Destroyed tanks due to unknown reason

[Wiedrock]

Elements disrupted in a battle will recover at the end of that action. Some extra fatigue will be assigned to reflect the impact of light wounds or limited damage to equipment.

In the attached file you can find 9 entries for "KV-1 M1940", those are:

Code: Select all

3,6,KV-1 M1940 disrupted by 37mm Cannon BK3.7 (air)
3,6,KV-1 M1940 disrupted by 37mm Cannon BK3.7 (air)
3,6,KV-1 M1940 disrupted by 37mm Cannon BK3.7 (air)
5,6,KV-1 M1940 disrupted by 37mm Cannon BK3.7 (air)
9,6,KV-1 M1940 disrupted by 37mm Cannon BK3.7 (air)
9,6,KV-1 M1940 disrupted by 37mm Cannon BK3.7 (air)
11,6,KV-1 M1940 disrupted by 37mm Cannon BK3.7 (air)
19,6,KV-1 M1940 disrupted by 37mm Cannon BK3.7 (air)
23,6,KV-1 M1940 disrupted by 37mm Cannon BK3.7 (air)

As you can see, there was no KV-1 tank damaged or destroyed in all those GA-UNIT-Missions.

For some reason the System still counted 1 KV-1 killed.

"Whooo let the tanks out?"

battle_result_KV-1.png (1.1 MiB) Viewed 1241 times

BK3.7 throwing pens at tanks.

KV-1_arm-pen.png (194.74 KiB) Viewed 1241 times

Since the BK3.7 (at it's current unhistoric pen-values) can't penetrate a KV-1 anywhere...
... I wonder how that KV-1 found it's end.

[DarkHorse2]

You can't expect those pen / armor numbers to actually mean anything, can you?

That is all window dressing, it appears.

Window dressing -
* the arrangement of an attractive display in a shop window.
* an adroit but superficial or actually misleading presentation of something, designed to create a favorable impression.

I have experience the same thing.

[Denniss]

BK 3.7 has an armor penetration rating of 69

[Wiedrock]

It doesn't. That's why it will never Damage/Destroy a tank with more than ~10mm-13mm (need to test the 'max' treshold for 6mm-pen, 13mm is a guess) that of armor.

[Denniss]

you may want to closely examine your attached file and you'll find several destroyed lighter tanks which clearly state pen 69 and the armor rating of the plate it fired against.
for the OT-133 type its atttacks are mixed at top and front armor but have not seen anything vs side
A T-26 is simply reported as destroyed by the BK 3.7 but no further details are given
Vs BA-10 it seems to prefer top armor but some attacks reported vs front armor plate, again nothing vs side
Vs T-26 all damage/destroy attacks are reported vs front armor plate
all of this may point to a problem of selecting proper attack vector for certain tank types (or the combat data sample is too small for conclusions)

[Wiedrock]

As title states I am more concerned about Tanks not being damaged/destroyed, being destroyed.

Edit: FYI, attached is a file which shows 3 AOGs of Ju 87Gs attacking a tankdivision with 500 KV-1's + 500 Rifle Squads + 500 Support being shot at with BK3.7 with "switched values" for the BK3.7, so 69pen and 6 anti-armor. Sice you are referring to anti-armor being the "pen" in the combat logs. The highest of the two values may be used to "log" it as "pen" in the logs, while they are obviously being handled completely different by the combat-system.
There must be some "mishmash" going on between (at least) those two values which just isn't as intuitive as we'd like it to.

[Denniss]

something changed in this log, it does not show a static unmodified penetration value of 69 but now it differs with range - the longer the combat range the less penetration value is shown. This is confusing why it does use correct dynamic range-modified values in this attack but used wrong static values in the previous log.

[Wiedrock]

Yea I am confused as well.

If "you read over it" in this recent log the vanilla pen & anti-armor-values were exchanged by me editing the "Device"!

When something is penetrated (how I interpret penetration, so penetration > armor (or little under the armor value, so maybe "partially penetrated")) then there is always this "range" taken into account.

