutes ago, Draco said:

Yes, flash fires, lots and lots of flash fires!

Did you have enough protection for your turrets (Side and Top armor). Turret top is very important for keeping your guns protected at long range).

[Draco]

No it wasn't that, although I have had instances of this too, but here you asked specifically about deck pens, and so I only specifically referred to those specifically deck-pen related flash fires.

15 minutes ago, o Barão said:

Maybe the shell penetrated the barbette or turret top?

I can recall one flash fire from a turret top pen against 22.5" from an 18" heavy shell with tube powder at 27km whilst testing the theory out originally. But this was before I started taking screenshots of everything.

Edited March 9, 2021 by Draco

[Draco]

As I said in the original post, You literally need 27.5" of turret top armour to fully protect against a 20" shell at 30000m.

This scales downwards more or less in relation to the pen values presented in game, so for an 18" you need approx. 25" of turret top to reliably bounce at 30km, for a 16" you need approx. 23", ect.

Most IRL battleships had less than 10"...

Edited March 10, 2021 by Draco

[madham82]

3 minutes ago, Draco said:

As I said in the original post, You literally need 27.5" of turret top armour to fully protect against a 20" shell at 30000km.

This scales downwards more or less in relation to the pen values presented in game, so for an 18" you need approx. 25" of turret top to reliably bounce, for a 16" you need approx. 23", ect.

Most IRL battleships had less than 10"...

Right, Iowa for example was just over 7". But a hit to the top of a turret was rarer than rare. The chance of that hit resulting in a flash fire that would sink the ship is even more rare. Only gross negligence like at Jutland by the British would result in that occurring. 

As far as the numbers, if Iowa's super heavy AP can't pen more than 14" of steel at max range/angle, why do we need 23" on the top? 

[Draco]

Don't get me wrong it's a REALLY GOOD system for dreadnought era battleship combat, one of the best and easiest to comprehend I've ever seen for such a complex topic, it could just use a liiiittle bit of balancing at the long-to-very-long range spectrum, where flash fires become abnormally common due to the skewing of deck and turret top armour efficiency as of right now. You don't have to radically change anything about the system itself except adjust the rate at which deck pen values overtake belt pen values by a few kilometers here and there. That's all I'm saying.

[Draco]

One more rant and I promise I'll stop.

Back in 0.7 the "US super battleship" mission had Mk.V guns available if you went with "boost firepower", and they ran all the way up to 18". Those things could shoot like 45km and start hitting reliably at 35km, and if you went with full scope upgrades and the best towers you could reliably begin spotting the japanese battleships at 30-35km if they were big enough.
Back then plunging fire was, well maybe a bit less OP than rn, and you could get away with like a 14" deck and a 18" turret top and still be more or less immune at around 30km dist. however the AI ship designer would usually run historical armour values like 6-8" of turret top and deck, and so logically if you just stayed at 30km range, you'd start getting hits within the first dozen salvos, and start getting flash fires within the first dozen hits. I had a lot of fun with that. Not saying it was perfectly realistic or even desirable, but it was a lot of fun, and it made me aware of the dynamic.
but since then it feels like there's been a gradual creep towards the current state of things, which really took a jump when 20" guns were introduced, and when you combine it all, maximum bulkheads being so heavy now, 20" guns penning any realistic sized deck/turret top at range, their even longer max range as compared to the old king the 18", the fact that the deck/turret armour also seems to be heavier now pr. inch you add, and finally the fact that if the enemy fleet has oxygen torpedoes, you may very well be forced to stay 20km away from them to keep safe anyway. It makes it all very hard indeed to make even your turret tops immune, much less your entire deck.

All the deck pen flash fires I've had have all been very close to the turret that then blew up a few seconds later and had a trajectory pointing toward the barbette and therefore magazines of that turret, and btw I love that this can happen in game and please don't remove that feature, but maybe make sure that if it cannot pen the turret top, then it can't pen the barbette under the deck either, and if this is already the case, then we are back at turret top and deck armour underperforming rn.

a 14" turret top should make you more or less immune to 20" fire almost all the way out to it's max range of 40 some kilometers, not start to get a chance to deflect shots at 20km.

The system and it's features are awesome, but currently skewed in favour of plunging fire, making even the most heavily armoured battleships sitting ducks above 20km where you need 20 or above inches of turret top to prevent them from going up like matches from a single lucky hit.

Apologies for not being able to summarize it better, it is a slightly complex topic after all, and I wish I could have gotten the points across with less text for you to read through.

Drach could probably have explained it in a much shorter time.

Edited March 10, 2021 by Draco