Gun destroyed?

[Dr Dan]

O.K., in accordance with forum bylaws, as well as currently accepted etiquette, I did a search on this but couldn't answer my question:

How does CM model 'Gun Destroyed?'

I understand the method used to determine armor penetration, but I can't find anything that indicates how gun hits are determined and what makes a 'hit' become a 'destroyed.'

Dan

CM is STILL fun!

[Slapdragon]

I have been testing gun damaged and am running in on a formula -- but I can tell you what are likely the variables. The size of the weapon is the big one -- and next it has to not kill the tank it hit. Gun damaged are rare on Shermans because they die like red shirts on Star Trek, they are much more common on German monsters because they can often take a blizzard of hits. Hits by large HE weapons which rarely penetrate will sometimes kill the main armament.

What I assume is that the gun might be damaged, but it may also be the optics or the positioning gear, either of which cooks the coaxial mg.

In testing the 76 against the Panther about 1 hit in 40 resulted in a weapon killed at 800 meters using standard AP. A 95 HE fired at a Hetzer resulted in one chance in 25 of killing the main armament, while a 150 fired at a Jumbo was 1 in 30. Closing the distance was harder to test because kills increaded, but a 95 at 20 meters was an 25% chance to kill armament and another 20% to kill the Hetzer 00 but this test was just to see what would happen because 20 meter hits by a 95 are going to be rare in daylight.

The variable is complex enough that it resists simple tests to decode, so there may actually be a lurking variable (like range) and maybe a third variable (like armour or something like that).

[This message has been edited by Slapdragon (edited 10-03-2000).]

[Guest *Captain Foobar*]

Redshirts... LOL!!

"Ensign Johnson, go look over that hill over there...."

"AAIiiiiigggghhhh"

[Schugger]

LOL!!!

* Bones kneels beside another nameless Redshirt, feels his pulse and shakes his head*

" He's dead, Jim!"

[M Hofbauer]

LOL...

yes, and minutes later Kirk is hit *repeatedly* by same weapon, Bones frantically goes to work, Kirk is immediately beamed up and revived.

Slapdragon, when putting together data for your formula research, please make sure to include the 50cals, as I have repeatedly stated before, my first QBs ever in CM involved a KT which got a "Gun Hit" from a pathetic M3 HTs 50cals in the first turn. Make sure your formula can account for that, too, I have been traumatized forever.

------------------

"Say i think u all need to chill out." (GAZ_NZ)

[This message has been edited by M Hofbauer (edited 10-03-2000).]

[Dr Dan]

Slapdragon, thanks for your observations. What you describe about heavy German tanks certainly appears to be true, although I have seen 'gun destroyed' a couple of times even on a Sherman. Also, large HE or similar IS a factor; it seems that tanks like the Cromwell VI or the Churchill VIII, which use HOLLW © ammo, are particularly good "gun destroyers." As for what actually is taken out, I think Steve said earlier that 'gun destroyed' covered more than just the barrel, including the sights/optics, mechanical failures, etc.

Hey, Steve, BTS, can you guys comment? One thing that I would really like to know is where you can find raw data that indicates specific types of tank damage throughout the war. I've tried searching ( ) the web, but I'm not exactly sure what to seek. Any WWII historians know where to look?

Dan

CM is STILL fun!

[Schugger]

I still bet that amour quality has something to do with the numbers of gun damages.

Slapdragon, when you run your testings I would like to know if you've made the same observations.

[Mannheim Tanker]

I recently had a Panther G taken out - from the front! - by a friggin' M8. My opponent was using the rarely seen, but terribly effective Uberscoutcars.

------------------

"Gun damaged are rare on Shermans because they die like red shirts on Star Trek" - Slapdragon

[Schugger]

Sh*t! I knew someone would make a sig out of this one.

[Mannheim Tanker]

Beat you to it?

------------------

"Gun damaged are rare on Shermans because they die like red shirts on Star Trek" - Slapdragon

[Dr Dan]

Topic Drift --

Maybe Sherman tanks can survive the same way SOME red shirts did on Star Trek -- they need to hang out with the engineers. (Aye, that and a wee bit more, Captain). As the historical reality folks might remind us, combat engineers had much lower casualty rates than (most) other unit types.

Dan

CM is ALWAYS fun!

[Slapdragon]

Ok -- I tested the .50 cal against the King Tiger for armament hits.

First, I lined up 12 KTs and 12 HMGs. The HMGs were placed in buildings because on the first test they were slaughted by the KTs without ammo. The KTs were dug in to encourage weapon hits. I ran the test at 3 ranges, a mean of 51 meters, a mean of 192 meters, and a mean of 398 meters. Each range got 4 repetitions for 48 trails. I used my standard range (available on request but maybe not with the right units since I use it over and over.)

In 10 turns of firing at 51 meters with 48 individual trials 1 king tiger lost its armament. At 192 amd 398 I observed no loss of armaments. With the high standard error of these tests I can conclude that at 51 meters you will have a KT - A/Kill every 400 to 440 turns of firing by 1 HMG. The other tests were inconclusive because my standard error exceeded my observed results, and it may be possible to A/Kill with a 50 cal at that range but it will be so seldom that it would take 400-500 runs to determine a close order hit.

A repeat of the test with an M3 Stuart gave 7 g/kills in 48 runs at a range of 192 meters and 1 completely surprising k/kill that I could not create again in over 160 more repeats. Here note that the very high ROF of the stuart means that it was firing off most of its AP in 4-5 turns, so the 37 may only be slightly more effective than the .50 cal.

So, Hofbauer, you are the victim of a cosmic accident -- it is possible with a 50 cal kill on a KT, I observed 1, but it is extremely highly unlikely you will see it again. If you do, then you may consider Las Vegas gambling as a method of winning big buck, betting against yourself of course .

An odd thing about my testing is because higher penetration to armour ratios result in kills, you rarely see gun kills on lightly armoured tanks, or caused by heavy duty weapons like the 88L71 or 128mm cannon. They just plain old kill the tanks involved, which of course is also a A/Kill.

So far my formula looks like this:

Weapon power (in FP points) accounts for .4 of observed effect which is the greatest correlation in a nonlinear relation: ie -- get to powerful and it just kills the tank.

Range accounts for .2 of observed effect also with a nonlinear relationship (ie. to close and you kill the tank).

Armour has .1 observed effect with a very confusing blip at 60-80mm

My residuals in all case are fairly high, telling me I am missing something somewhere.

Now to flame my own results. I am observing nonlinear results that do not fit a good scatter line because of the fact that guns that make a kill could have also knocked out the armament, but the kill does not tell me if they did. Armour at one point has a negative correlation, then has a positive correlation, making me work with nonlinear stats. I am fairly confident that the two variables above are big factors. I am not confident yet on armour since it may be a result of the lurking variable of killing power.

My bet is if I could hack the application model I would find a 2 variable "does it kill the gun" loop in the hit effects model, effected of course by the "does it kill the tank" question. If there truly is a 3 variable confounding me, it is likely to be a small one, or several small ones that my testing is not accurate enough to hack out.