When something is using "your penetration" of anti-armor this never is shown + the enemy is losing tanks (at a MUCH lower rate) that are not shown in the log (no damaged/destroyed in log) while still losing tanks on ground (like the "blue tank" in my first post).


Added to this confusion there is the Range value of plane-guns (but let's leave that out for now, lol).

[Denniss]

can you re-run this latest version with T-34/40 and T-70 in the target unit?
I'd like to see target armor plate selection distribution, especially the T-70 should be penetrated from all directions

[Wiedrock]

can you re-run this latest version with T-34/40 and T-70 in the target unit?
I'd like to see target armor plate selection distribution, especially the T-70 should be penetrated from all directions

Here we go.

I am overall not that concerned about "which plate the plane shoots at", since I find it realistic to randomly choose a "side of the tank".


Concerning the initial reason of this post (Ghosting tanks). If I do the same mission (with 500 KV-1) with vanilla Ju87G's I get like 15 KV-1s killed BUT none of them is tracked in 1.combat log, 2.battle reports, 3."AD-Results-window" ....I can only see those 15 killed KV-1's in the "Ground Losses - Axis-Airphase"-window! ...

[Wiedrock]

As mentioned I managed to replicate "Ghosting tanks" not tracked to be killed in any screen but the "Losses Screen".

Ghosting_Tanks_result.png (673.33 KiB) Viewed 1101 times

As you can see there are no AFV/Tanks being killed. Same goes for the other Ghostroyed Elements.

While the Losses screen shows that there was something lost, due to which reason?!

Ghosting_Tanks_losses.png (254.27 KiB) Viewed 1101 times

[Denniss]

Re-visiting the A2G reports made me think game engine is using ground combat code for those attacks so Anti-Air/Soft/Armor would be HE effects and Penetration/HEAT/HVAP would be AP effects.
That's the only sense I can make of the static penetration shown for the anti-armor data and the variable-by-range penetration shown for the test with modified (AP) penetration.

[Joel Billings]

I recently saw a save where there were a few ground elements destroyed in the enemy air phase, when the only missions flown were recon. There was flak firing, but no bombing. I will ask Gary, but I know there are routines in ground combat that can kill or damage elements not due to enemy fire, but due to "mechanical breakdowns", mines, etc. Now exactly how and when these die rolls happen is a mystery to me. It's possible something is triggering these kinds of die rolls during the air phase, maybe even by units firing flak. That's the only explanation I can come up with, unless it's some kind of reporting error that is reporting losses previously taken. Assuming your tests above are indicative of real, new, losses that happen during the air phase, that would make me think it's due to these other ways to lose a unit in combat. One thing we've never modeled is friendly fire, which would be a major factor in ground combat, but not so much in air combat. There could be a bug in there, but if so it could be very hard to get at.

[Wiedrock]

"mechanical breakdowns"

From another combat log: "1,6,Panther G breaks down in combat"

At least this line is not being tracked in the log file for those KV-1's. I'll edit the reliability for the KV-1's to be worst/awesome and retest and see if the loss ratio decreases/increases.

[Joel Billings]

Do you have a save where you get the losses showing only on the left of the losses screen that I can run and see this happen (at least a decent amount of the time)? Or is this something that you can never recreate from a save? If we've got a save where it happens often enough, Pavel might be able to track it down.

[Wiedrock]

Do you have a save where you get the losses showing only on the left of the losses screen that I can run and see this happen (at least a decent amount of the time)?

It's no specific save, you can recreate it in GC 41.

1. delete all ADs
2. set up a GA mission (Luftflotte 2) on Hex 177,161 (It's a SWAMP with lots tanks in it)
3. run the one AD
4. count tanks/compare

As an example here the numbers for the T-34 M1941 of which two got officially destryoed, while 10 Ghosted away. NOTE that the "Last Action" seems to always represent the "ghosted" stuff

T-34 M1941_GC 41 T1_AD-GA-UNIT_Hex 177-161.png (419.94 KiB) Viewed 1032 times

LAST ACTION.png (1.44 MiB) Viewed 1032 times

So obviously, ghosting happens in/during/while/after Last Action, obviousl also in my previous posts, this fact just did not make it into my attention as of now.

