v90 Guide: HE vs. AP

[Evil4Zerggin]

Penetration types

• Ricochet: Shells have up to a 60% base chance to ricochet at glancing angles. However, if the penetration exceeds twice the armor (including angle and tech), it will never ricochet. HE shells also never ricochet.
• Overpenetration: If the penetration is more than 60x the armor then the shell overpens, dealing 35% damage. There's an extra modifier to the overpeneration threshold for ship size but it appears to generally lie in the range of +/- 40% even for very large or small ships. Armor is always treated as being at least 1 mm for this calculation.
• Penetration: Otherwise, if the penetration exceeds the armor, the shell does a full penetration for 100% damage.
• Partial penetration: Otherwise, if the penetration exceeds 18.5% of the armor thickness, the shell does a partial penetration for 33% damage.
• Shatter: Otherwise the shell shatters for no damage.

Overpens and full pens are required for the following effects---richochets, partial pens, and shatters cannot cause these:

• Internal fire. (Any deck hit can cause a deck fire.)
• Flooding.
• Detonation.
• Flash fire.
• Module damage (engine, rudder, etc.)

AP vs. HE

AP is considered the "normal" shell type. Compared to this:

• HE does 1.35x base damage, and is affected by additional modifiers from Shell Charge components.
• HE has 25x chance to cause fires, also affected by additional modifiers from Shell Charge components.
• HE has only 25% the penetration.
• HE is not affected by azimuthal angle. This is why it sometimes works better than AP in a chase.
• HE never ricochets.

Auto ammo selection

Auto ammo selection works as follows:

• Compute whether it's more likely to hit the belt or the deck.
• Compare the ratio of penetration to the corresponding armor thickness (using the mid section).
• If the result is between 0.3 and 10, use AP. Otherwise, use HE.

AFAICT this doesn't take azimuthal angle into account, i.e. it assumes the target gets hit broadside or by HE. It also doesn't take bow/stern armor into account, nor superstructure/turrets.

Commentary

As discovered in the main thread, auto shell selection can result in suboptimal choices if the mid section has considerably heavier armor than the bow/stern, which may be overpenned by the same AP shell.

The recent patches really cut down on overpenetration. I believe the minimum 1 mm effective thickness corresponds to around 1-3" penetration threshold for overpenetration, but even 1" of real armor would be more than 25x as much. So if you go all-or-nothing, make sure that the "nothing" literally is nothing.

Under the current mechanics, I think it would have been better strategy to reduce the upper auto ammo threshold to 4-5, slightly above the reciprocal of the HE penetration multiplier. Some way of accounting for fore/aft would be good as well; the simplest method would be to take the most likely similar to how deck/side is handled.

By the logic of not being affected by azimuthal angle, it would seem like HE shouldn't be affected by angle of fall either. However, I'm not sure this would be more realistic on net and it could upset game balance.

Edited October 5, 2021 by Evil4Zerggin

[Skeksis]

...

Edited November 9, 2021 by Skeksis

[Evil4Zerggin]

Given that it's not feasible to carry enough armor to fully shatter BB shells, I would just go for effective armor = expected incoming penetration plus 10% or so to account for the random factor in penetration. This will at least reduce the damage by 3x compared to full pens and protect you from the extra effects of penetration (internal fire, flooding, flash fire, detonation, module damage---edited OP with this info).

[Skeksis]

...

Edited November 9, 2021 by Skeksis

[Evil4Zerggin]

18.5% is the threshold between shatter and partial pen. You want 100% to downgrade full pen to partial pen. Also, if you're solving for armor thickness you need to divide rather than multiply---you have the overpen correct but the others flipped. So against 100 mm penetration:

• < 1.67 mm armor for overpen (ignoring size factor)
• > 50 mm to allow possibility of ricochet
• 1.67 - 100 mm for full pen
• 100 mm - 540.5 mm for partial pen (i.e. 100 mm / 0.185 = 540.5 mm)
• > 540.5 mm for shatter

[Skeksis]

...

Edited November 9, 2021 by Skeksis

[Evil4Zerggin]

5 minutes ago, Skeksis said:

that's alot of armor? What am I missing?

You're not missing anything---it's very difficult to fully shatter BB shells (unless the shell hits with a lot of angle but IMO that's not to be counted on). Partial pens is the best you can hope for in most cases.

[Steeltrap]

33% for partial pen on the basis of exceeding 18.5% armour thickness strikes me as 99% bollocks.

I think I've posted this before re HE. The effectiveness of HE against ANY decently armoured target seems always to be grossly overstated IMO. Most certainly is in this game, sometimes hilariously so in earlier versions (no idea how it is now as I never play uber ships and I don't fire HE at armoured targets).

For example, a 16" 1900-lb Mark 13/14 instantaneous-nose-fuzed ("PDF") High Capacity (HC) US Navy WWII shell cannot penetrate 3" of homogeneous armor!!! It makes a big dent, but the shell destroys itself before its nose tip can move more than a few inches forward - against a 17.3" Class "A" plate the effects would hardly be noticeable (definitely no dimple in that circa 9" of rigid face layer!!!).

Source: Nathan Okun in an article on IJN Kirishima's hit on USS South Dakota  http://www.navweaps.com/index_tech/tech-092.php

Edited October 17, 2021 by Steeltrap

[SpardaSon21]

On 10/16/2021 at 7:27 PM, Steeltrap said:

33% for partial pen on the basis of exceeding 18.5% armour thickness strikes me as 99% bollocks.

I think I've posted this before re HE. The effectiveness of HE against ANY decently armoured target seems always to be grossly overstated IMO. Most certainly is in this game, sometimes hilariously so in earlier versions (no idea how it is now as I never play uber ships and I don't fire HE at armoured targets).

For example, a 16" 1900-lb Mark 13/14 instantaneous-nose-fuzed ("PDF") High Capacity (HC) US Navy WWII shell cannot penetrate 3" of homogeneous armor!!! It makes a big dent, but the shell destroys itself before its nose tip can move more than a few inches forward - against a 17.3" Class "A" plate the effects would hardly be noticeable (definitely no dimple in that circa 9" of rigid face layer!!!).

Source: Nathan Okun in an article on IJN Kirishima's hit on USS South Dakota  http://www.navweaps.com/index_tech/tech-092.php

Agreed.  I've seen some very odd effects when firing pure HE against heavy armor in custom battles.  Apparently you completely destroy a ship's machinery rooms with nothing but HE impacts to the belt... and somehow not a single lick of water will have been let inside the ship at the same time!

[o Barão]

"HE has only 25% the penetration."

This is 25% the gun caliber in all distances, or 25% the penetration in comparison to AP performance?

[Evil4Zerggin]

25% the penetration in comparison to AP performance.

There have also been numbers/mechanic changes since v90 when this was written, so this may not hold exactly today.