[Wiedrock]

"mechanical breakdowns"

Using my initial 500 KV-1 setup with Ju 87G (unable to kill any tank with their BK3.7) in Clear Terrain and

• setting the Reliability of KV-1's to 100 gives 6 Ghosted tanks
• setting Reliability to 9000 gives 7
• my initial test with vanilla ~5k Reliability gave 9 ghosted KV-1's

....so, IF it is related to breakdowns, those breakdowns do not seem to be calculated using the Reliability value (at least in clear terrain).

Here how it looked with 9k reliability:

KV-1_9k-reliability.png (435.78 KiB) Viewed 1025 times

[Joel Billings]

I've done as you suggested, and in that 1 attack, I'm seeing the losses screen (left side and right side for the air phase equal each other (ran it 5 times). I am seeing the amount listed in the battle report being lower, so clearly some losses are happening that are not listed in the line battle report. I'll ask Pavel to look into that.

[Joel Billings]

Gary, Pavel and I discussed the original question re the 37mm penetration vs anti-armor rating in A2G combat. To say the code is confusing is an understatement. Gary thinks the penetration rating was never intended to be used in A2G combat, as the numbers are all very low values. He and Pavel think they could see where it might get used in A2G combat, but it wasn't clear exactly when or how often it would use it. Gary didn't think the range should even be a factor in this kind of strafing A2G attack. It was not clear to me from the posts above whether you were saying the that range was impacting the value or it wasn't, as I lost track of what you were saying about your tests and your changing the values. Did you post a save before one of your bombing attacks with the 37mm? It would be good to have that save (with the standard data). If I have a save, I will ask Pavel to try to look into adjusting the formulas so it's using the anti-armor value.

As for the unknown loss issue, I've given a save to Pavel and asked him to look into it. Thanks for all your detailed tests.

[Wiedrock]

As for the unknown loss issue, I've given a save to Pavel and asked him to look into it.

Sounds good, that's the issue this Thread was supposed for.

---

Gary, Pavel and I discussed the original question re the 37mm penetration vs anti-armor rating in A2G combat. To say the code is confusing is an understatement. Gary thinks the penetration rating was never intended to be used in A2G combat, as the numbers are all very low values. He and Pavel think they could see where it might get used in A2G combat, but it wasn't clear exactly when or how often it would use it. Gary didn't think the range should even be a factor in this kind of strafing A2G attack. It was not clear to me from the posts above whether you were saying the that range was impacting the value or it wasn't, as I lost track of what you were saying about your tests and your changing the values. Did you post a save before one of your bombing attacks with the 37mm? It would be good to have that save (with the standard data). If I have a save, I will ask Pavel to try to look into adjusting the formulas so it's using the anti-armor value.

The whole topic about A2G penetration values came from my post here.

But I will supply you with some quick points which should be able to clarify the issue.

1. many/all Aicraft cannons have their historical Penetration values put into the Anti-Armor
   ...so instead of penetrating ~60mm the gun only penetrates ~6mm and so on.
2. This can be fixed by simply exchanging Penetration and Anti-Armor values with one another
3. I can't make any statements without confusing you/going in depth concerning the range values of aircraft cannons (used in A2A and A2G(when penetrating)).
   But range is strange.

Example Devices affected by this Issue:

• 37mm Cannon BK3.7 (ID: 48)
• 30mm Cannon MK103 (ID: 46)

Attached is my testing setup + pre save + post save, testing Ju 87G's with BK3.7 Cannons aganst KV-1 M1940 Tanks with a top armor of 35mm, where you will never see any kill (since the BK3.7's pen in game is only 6mm). ...only the "Ghosting Tanks" may occur to be lost.
Just click F12 to activate the AD which is set up.